Está en la página 1de 615


PI20=1 P >=1 PI20=P P <1 PROGRAMME
DAY 1 - 12 th April, 2011 (Tuesday)

ISBN 978-967-5418-17-4


Professor Ir. Dr. Mahyuddin Ramli

Associate Professor Dr. Hassim Mat

Associate Professor Dr. Kausar Hj. Ali
Associate Professor Dr. Nurwati Badarulzaman

Committee Members
Associate Professor Jamel Ariffin
Dr. Ahmad Puad Mat Som
Dr. Azizan Marzuki
Dr. Norazamawati Md. Sani@ Abd. Rahim
Dr. Nor Zarifah Maliki
Abdul Ghapar Oth man
Nor A’zam Shuib
Subramaniam Govindan

DAY 1 – 12 th April, 2011 (Tuesday)
Venue: Conference Hall
School of Housing, Building and Planning (HBP), USM

Registration of Participants

08.30 am – 09:15 am Arrival of Guests of Honour

Recital of Doa and Welcoming
Remarks by ITRC 2011 Chairman
Associate Professor Dr. Hassim Mat
09.15 am – 09.30 am
Opening Remarks by Professor Dr.
Susie See Ching Mey,
Deputy Vice Chancellor Universiti
Sains Malaysia, Industry and
Community Network
Keynote Address 1
Mr. Khairil Anwar Abu Kassim Head
09.30 am – 10.30 am
of Crash Safety Engineering Unit,
Vehicle Safety and Biomechanics
Research Centre, Malaysian Institute
10.30 am – 11.00 am
of Road Safety Research (MIROS)
Refreshment Break

11.00 am – 1.00 pm Parallel Sessions 1 & 2

HBP Conference Hall & Resource Centre
1.00 pm – 2.30 pm Lunch

2.30 pm – 5.30 pm Parallel Sessions 3, 4 & 5

HBP Conference Hall, Resource Centre & Viva Room

5.30 pm – 6.00 pm Refreshment Break

09.00 am
10.00 am -- 09:00
10.30 am
10.00 am


DAY 2 – 13 th April, 2011 (Wednesday)

Venue: Conference Hall

School of Housing, Building and Planning (HBP), USM

Opening Remarks by Master

of Ceremony

Keynote Address 2
Mrs. Umairah Saad Senior
Human Capital Manager Rapid Penang

Refreshment Break

10.30 am – 12.30 pm Parallel Sessions 6 & 7

HBP Conference Hall & Resource Centre
Closing Ceremony by
12.30 pm – 1.00 pm the Dean, School of Housing,
Building and Planning
1.00 pm – 2.30 pm Professor Ir. Dr. Mahyuddin Ramli


2.30 pm – 5.00 pm Rapid Penang bus trip

(Open to all participants)

Refreshment Break
5.00 pm – 5.30 pm End of Programme

DAY 1 – 12 th April, 2011 (Tuesday)
Time: 11.00 am - 1.00 pm PARALLEL SESSION 1
HBP Conference Hall
Theme: Transport Operation and Management

CODE Paper Title and Authors

Simulation Of Traffic Operation And Management At

001 Malaysian Toll Plazas Using Vissim by Ahmad Hilmy
Abdul Hamid (p.13)

Operational Efficiency And Automation In Public

Transport Through Integrated Transit Management
System (Itms) – A Case Study Of Ahmedabad, India
by Vivek Ogra and Aurobindo Ogra (p. 22)

Parking Pricing And Its Effect On Urban Traffic

Management System Specially Traffic Congestion by
Hamed Eftekhar and Riza Atiq Bin O.K. Rahmat

Strategic Option Of Project Consultant Procurement

Between Single And Multiple Contracts: Case Study Of
MRT Construction Project by Athiwat Noonma and
Vachara Peansupap (p.43)

Cargo Flow In Malaysia: Analysis Of Current Status

040 On Federal Roads by Mimi Suriani Mat Daud, Intan
Rohani Endut and Harlina Suzana Jaafar (p.53)

Driver`s Responsiveness To Congestion Pricing Policy

070 by Kian Ahmadi Azari, Sulistyo Arintono, Hussain
Hamid and Riza Atiq O.K. Rahmat (p.64)

DAY 1 – 12 th April, 2011 (Tuesday)

Time: 11.00 am – 1.00 pm PARALLEL SESSION 2

HBP Resource Centre
Theme: Environment and Sustainability
CODE Paper Title and Authors
Analysis Of Travel Time Variability For Multi-Lane
Highways by Khoo Hooi Ling ( p.76)

Yes We Can ! Reduce Road Congestion And Co2

009 Emission By Introduction Of A New Intermodal Logistics
Chain by Natasa Gojkovic Bukvic (p.85)

Comparative Assessment Of Trip Distribution In Skudai

021 Town, Malaysia by J. Ben-Edigbe and A. Pakshir

Traffic Flow Improvement At Signalised Intersections By

Coordinating Signal Phases In Batu Pahat, Malaysia by
Zareda Abu Bakar, Ismail Yusof and Mohd Erwan
Sanik (p.107)

To Promote Future Sustainability With Integrated Design

049 Of Urban And Transportation System by M. Z. Maleki.
and M. F. M. Zain ( (p.117)

The Assessment Of Significant Aspects And Impacts At

Highway Construction Towards Sustainable
054 Development by Alea Wahida Ismail, Sumiani Yusoff
and M. Rehan Karim (p.126)

DAY 1 - 12 th April, 2011 (Tuesday)

Time: 2.30 pm – 5.30 pm PARALLEL SESSION 3

HBP Conference Hall
Theme: Road Safety

CODE Paper Title and Authors

Investigation Of The Effects Of Rainfall And Traffic On

012 Road Accidents by Md. Mahmud Hasan, Shamas
Bajwa, Edmund Horan and Edward Chung (p.145)
Design And Development Of Smart Motorcycle Safety
Vest (Sms-V) For Motorcyclist In Malaysia by Mohd
033 Farriz Basar, Khalil Azha Annuar, Firdaus Ab Halim,
Norhaslinda Hasim and Muhamad Khairi Aripin
Study On Reliability Of Wire Rope Installation Along
036 Malaysian Roads by Rashidah Ab. Rashid, Ismail
Yusof and Mohd Erwan Sanik (p.166)
Prediction Of Accident Trend At Two-Lane Federal
Highways Using Statistical Approach by Cik Wan
037 Norbalkish Jusof, Ismail Yusof And Mohd Erwan
Sanik (p.178)
Centerline Rumble Strips: A Review Of Application And
Effectiveness by Mohd Hanifi Othman, Zaiton Haron,
038 Khairulzan Yahya, Haryati Yaacob and Shamila Azman
Shoulder Rumble Strips: A Review Of Application And
Effectiveness by Mohd Hanifi Othman, Zaiton Haron,
039 Khairulzan Yahya, Haryati Yaacob and Shamila
Azman (p.203)
Investigation Of Type Of Damages Occurred On
Malaysian Federal Road Route One (Ft01) At Selangor
042 by Nurul Elma Kordi, Intan Rohani Endut, Bahardin
Baharom and Md Yunus Ab Wahab (p.217)
Using Finite Element Analysis To Determine The
Increasing Of Axle Load Factor Due To Increasing Axle
076 Load Limit by Osama Mahmoud Yassenn, Mohamed
Ahmed Hafez,Md Yunus Ab Wahab, Intan Rohani
Endut and Bahardin Bin Baharom (p.228)
DAY 1 - 12 th April, 2011 (Tuesday)

Time: 2.30 pm – 5.30 pm PARALLEL SESSION 4

HBP Resource Centre
Theme: Pedestrian and Public Transport

CODE Paper Title and Authors

Study Of Bus Lane Layout Efficiency In Tehran, Iran by

015 Amiruddin Ismail, Mohammad Hesam Hafezi and
Omran Kohzadi Seifabad (p.243)
Study On Organizing And Improving Walkways In City
Centre Of Yasuj, Iran by Amiruddin Ismail And Omran
016 Kohzadi Seifabad and Mohammad Hesam Hafezi
(p. 256)
Developing A Methodology To Evaluate Impact Of Road
Infrastructures On Different Modal Trips – A Scenario
075 Based Study by Mukti Advani and B.Kanagadurai

Practical Evaluation Method For Pedestrian Level Of

078 Service In Urban Streets by Zohreh Asadi Shekari and
Muhammad Zaly Shah (p. 280)
Student Intercity Travel Characteristics By Stated
Preference Method: A Case Study For Intercity Travel
081 Between Parit Buntar, Penang And Kuala Lumpur by
Angelalia Roza, Bayu Martanto Adji, Raja Syahira
Raja Abdul Aziz and Mohamed Rehan Karim (p.293)

Potential Of Bicycle As Transportation Mode For

Activities Around Campus by Bayu Martanto Adji,
Angelalia Roza , Raja Syahira Binti Raja Abdul Aziz
and Mohamed Rehan Karim (p.304)

Modes Of Transport Choice And Its Dependency In

084 Pulau Pinang, Malaysia by Irin Caisarina and Hassim
Bin Mat (p.3 14)

Travelling Characteristics of University Students' Bicycle

098 Excursion to Cultural Heritage Sites: A Case of Lijiang,
China by Xing Huibin and Azizan Marzuki (p.325)
DAY 1 - 12 th April, 2011 (Tuesday)

Time: 2.30 pm – 5.30 pm PARALLEL SESSION 5

HBP Viva Room
Theme: Issues in Transportation

CODE Paper Title and Authors

Telecommuting’s Potential Contributing To Reducing

053 Traffic Congestion: A Malaysian Perspective by Diana
Mohamad and Matthew W. Rofe (p.340)

A Novel Road Extraction Method From Satellite Images

As An Effective Step Toward Digital Map Generation,
055 GIS And Intelligent Transportation System by B.
Yousefi, Seyed Mostafa Mirhassani, M.Soltani, M.J.
Rastergar Fatemi and H. Hakim (p.353)

Halal Transportation Technologies For Malaysian Halal

064 Logistic by Raziah Noor Bt Razali, Mohamad Iskandar
Bin Illyas Tan and Mohammad Ishak Desa (p.365)

The Needs Of Halal Transportation Control In Malaysia

by Zuhra Junaida Bte Mohamad Husny, Mohamad
Iskandar Bin Illyas Tan and Zaly Shah Bin
Mohammed Hussein (p.378)

The Possibility Of Implementing Road Pricing Policy In

083 Georgetown, Penang, Malaysia by Abu Bakar Bin Mat
and Hiroo Ichikawa (p.388)

A Productivity Analysis Of Medium Seaport Container

Terminal by Kasypi, M. and Shah, M. Z. (p.398)

The Assessment of Visual Impact On Highway

Landscape Case Study: Butterworth-Kul i m Expressway
(BKE) Highway by Munira Zainol Abidin and Jamel
Ariffin (p. 417)

Why Motorcyclist Neglect Helmet? A Study From

031 Psychological Perspectives by Nur Sabahiah Binti
Abdul Sukor and Mohd Ezree Abdullah (p. 430)
Time: 10.30 am – 12.30 pm PARALLEL SESSION 6
HBP Conference Hall
Theme: Transport Design and Technology

CODE Paper Title and Authors

The Use Of Pretorsional Suspension On Monorail

006 System by Mahmud Iskandar Seth A Rahim and
Asyraf Ismail ( p.443)

Finite Element Analysis Of Asphalt Pavement Rutting

Using Viscoplastic Model by Amiruddin Ismail,
Mohammed Hadi Nahi, A. K. Ariffin and Ramez Al-Ezzi
Al-Mansob (p.455)

Performance Evaluation Of Asphalt Mixes Modified With

050 Coconut Shells by Amiruddin Ismail, Ramez Al-Ezzi
Al-Mansob and Mohammed Hadi Nahi ( p.468)

Review of Fiber-reinforced Bituminous Pavement and

Possibility of Using Oil Palm Fiber as Alternative
Reinforcement Material by Zargar M., Ahmadinia E.,
Karim M.R., Mahrez A. (p.478)

Effect Of The Elongated Aggregate On The Marshall

059 Properties Of The Pavements by Ahmadinia E., Zargar
M. , Karim M.R. and Mahrez A. (p.495)

Typical Types Of Hot Mix Asphalt (H ma) Pavements

071 Deterioration (Distress) And Maintenance Methods by
Pourtahmasb, M.S. and Karim, M.R. (p.503)
1 - 112
– 12 th April, 2011 (Tuesday)
th April, 2011 (Tuesday)
Time: 10.30 am – 12.30 pm PARALLEL SESSION 7
HBP Resource Centre
Theme: Urban Transport Development

CODE Paper Title and Authors

Factors Affecting Transportation In Africa by Massuod

008 Ali Ahmed Abuhamoud, Riza Atiq O.K Rahmat and
Amiruddin Ismail (p.515)

Growth And Challenges Of Malaysian Urban

013 Transportation by Aldukali Salem I. Almselati, Riza
Atiq O.K Rahmat and Othman Jaafar (p.522)

Comparison Of Accessibility Between Honeycomb And

044 Terrace Housing by Rahaya Bt Md Jamin and Hassim
Bin Mat ( p.530)

Analysis Of Traffic Flow Characteristics In A University

Setting by Abdul Azeez Kadar Hamsa (p.543)

Urban Growth And Its Implication On Home And

074 Workplace Location: A Conceptual Framework by Syra
Lawrance Maidin and Jamilah Mohamad (p.554)

Regression Analysis Of Land Use And Private Vehicle

077 Usage In Urban Area by Mehdi Moeinaddini and
Muhammad Zaly Shah (p.566)



Ahmad Hilmy Abdul Hamid

Universiti Sains Malaysia, Pulau Pinang, Malaysia
hilcom@usm ,my

ABSTRACT: Congestion problems at toll plazas are becoming a pressing issue in Malaysia.
VISSIM, a micro-simulation software, is used to investigate various traffic operations and
management configurations at several toll plaza locations in Malaysia. The study made
simulation models of the selected toll plaza operations. It was found that traffic volumes, toll
booths orientation, storage capacity and types of toll service have influence on traffic
operations and efficiency of the toll plaza. A preliminary table of recommended
configurations for different traffic conditions and toll plaza setup is proposed. A more detailed
study is suggested to enhance the proposed recommendations.

Keywords: Toll Plaza, Simulation, Traffic Operations, VISSIM, Payment System, Traffic

Within the decade of 1990 to 2000, toll roads became the preferred financing
model for Malaysian policy makers following the success of the North-South
Expressway project (PLUS) which used the closed-toll road system. Road users pay
for the distance they travel in the closed-toll system at their exit points. Whilst the
PLUS highway served to connect the North and South states of Peninsular Malaysia,
a significant number of toll roads were introduced within the city of Kuala Lumpur, the
economic capital of the country. These toll roads were mostly of the open-toll system
where road users pay a fixed amount when they enter and exit the road system.
Initially, the toll roads received lukewarm acceptance as the city dwellers frowned on
the need to pay for the privilege of using the toll roads. Some journeys required that
users pay several Ringgits per direction using routes that may have more than one
toll plazas. As the city grew and new developments started on the fringes of the city,
the toll roads began to be the preferred routes for commuting traffic. Over the years,
more and more users opted for toll roads in search of faster and smoother journeys
to and from their homes.

The PLUS highway too became a life line to many towns and upcoming cities in
the country. Previously unthinkable inter-city commuting journeys became quite
attractive as the highway provided fast and relatively economically transport option.
The absence of a strong medium distance public transport system virtually forced
commuters to use the toll highway.

Toll roads were put in place primarily to offer the following:
i)A quick approach to providing needed infrastructure by utilizing private
investment funding in lieu of the limited resources available to the federal or local

ii)Alternative routes for users who are willing to pay against the free but older and
congested routes

As the years went by, more and more users ply the toll roads accepting that
comfort was paramount as oppose to the daily back-breaking slow moving and hectic
traffic. Recently, what was meant to ease congestion is itself a cause of congestion.
The sheer volume of traffic during the morning and evening peak hours has stretched
the present toll system in Malaysia teetering on the brink of ‘failure’ in terms of level
of service and comfort. When previously, congestions may occur on the road link, it is
common these days in Malaysia to have heavy congestion at the entrance or exit of
toll road systems

Various measures such as the use of ‘Touch-N-Go’ cards and Smart Cards for toll
payment system and multiple toll booths for better traffic management have been
introduced to avoid or at least reduce congestion at toll plazas. Presently, however,
those measures are only useful during the off-peak hour periods. It was thought that
better understanding of the cause and effect besides the obvious over-saturated
demand on the toll plaza setup needed to be made.

Polus and Reshetnik (1997) proposed a chart for determining the number of
automatic lanes against desired throughput. Tables of toll booths gate assignments
for various flow scenarios was also proposed by Pratelli and Schoen (2003). An
optimal lane configuration model using Queuing Theory and Mathematical
Programming was presented based on historical data by Kim (2007). Those studies
relied on site specific inputs for their outcomes but the methodologies provide an
avenue for us to develop and compare results. Some simulation approaches to toll
studies were also made Xiuli (1999) and Ito (2004) but not for forecasting toll system


The design of toll plazas in Malaysia is subject to guidelines provided by the
Malaysian Highway Authority (MHA). These guide notes are adapted from other
international guidelines for toll facilities. Inherently, local needs in terms of level of
service requirements or physical layout configurations are subject to the operators
own interpretation. The most likely approach would be the use of demand estimates
versus service capacity which in turn is based on available recorded experiences.
The more recent toll plazas are built in retrospect of the older toll plazas which are
found on the PLUS toll road network.

The one advantage of the toll traffic system despite the ‘subjective’ design
reference is that vehicles will keep on moving along until the system becomes
saturated and even that saturated situation will subside after a certain period until the
next cycle of saturation recurs. This apparently ‘working’ condition however may
camouflage any lacking or improper design that exists in the system. Often, the retro-
fitting approach is applied once the conditions are intolerable to the users. In fact, the
threshold for determining ‘intolerable’ conditions is itself quite dubious and unclear.


An on-going study is presently underway to address several of the issues raised
above. The most appropriate approach for studies of this kind is simulation modeling.
Obviously, an actual on-ground or ‘live’ experiment is almost impossible not because
that they might not work but mainly due to the skepticism and reluctance of the toll
road operators to test new ‘unproven’ proposals from third parties. A simulation study
does away with the dangers of blunders which are the main fear of the toll operators
whilst providing a safe haven for studying various toll plaza configurations and traffic
management systems.

VISSIM is a software package capable of multi-modal micro-simulation of traffic

movements and interactions. The modes of traffic relevant to toll plaza simulation
include motorcycles, cars, vans, lorries and buses. Further breakdowns of vehicle
types are possible via a vehicle definition module within VISSIM. The software also
provides 3D model views which gives clearer scenario appreciation besides collating
the necessary traffic data deemed relevant for the study.

Table 1 below shows the input and output parameters that are required and
obtained for a typical simulation run for VISSIM.

Table 1. Toll Plaza Simulation: Input and Output Parameters

Input Output
1.Vehicle Volumes 1.Average delay
2.No of Toll Booths 2. Total delay
3.Size of waiting area 3.Average no. of vehicles
4.Types of payment system processed per booth
5.Traffic access arrangement 4.Total no. of vehicles processed

Figure 1 shows a typical toll plaza layout that can be configured in VISSIM.
Besides actual site layouts, proposed or future layouts may also be configured for
testing as long as the scale or dimensions of the study area with respect to vehicular
sizes are maintained. This flexibility allows for changes in the toll plaza layout to be

The primary elements within VISSIM that needs to be calibrated to site or local
conditions are speed distribution, vehicle type distribution and if necessary vehicle
dimensions. Once these parameters are established the simulation run can proceed
as the software caters for vehicular interaction automatically. Using probabilistic
conditions, vehicles will select available toll booths randomly subject to space
constraints in the queuing area. Any preset traffic arrangements such as heavy
goods vehicles (HGV) keeping to the left (right-hand drive) of the toll plaza maybe
implemented using a routing module provided in VISSIM. Safety conditions in the
queuing area are maintained via traffic interaction rules which are placed at potential
merging and diverging areas before and after the toll plaza.
Figure 1. Toll Plaza Layout Configured in VISSIM

Figure 2 shows the 3D model screen capture of the toll plaza operation running
using VISSM.

Figure 1. 3D Model of Toll Operation using VISSIM

It was hypothesized that toll plaza configurations are primarily related to the
volume of vehicles and type of payment system implemented. Toll plaza
configurations include size of queuing area, toll booth arrangements and traffic
access instructions. Simulation runs were made for several combinations of vehicle
volumes and toll plaza configurations to obtain some preliminary data for analysis.
Case studies of several existing toll plazas on the PLUS highway were made to
establish current operating conditions for reference. The configurations to those
existing toll plazas were then modified within the simulation model and their
operational differences ‘observed’. Table 2 shows the case studies data from site and
simulation runs.

Table 2. Case Study Data: On-site and Simulations Runs

Case Study No of Toll Parameters On Site Data Simulation Data

Location Booths Observed (Modified toll
Sungai Dua Toll 7 parallel Delays 8 mins 5 mins
Plaza 5 series Nos.Vehicles 1500 / hr 1800 / hr
Juru Toll Plaza 4 parallel Delays 16 mins 12 mins
8 series Nos.Vehicles 2600 / hr 2800 / hr
Sungai Petani 4 parallel Delays 17 mins 13 mins
Toll Plaza 2 series Nos.Vehicles 900 / hr 1100 / hr


The provision of toll booths and its payment collection system at toll plazas in
Malaysia has so far been a trial and error exercise. Often, after several years of
operation, toll operators are faced with high occurrence of congestion at their entry
and exit points.

A proposal is made here to recommend a suitable configuration that maybe

adopted for a particular toll plaza site (new or existing) depending on the volume and
composition of traffic expected or prevailing at that site. Table 3 shows the
preliminary proposed design table for toll plazas.
Table 3. Preliminary Proposed Design Table For Toll Plazas

Number Traffic No of Payment Collection

of Composition Booths system
vehicles (% HGV) required Type Nos.

< 5% Manual 1
5-10% TNG 1
< 1000 10-20% 3 Smart Card 1
20-30% Hi-Speed 0
Auto Detect
< 5% Manual 2
5-10% TNG 2
10-20% 4-6 Smart Card 2
3000 20-30% Hi-Speed 0
Auto Detect
< 5% Manual 3
5-10% TNG 6
3000- 10-20% 6-12 Smart Card 2
5000 20-30% Hi-Speed 2
Auto Detect
< 5% Manual 4
5-10% TNG 8
10-20% 12-20 Smart Card 4
10000 20-30% Hi-Speed 4
Auto Detect

The advent of toll roads provided an alternative route choice for smooth, high
speed and safe travel for daily commuter within a city or between cities. As the
number of road users opting for toll roads increase the toll gates become hot-spots
for congestion sometimes to intolerable conditions. Efforts should be made to
facilitate the provisions of toll plazas with anticipated design requirements that will
allow toll gates to be responsive to the traffic demands. While admittedly, over
saturated demand can never be satisfied, the amount of traffic below the saturation
volume can still be catered for if careful considerations are made on time at the right
location. Further studies are necessary but intuitively, not all toll plazas congestion is
caused by over saturated demand. The precise remedy however requires site
specific scrutiny and is locality dependent. The proposed preliminary toll plaza design
table is hoped to be a starting point of reference for such a remedy for Malaysia and
other countries. Simulation modeling in particular the VISSIM package can help
operators and policy makers determine their best option for implementation.
This research was funded by Universiti Sains Malaysia via its Research Creativity
and Management Office (RCMO) short-term grant.


Seongmoon Kim (2007) Modeling Decisions for the Time-Dependent Optimal

Lane Configuration Policy with Queueing Theory and Mathematical
Programming, V. Torra, Y. Narukawa, and Y. Yoshida (Eds.): MDAI 2007,
LNAI 4617, pp. 489–499, 2007, Springer-Verlag Berlin Heidelberg

Ito, T. (2004) “Simulation-based analysis of traffic jams at toll plaza with ETC
gates”, Proceedings of the Japan/USA Symposium on Fexible Automation ,
Denver, CO, July 19-21

Ito, T. and T. Hiramoto. (2004) “Process simulation model towards analysis of

traffic jams around toll gates”, Information Technology Letters, Forum on
Information Technology 2004 , LO-002. (in Japanese)

A Pratelli, Università di Pisa; F Schoen, Università di Firenze (2003) IT, Optimal

design of motorway toll stations, European Transport Conference

Krajzewicz, D., G. Hertkorn and P. Wagner. (2002) “An example of microscopic

car models validation using the open source traffic simulation SUMO”,
Proceedings of the 14 th European Simulation Symposium , October 23-26,
Dresden, Germany, pp.318-322.

Horiguchi, R. and M. Kuwahara. (2000) “A theoretical analysis for the capacity

of toll plaza partially with ETC tollgates”, J. of Japan Society of Civil
Engineers , No.653/IV-48, pp.29-38. (in Japanese)

van Dijk, Nico M., Mark D. Hermans, Maurice J. G.Teunisse, and Henk
Schuurman. (1999) Designing theWesterschelde tunnel toll plaza using a
combination of queueing and simulation. In Proceedings of the 1999 Winter
Simulation Conference , pp.1272-79.

Pursula, Matti. (1999) Simulation of traffic system: An overview. Journal of

Geographic Information and Decision Analysis 3 (1): 1-8.

Chao, Xiuli. (1999) Design and evaluation of toll plaza systems. Technical
report, Department of Industrial and Manufacturing Engineering, New Jersey
Institute of Technology.

Polus A.(1998) Dynamic equilibrium and concepts of toll-plaza planning. Traffic

Engineering and Control, 230-233.

Belenky, A. S. (1998) Operations research in transportation systems: Ideas and

schemes of optimization methods. Norwell, MA: Kluwer Academic.
Abishai Polus and Israel Reshetnik (1997) A new concept and a manual for toll
plaza planning, Can. J. Civ. Eng. 24: 532.538

Al-Deek H.M., Mohamed A.A. and Radwan A.E. (1997) Operational benefits of
electronic toll collection: case study. ASCE J. Transpn Engng, vol. 123, 467-

Burris, M.W., and E. D. Hildebrand. (1996) Using microsimulation to quantify

the impact of electronic toll collection. ITE Journal 66 (7): 21-25

Polus A. (1996) Methodology and simulation of toll-plaza analysis. Road and

Transport Research, vol. 5.

Matstoms, Pontus. (1995) Queue analysis for the toll station of the Öresund
fixed link. Technical report, VTI, SE-581 95, Linköping, Sweden.

Gulewicz V. and Danko J. (1994) Simulation-based approach to evaluating

optimal lane staffing requirements for toll plazas. Transp. Res. Record 1484,
T.R.B., 33-39.

Betrò B., Schoen F. and Speranza M.G. (1991) Discrete-Time Point Processes:
Applications to Road Traffic. in Papageorgiou M. (ed.) “Concise
Encyclopedia of Traffic and Transportation Systems”, Pergamon Press, New
York, pp. 101-106.

Betrò B., Schoen F. and Speranza M.G. (1987) Dynamic estimation of queue
behaviour in urban traffic, EJOR 31, pp. 368-375.

Edie L.C. (1954) Traffic delays at toll booths. Journal of the Operations
Research Society of America, vol. 2, 107-138.

Vivek Ogra 1
Aurobindo Ogra 2

Director – Technology & Innovation, Vbsoft India Ltd., Ahmedabad, India
Lecturer – Department of Town and Regional Planning
Faculty of Engineering and the Built Environment
University of Johannesburg, South Africa
Email:; 2 Email:

Abstract: Enhancing the efficiency of public transit operations is a continual challenge for
public transport authorities and operators. Key tasks associated with the system involve
number of different people performing individual tasks like: Ticket sales, the cashier, the
ticket inspector on the vehicles, control center operations team etc. The transit operations
have over a period of time moved from standard tickets through manual system or other
informal basis to completely automatic systems involving electronic fare collection systems,
automated vehicle location system, vehicle scheduling, passenger information etc. The
upward transition of infrastructure in terms of quality and availability demanded introduction
of more automated methods and tools to be put in place for transit managers and operators
ensuring high productivity environment which is driven by predictable management system.

The paper discusses a case study of Integrated Transit Management System (ITMS)
adopted for Bus Rapid Transit System (BRTS) for City of Ahmedabad, India. ITMS is
expected to meet the corporate objectives of enhancing service standards, bring in
commuter market approaches, better organization of planning and operations; integration of
Para-transit, integration of control, capital improvements, marketing, and automate collection
and payment of transit fares. Based on the initial results, the paper discuss the components
of the ITMS with its aim to provide world class transit experience embarked on the project to
implement which would provide stakeholders and operations with a system to bring in world
class operational efficiency and automation for its transit operations.

The paper discuss how ITMS enables transport organization to automate its financial
characteristics, operational characteristics, better insight into passenger profiles, perform
route analysis to optimize on operational efficiency, service consumption, perform functional
area productivity analysis and thereby creating a sustainable public transport as a preferred
user choice by citizens.

Keywords: Public Transport, Operational Efficiency, Integrated Transit Management System

(ITMS), Sustainable Transport


Enhancing the efficiency of public transit operations is a continual challenge for

public transport authorities and operators. Key tasks associated with the system
involve number of different people performing individual tasks like: Ticket sales, the
cashier, the ticket inspector on the vehicles, control center operations team etc. The
transit operations have over a period of time moved from standard tickets through
manual system or other informal basis to completely automatic systems involving
electronic fare collection systems, automated vehicle location system, vehicle
scheduling, passenger information etc. The upward transition of infrastructure in
terms of quality and availability demanded introduction of more automated methods
and tools to be put in place for transit managers and operators ensuring high
productivity environment which is driven by predictable management system. (Ogra

The public transit requires increased coordination and mechanisms like –

organizational coordination, contractual coordination, partnership coordination, and
discursive coordination (Sorensen and Longva, 2010). The Intelligent Transportation
System (ITS) is the most important approach accepted worldwide to solve traffic and
transportation issues at intra-city / inter-city level. The Advanced Public Transport
System (APTS), forms one of the sub-sector of transportation system which has
remained in focus during the recent years. Currently, the most significant in public
transportation system in cities is the inadequate application of appropriate
technology environment in which it’s working in terms of technologies of modern
communication system, information, electronics controlling, Global Positioning
System (GPS), Geographic Information System (GIS) etc. The APTS should
essentially be perfected in terms of : Intelligent public transport dispatch model
restricted by urban transit capability, dynamic evaluation model of public transit
service level according to the road status, automatic public transit dispatch model
according to the demand status of the public transit, intense combination of the
actualization of the intelligent public transit system’s construction, operation, and
urban planning, and integration of MIS with intelligent public transit dispatch system
(Guohua et al 2007).

Indian cities rely predominantly on public transport and public transit service is
important in order to meet the rapidly growing mass mobility. Most of the cities
represent the public bus transit service as inadequate, operational issues, declining
ridership, declining productivity, persistent losses despite rapidly increasing fares
(Badami and Haider 2007).

The paper highlights the role of technology perspective to manage the public transit
activities in a highly coordinated manner leading to a high productivity environment
and reliable services to the users. The analysis of the official documents on BRTS
has been made to identify the dimensions of public transit and technology
applications. The paper focuses on the Integrated Transit Management System
(ITMS) of the technology innovations / applications landscape of Bus Rapid Transit
System (BRTS) of Ahmedabad city in Gujarat, the seventh largest metropolis in


3.1 City Profile & Public Transport

The city of Ahmedabad is the seventh largest metropolis in India founded in 1411
AD as a walled city on the eastern bank of river Sabarmati. As per 2001 census, the
population of the urban agglomeration was recorded as 4.5 million, as compared to
3.31 million in 1995. The city once recognized as the ‘Manchester of India’ on
account of its textile industries, continues to be one of the most important centers of
trade and commerce in western India. The city can be well understood by its
institutional structure and represents the jurisdictions of Ahmedabad Municipal
Corporation (AMC), Ahmedabad Urban Agglomeration (AUA), and Ahmedabad
Urban Development Authority (AUDA) areas.

The AMC covers a jurisdiction area of around 190 sq. km’s, AUA covers 350 sq.
km’s and AUDA covers 1330 sq. km’s. The city continues to be recognized as
having large concentration of economic activities. Of the 4859 factories in the city,
the largest share is chemical and petrochemical industries which accounts to 29% of
the industries, followed by textile industries accounting to 12%. Around 40% of the
city area is under residential, followed by 15% area under industrial.

Of the total population of AUA, around 78% resides in the municipal area. As
observed from the land use and spatial structure, the city continues to be relatively
compact, and due to the rapid urbanization, this has led to the spillover of population
and spread of new settlements outside the city limits (CDP, Ahmedabad 2006).
Figure 1. Spatial Distribution of Residential and Commercial Areas, Ahmedabad
Source: BRTS, Ahmedabad (2006)

Public Transport facilities in Ahmedabad are facilitated by Ahmedabad Municipal

Transport Service (AMTS) under the municipal body. The AMTS operates 550
buses per day, catering to more than 650,000 passengers. The buses cover around
150 routes and makes around 250,000 trips per day. The services area of AMTS is
more than 375 sq. km’s and covers city and peripheral areas. According to AMTS,
as per 2005, the fleet size of buses, number of routes, service kilometers per day,
buses per 100,000 population, and number of passengers per day has decreased
and number of services were cancelled on account of: poor connectivity in
peripheral areas, inadequate parking facilities, lack of pedestrian facilities, inefficient
public transport system etc.

Table 1. AMTS Service Indicators (2000-2005)

Source: City Development Plan (CDP), Ahmedabad, 2006
Year Fleet- No. of Service km’s Buses per No. of
size(Buses) Routes per day 100,000 of passengers
population per day
2000 942 144 155675 22 757852
2001 886 140 151245 21 678861
2002 801 136 124375 18 574257
2003 687 115 81802 15 385682
2004 601 110 76028 13 325378
2005 540 117 77411 11 349653
3.2 Need for Operational Efficiency
The declining occupancy ratio from 71% to 54% from 1992 – 2005, were clear
indicators of declining public transport efficiency. Looking at absence
existing the declining issues
related to fleet size, number of routes, serviceability and ridership etc, a need was
felt for providing a sustainable solution which could address all system,
such issues which
were facilities,
etc inefficient
responsible for

transport. resulted
The AMTS was
to cater to
the financial
related to
maintaining the
fleet. The
of effective

monitoring infrastructure fleet utilization

in the
deterioration of operational efficiency and efficient management of services.

Figure 2: AMTS Routes, Ahmedabad

Source: BRTS, Ahmedabad (2006)
3.3 Models Identified and Options Adopted

In order to address some of the issues associated with public transport, and to
promote sustainable intra and intercity, regional transport facilities, a number of
initiatives were evaluated to provide quality transit services and solutions. The key
stakeholders associated from the Government of Gujarat considered number of
options for providing efficient public transport facilities like introducing: metro rail
system, regional rail system, bus rapid transit system, and regular bus system (ITF
2010; BRTS 2006). For intra city needs, BRTS was considered as the option to be
progressed with for implementation. The BRTS was developed as a strategic
intervention for addressing the issues related to public transport transit besides
environmental, operational, m a n a g e m e n t a n d o t h e r operational issues.

3.4 BRTS – Introduction

The BRTS started its first operations in 2009, and was awarded with ‘Best Mass
Transit Award-2009’ by

Ministry of Urban
D e v e l o p m e n t ( M o U D ) , Government of India and “Most Innovative Initiatives
Award” in Dec 2010. The system has been in operation since 2009, and carried more
than 90000 passenger daily with its fleet of 55 buses covering 40 kms of BRTS

network within its first year. Figure 3: BRTS Corridors

This BRTS is the first full Source: BRTS, Ahmedabad (2006)
BRTS systems in India operating as a closed system. Within first four months of its
inauguration, the ridership doubled, and gained significant public response for its
efficient, reliable, and quality service. The public rating for the system has been
recorded on an average of 9 on a scale of 10.

Key Statistics – BRTS Operations

Table 2: Key Operational Indicators of BRTS, Ahmedabad
Source: BRTS, Ahmedabad (2006, 2010)
Key Indicators 1st month 8th month After 12
Fleet Size (Buses) 18 39 50
Route- Total Km’s of length 12 km’s 25 km’s 40 km’s
Average Total Km’s / day 3640 km’s 8665 km’s 11500 km’s
Operational Timings 07:30 hrs – 06:00 hrs – 06:00 hrs –
22: 30 hrs 23:30 hrs 23:30 hrs
Frequency of Buses (peak 5 min 3-4 min 2.5-4 min
Frequency of Buses (off-peak 10 min 6-12 min 6-8 min
Total passengers 536749 1587426 2790000
(196% increase) (420% increase)
Average ridership/day 17315 51207 90000
(196% increase) (420% increase)
Average pax/bus/day 962 1313 1800
(36% increase) (87% increase)
Total Revenue in 1 month 2511888 8684495 17980000
(IN R) (24 6% increase) (616% increase)
Average Revenue per day 81029 280145 580000
(IN R) (24 6% increase) (616% increase)

3.5 Technology Drivers behind Successful Operations of BRTS

Besides a well thought planning and scaling of BRTS, the earlier focus remained on
the designing and construction of the BRTS. The system included in its early and
later stages much focus on delivery of superior quality transit through innovative
technology / applications using indigenously developed methods and innovations.
The technology applications adopted for the BRTS are in three broad areas:

RFID based bus docking mechanism

Integrated Transit Management System (ITMS)
Area Traffic Control System

3.6 ITMS Components

In order to enable BRTS automate its financial characteristics, operational

characteristics, better insight into passenger profiles, route analysis to optimize an
operational efficiency, service consumption, functional area productivity analysis, the
following components were considered in the ambit of ITMS solution landscape :
• Automated Fare Collection System (AFCS)
Passenger Information System (PIS)
Automated Vehicle Location System (AVLS)
Infrastructure Landscape
Other ITS Information Sub-systems

Figure 4: ITMS Integrated View, BRTS, Ahmedabad

Source: Ogra, V. (2010); AMC (2010)

3.6.1 Automated Fare Collection System (AFCS)

AFCS is an automated revenue collection system facilitating purchase and use of
pre-tickets through an electronic systems to permit access to/ from the transit
stations and buses. The 62 bus stations facilitate easy access to transit services.
The present system provides information on passenger flows by each station, ticket
statistics and other details in real time by the central server. The automated system
has enabled in avoiding revenue leakages and has facilitated easy way of revenue
collection. The passenger origin-destination data provides route structure for BRT
operations and also provides rationalized routing plan for AMTS. Within one year of
successful operations, the average revenue per day has increased to 616% since its
initial operational month.
3.6.2 Passenger Information System (PIS)
The buses and the bus stations are equipped with LED panels, audio systems
providing route information, arrivals, departures, next vehicle, next station, display
announcements etc. The user satisfaction survey carried out by operating agency
on the passengers at the end of its 11 months commercial operation showed an
average rating of 9.0 out of 10 from its users. The survey is focused on various
dimensions like safety, operator driving, frequency of service, ease of fare payment
and cleanliness at stations.

3.6.3 Automated Vehicle Location System (AVLS)

The buses are equipped with GPS Console and GPRS technology is used to track
the vehicle operations. The status of bus schedules is controlled by control center
which monitors all the buses on real time, and this facilitates operations of PIS,
vehicle kilometer count, speed monitoring and instructions to slow down / go fast
depending on the bus schedule. At the end of 11 months of commercial operation,
over 95% of departures are on time (+/- 90 sec time), 65% of arrivals were on time,
22% arrived before time and 13% delayed.

3.6.4 Infrastructure Landscape

The infrastructure landscape of ITS includes data center, disaster recovery site,
control center, training center, bus station infrastructure, bus infrastructure,
communication infrastructure, operations, management and management services,
ticket operations, control center operations, and depot management system (under
implementation). The analysis of data through specialized reporting structures on
multiple dimensions facilitates transit managers to plan and manage transit services
in a highly efficient manner.

3.6.5 Other ITS Information Sub-systems

This includes operations, management and maintenance services like: ticket
operations, control center operators at various levels like control center, technical
operators, and bug fixing and roll out; system administration and database
administration, ticket terminal operations, overall operations of AFC, vehicle
monitoring and communication services, incident management system, vehicle
scheduling and dispatch service, passenger information system, bus terminal
management system, depot management system.
It is evident from the case study that a considerable number of urban transit
commuting can undergo a transformational stage in terms of ridership, patronage,
revenue enhancement with high quality transit service. The appropriate
technological landscape through ITMS can improve reliability, frequency, customer
convenience, safety, user satisfaction and improved travel speed. The adoption of
such system can improve the rapid mobility across cities without compromising in
terms of flexibility, cost and convenience.

Badami, M.G, Haider, M. (2007). An Analysis of Public Bus Transit Performance in
Indian Cities. Transportation Research. Part A 41, 961-981
Bus Rapid Transit System, Ahmedabad (2006). Detailed Project Report, Phase-1 :
GIDB, AMC, AUDA and CEPT University.
Bus Rapid Transit System, Ahmedabad (2010). Monthly BRTS Reports, Accessed
on 10 Jan 2011:
City Development Plan, Ahmedabad (2006). Jawaharlal Nehru National Urban
Renewal Mission. Ahmedabad : AMC, AUDA, CEPT University.
Guohua, Z., Ming, L. and Jingxia, W. (2007). Application of the Advanced Public
Transport System in Cities of China and the Prospect of its Future Development.
Journal of Transportation Systems Engineering and Information Technology, 7(5)
International Transport Forum (2010). Transport and Innovation – Unleashing the
Potential, 26-28 May, Leipzig.
Ogra, V. (2010). Integrated Transit Management System (ITMS), Ahmedabad
Janmarg Limited (unpublished case paper).
Sorensen, C.H, and Longva, F. (2010). Increased Coordination in Public Transport –
Which Mechanisms are Available ?. Transport Policy, 18 (2011) 117-125.


Hamed Eftekhar 1 And Professor Ir. Dr. Riza Atiq Bin O.K. Rahmat 2

1, 2 University Kebangsaan Malaysia (UKM), Bangi, Malaysia

ABSTRACT : These days large cities are facing highly increase in populations and car
dependency. In this condition an efficient urban traffic management system playing very
important role to handle issued problems. Urban traffic management system divided to two
main parts. First one which is considered more yet is; current traffic and second one; quite
traffic (parked vehicles) and its effects on current traffic have been considered less .Parking
management is known as one of the most important tools in urban management. There are
different items regarding to parking management that one of the most important one of them
is parking pricing which has tremendous effects in an efficient urban traffic management
system. In this paper we will discuss on why parking pricing and with an study case (UKM)
will see how we can implement and predict about parking pricing and its affects parking
demand and traffic volume and also mention how parking pricing can be considered as the
second most effective congestion reduction strategy (after peak-period congestion pricing
and before Vehicle Miles Traveled (VMT) taxes, fuel taxes and pollution emission fees).

Keywords: Parking, Parking Pricing, Parking Management, Parking Pricing Strategies,

Parking Revenue.

As usual usage of automobile during 24jam is almost 1-2 hours and it’s parked for
remain almost 22 hours which requires lots of parking. There are 3 to 6 parking
spaces per vehicle (one at workplace, one at home, and various places such as
street side and shopping’s, school, and parks area) in most communities. Providing
this parking facilities even in an ordinary urban area are costly since there are land,
operation and construction costs that come to almost $500 to $for each parking
space. Sometimes even value of parking spaces is more than vehicles that occupied
them, yet mostly parking facilities are unpriced since there are indirect or hidden
costs such as taxes, lower usage and so one.

There are direct or indirect payments for parking facilities so really free parking
not exists. Direct pricing of marking facilities are more recommended by many
experts and for variety reasons many cities, campuses and commercial buildings
are expanding when and where parking’s priced. Several current trends increase the
justification for parking pricing, including increased urbanization and land costs,
increased concern about vehicle traffic external costs ( congestion, sprawl, accident,
pollution), and improved pricing technologies. However, unpriced parking is well
established, so parking pricing implementation requires overcoming various political,
institutional and technical obstacles. Care is required to address potential problems
and communicate the benefits.

Parking pricing is just one of the several parking management strategies, as

Summarized in table 1. If parking pricing implemented as a part of an integrated
parking management program that includes support strategies such as improved
user information, increased parking options, and better enforcement it tends to be
most effective and beneficial. Based on our goals from parking pricing and their
priority we make decision which and what kind of parking rates are more proper for
any specific case. Main goals of parking pricing are:

Car dependency reduction; tends the motorists to switch transportation

especially in the existence of alternative transportation (public) modes.

Demand management; set charges to attain above 80% parking-lots

occupancy. Encourage users to chose uncongested times rather congested
ttimes by offering variable rates.

Revenue generation; set the rates to gain maximum revenue. Expand when
and where parking is priced.

Motorist convenience; charges only when it’s needed, minimizing cost and
provide discounts and exemptions, such as low monthly passes.

Parking pricing is a suitable choice almost anywhere parking is congested. Experts

recommend setting prices to maintain 85-90% occupancy rates; this is called
performance-based or responsive pricing (Shoup 2005).

Parking pricing affects transportation system from several aspects such as:
vehicle ownership reduction (especially residential parking pricing); destination
(switch’s to the cheaper parking areas); mode shifts from private vehicle to public
transport, ridesharing, cycling, walking; changes in trip making schedule from priced
to free or underpriced periods; change in parking location (to the cheaper or free
parking lots); and reduce stop duration. Also Demographic, geographic and
economic factors tends to these changes; Greater impacts are likely to result of
drivers lower income, more optional trips, and if the commuters have a better mode,
destination and parking options.

Parking pricing also reduce traffic congestion by decreasing traffic caused by

motorists who are looking for an unoccupied parking lots and shifting to the
alternative modes, particularly if implemented widely throughout an urban region
and in conjunction with other demand management strategies (Booz Allen Hamilton
2006). This leads to increase economic productivity (Roth 2004 &1 965).
Surveys shown 8-74% of traffic congestion in commercial centers is made by
vehicles cruising for an on-street parking space since its usually unpriced (Shoup
2004). Modeling by Deakin (1996) estimated that in Southern California (all values in
1991 dollars):

A 1 0¢ per vehicle-mile congestion fee reduces VMT (Vehicle Miles Traveled)

2.3% and congestion delay 22.5% (Congestion delay/VMT ratio = 9.8).
A $3.00 per day parking fee would reduce VMT 2.7% and congestion delay
7.5% (Congestion delay/VMT ratio ratio=2.8).

A 2¢ per vehicle-mile VMT fee reduces VMT 4.4% and congestion delay
9.0% (Congestion delay/VMT ratio = 2).
A $0.50 fuel tax increase reduces VMT 4.1% and congestion delay 6.5%
(Congestion delay/VMT ratio ratio=1 .6).
A 1 .0¢ per vehicle-mile emission fee reduces VMT 2.2% and congestion
delay 3.0% (Congestion delay/VMT ratio=1 .4).

Above ratios shown that parking pricing can consider as the second most
effective congestion reduction strategy, less effective than peak period congestion
pricing and more effective than flat VMT fees, fuel taxes and pollution emission fees.

1.1 Efficient Parking Pricing

Parking pricing is significantly appropriate:
Where parking facilities are costly, land is valuable or parking facilities are

Commercial centers with more than 5000 employees, since beyond this size
surface lots can’t satisfy total parking demand, so costly structured parking
facilities required.
Where they want to encourage commuters to the alternative modes of transit
to reduce traffic congestion, fuel consumption and pollution emissions.
In the areas where environmental protection and livability effort to decrease
impervious surface area and (portion of paced land) and total vehicle travel.
Where development affordability is an important object.
When owners or governments assets need additional revenues.
Generally efficient and fair parking prices are set to equal marginal costs, except
if a subsidy is particularly justified, for example to aim fairly or strategic
development. Marginal pricing also prevents municipality from spending $2 value of
resources to provide parking spaces for which users only worth at $1. Direct parking
payment enable that users to save money if decrease their parking expenditures.
For example if parking packaged with housing so tenants should pay for parking’s
facilities regardless of whether or not they need them, but if residential parking’s are
priced separately so tenants can save money if they decrease their vehicle
ownership. Correspondingly if employee directly pay parking fee so they can save
money by using alternative transit modes, an option not available if unpriced parking
is an automatic employee benefit. Under current parking pricing, saving through
parking costs reduction are dispersed through the economy while in an efficient
pricing returns more savings to individual users who discount their parking

Table1, Current parking pricing VS efficient parking pricing (Litman 2010)

Current parking pricing Efficient pricing
Motorists reduces parking costs Motorists reduces parking costs
(Reduce vehicle ownership and vehicle, (Reduce vehicle ownership and vehicle
uses less costly parking spaces) uses less costly parking spaces)

Reduced parking costs Reduced parking costs

( Reduced parking congestion, reduces ( Reduced parking congestion, reduces need
need to build and maintain parking to build and maintain parking facilities)

Money saving
Money saving (Returned to the individual motorist)
(Widely dispersed through economy)

As it’s shown in above table in an efficient parking pricing motorists affects directly
by saving more money if they reduce or optimize their vehicle ownership and
parking demand.

1.2 Use of Revenues

Revenues from net parking can use in various ways:
• Returning parking pricing expenditures (equipment, enforcement, user
information, etc.).
Returning parking facility construction, operation and maintenance system

Returning the equivalent of rent and taxes on parking facilities. For example,
an urban parking program can generate net revenues equivalent to what
would be earned if the facilities were privately owned and paid rent on the
land and taxes on facilities and profits.

Parking and transportation management system expenses, including trip

reduction plans and improvements to provide alternative modes that reduce
parking and traffic problems.
Urban transportation expenses (street and sidewalk assets and operating

Neighborhood and region improvements, such as street-scaping, improved

street and sidewalk cleaning and security, and commercial regional
Reduce general taxes or offset tax increases that would otherwise be
Financial support and founding special projects or programs, such as
municipal arena or center recreation.

Where parking policy is maximizing motorists convenience it means parking fees

should be less and also increasing parking supply and facilities with spending
revenues to finance additional parking supply so revenues generally are small and
generation less than 1% of total municipal or campus revenues. However, where
parking in managed to maximize revenues, parking can generate 5% to10% of total
municipal or campus revenues. The following actions lead to increase net parking

Dependant price more parking. Increase where and when parking is priced,
for example, to include smaller commercial districts, residential streets,
evening and weekends and public holydays.
Increase parking charges to the highest feasible rates.
li Alternative parking and transport options reduction (for
example limiting
nearby availability of free parking and minimizing public transit service).
i Use more cost effective pricing systems, such as multi -spaces
Increase enforcement and fines.
“[Parking] fees are largely associated with positive effects on the local economy over
the long term, though over the short term there may be a drop in the number of
visitors to such an area. The change from negative to positive effect is not only a
matter of years but also of extra measures that increase the attractiveness of the
shopping area (e.g., new shops and/or renovation of existing shopping). In relation
to the parking process, parking fees produce some benefits such as less time spent
looking for a parking space, more efficient use of parking spaces, and promotion of
‘short stay’ parking. (Van der Waerden and Timmermans 2009)”

To determine about an efficient parking pricing which is the most important step
before implementing any parking pricing we need a model that can help us since it’s
not possible we try many different prices to achieve the efficient one. What we can
do, is implementing some charges and monitors their effects on parking demand,
revenue and traffic volume. The model which used in this paper is Logistic
regression model that predicts the probability of occurrence of event by fitting data
to a logistic curve. It is a linear model used for binomial regression; making use of
several predictor variables which may be numerical or categorical. It describes the
main features data collection in a quantitative manner. The dependent variable in
the regression equation is modeled as a function of the independent variables,
corresponding parameters ("constants"), and an error term. Logistic regression is
commonly used in medical, social sciences and for predicting customer’s propensity
to purchase a product in marketing. This model in some occasion is being referred
to as linear model Logit or logistic model.

The techniques used for the modeling and analysis of numerical data consist of
values of a dependent variable (response variable) and of one or more independent
variables (explanatory variables or predictors). The error term is treated as a
random variable and represents unexplained variation in the dependent variable.
The parameters are estimated so as to give an accurate data. The best fit is
evaluated by using the least squares method though other criteria have also been

Regression can be used for prediction (forecasting of time-series data),

inference, hypothesis testing, and modeling of causal relationships although its
estimation relies heavily on the underlying assumptions being satisfied. The
derivative of pi with respect to X is computed from the general form:

Where is an analytic function in . With this choice, the single-layer network

is identical to the logistic regression model. This function has a continuous
derivative, which allows it to be used in back-propagation.

The following functional form is used in this paper to determine the dependent

2.1 Logit Function

Discussion about logistic regression necessitates advanced discussion and
knowledge about of the logistic function:

The logit function was invented in the 19th century for the description of the
growth of populations and the course of autocatalytic chemical reactions, or chain
reactions (Cramer 2003). It is an important part of logistic regression and is the
inverse of the “sigmoid” or “logistic” function.

Figure1, Sigmoid function of the Logistic curve

2.2 Case Study
The survey was carried out in Universiti Kebangsaan Malaysia (UKM) because of
the high car ownership and availability of public transport. The train station is located
on the northeast tip of the university's grounds, 1.5 km from the university's main
zone. Analytical description of the survey carried out focuses on students
(International/ local/ postgraduate/ undergraduate) because of high dependency on
private car mode of transportation among them because of poor public
transportation of campus including inside campus and from campus to KTM station
and its inverse. Below there is a picture which is shown campus map.

Figure2, UKM campus map

Following histogram illustrates the total number of registered vehicle in UKM from
years 2003 to 2009.
Figure3, cumulative number of registered vehicle by students and staffs



5000 Studa nt Staff

2003 2004 2005 2006 2007 2008 2009

Annual registerted vehicle

My survey results from implementing parking charges on switching
transportation mode in both after modeling and before modeling (just plot the data’s
we got from our survey) are shown in below graph:
Parking charges RM

Figure 4, switching transportation mode based on various parking charges in UKM

RM10; 98%
100 100
RM7; 94%
From above figures we can predict what will happen under different parking
89.9% 99%
charges and how much we will have swathing on public transportation, also we can
predict if we want reduce our parking demand for x% how much parking charge
should be imposed to achieve our minded reduction. Through this model it’s very
60 RM5; 60% 63.14%
easy and also trustable to achieve an efficient parking pricing based on our goals
which can be different, for example maybe is some places our percentage-our
goal is whether survey
maximizing revenues or minimizing parking demand or any other strategy which is
RM3; 28%
undertaken. 26.25%
As it mentioned this graph obtained very easy from plotting
Shiftour collected under
percentage data’sModelling
from our survey and modeling it under Logistic regression to gain modified graph
which is more adopt to real condition if they perform.
0 2 4 6 8 10 12 14

As it’s shown before at 2009 there were almost 15000 registered vehicles in UKM
that almost 9000 of these vehicles belong to staffs so they will be in university every
day, so we face a severe lake of parking supplies in campus and providing new
parking’s is costly. In this condition our policy about students will be trying to impose
parking charges as well as reduce their private transportation using by almost 50%
to have revenue from parking’s and use it in other purposes such as improving
campus public transportation, considering subsidies for the students who are using
public transportation (KTM or Rapid KL bus) and also providing new parking’s, also
reduce UKM parking demand and traffic volume which is made by students who use
private vehicles.

Based on our achieved graph we have to impose RM4 to RM4.5 per day to get
45-55% reduction in parking demand so under current condition the efficient parking
charge for UKM might be RM4 to RM4.5 per day, but in previous pages also we
discussed that if UKM improve its public transportation also they can impose parking
pricing to anybody who drive to campus whether student, staff, and even visitors
since the main policy regarding to transportation is reducing car dependency and
increase public transportation usage so parking pricing can be consider as one of
the quite but most important strategies to decrease traffic volumes. While motorists
seen their travel cost will be much cheaper when they use public transport rather to
private transport of course they will switch on public transport and as well as this
expenditures be higher (in the presence of wisdom) this swathing percentage will be


CRAMER, J. S. 2003. The origins and development of the logit model. Cambridge
University Press .

DEAKIN, E. 1996. Transportation Pricing Strategies for California: An Assessment

of Congestion, Emissions, Energy and Equity Impacts. California Air
Resources Board .

HAMILTON, B. A. 2006. Parking Restraint Measures. International Approaches to

Tackling Transport Congestion .

LITMAN, T. 2010. Parking Pricing Implementation Guidelines. Victoria Transport

Policy Institute .

ROTH, G. 1965. Paying for Parking. Hobart Paper 33 London.

ROTH, G. 2004. An Investigation Into Rational Pricing For Curbside Parking: What
Will Be The Effects Of Higher Curbside Parking Prices In Manhattan.

Masters Thesis, Columbia University.

SHOUP, D. 2004. The Ideal Source of Public Revenue. Regional Science and
Urban Economics, V.34, pp 753-784.
SHOUP, D. 2005. The High Cost of Free Parking. Planners Press .
VACA, E. & KUZMYAK, J. R. 2005. Parking Pricing and Fees. Transit Cooperative
Research Program, Chapter 13, TCRP Report 95.

WAERDEN, P. V. D. & TIMMERMANS, H. 2009. Introduction of Paid Parking in

Shopping Areas: Short Term versus Medium Term Effects. 16th International
Conference on Recent Advances in Retailing and Service Science. Niagara
Fall, Canada.

Athiwat Noonma 1 and Vachara Peansupap 2

1, 2 Department of Civil Engineering, Chulalongkorn University, Thailand

ABSTRACT : Infrastructure construction project is a complex work that requires engineering

consultants to provide services toward the work completion. Mass Rapid Transit (MRT)
construction project is the one of them, it contains technical works and has more project
value in which it is required engineering consultants, experts and specialists to provide
several types of services. One problem that government agencies should be considered is
the strategic option of project consultant procurement to provide services such as project
management, construction supervision, quality control of works, and so on. Generally,
strategic options of project consultant procurement could be separated into two options,
which are single and multiple contracts. Each option of contract procurement can be
advantages and disadvantages depend on work characteristics. From the strategic option for
project consultant procurement, the client should select it at the beginning stage. This
research focuses on factors’s identification and priority for selecting strategic options of
project consultant procurement. The three case studies of MRT construction projects were
used as the case study. The interview technique is used to collect data from experts. The
data analysis used Analytical Hierarchy Process (AHP) for identifying the weight and
prioritizing the options. The expected outcome of this study will be used for making decision
on selecting strategic option of consultant procurement for MRT projects during construction
period in single and multiple contract option in the future time.

Keywords : Strategic Procurement, Engineering Consultants, Analytical Hierarchy Process,

Single and Multiple Contracts

In general, construction stages of MRT project consist of two main components.
First component involves the construction of structural infrastructure works such
as superstructure, foundations, piers, beam girders, tunnels and etc. Second
component comprises with mechanical and electrical works such as signalling,
car body, power supply, lifts and escalators and etc. [Duangkamol, 2010]. As
the MRT construction is a complex project, the client should require engineering
consultants to provide services toward the work completion. Many reasons of
using engineering consultant’s service are that the government agency
inadequate of human resource and need more opinion from third parties. The
success of the project depend on the selection of appropriate consultant
because consultant involves in supporting owner’s decision. However, the
consultant service also relies on the service cost that owner provide [FIDIC,
The expansion of MRT construction in Bangkok, Thailand is increasing in the
recent years. The MRT extension and new lines MRT in Bangkok and vicinity
master plan for 10 years (2010 – 2020) consist of 7 lines with total distance of
154 km [OTP, 2010]. As the result, project consultant procurement should be
increased follow the construction project. Therefore, the government agencies
should concern the good strategy for selecting project consultant procurements.

From the literature review, the strategic contract options for project
consultant procurement can be categorized into two types. There are single and
multiple contract options [NSWG, 2008]. Single Contract option refers to only
one consultant that services all of the works in the project such as project
management services, construction supervision, etc. This option is suitable for
the project that clearly defined scope and unlikely to change after the contract is
awarded. Multiple Contract option consists of two or more consultant teams that
service works in the project such as project management consultant,
construction supervision consultant, independent certificate consultant, etc. The
option is suitable for the project that the construction work commences before
the design can be completed, unclear definition of technology, and high project
change. Each type of contract option should have advantages and
disadvantages in each characteristic of project [NSWG, 2008]. However,
strategic contract options for project consultant procurement should be
concerned with other relevant factors for adopt the suitable contract option.
Because contract option shall be effected to project management until
terminated the contract. Also, the client and the consultant have to work
together until the project completion.


This research focuses on exploring factors for selecting strategic options of
project consultant procurement between single and multiple contracts (PCP -
SMC). To develop the systematic approach for selecting the strategic contract
options, this study adopted AHP technique. The research used three case
studies of MRT construction projects under responsibility of Mass Rapid Transit
Authority of Thailand (MRTA). These are the M.R.T. Claloem Ratchamongkhon
Line Project, MRT Purple line project and MRT Blue line extension project. The
research based on two main assumptions. First, there is no influence from
politics. Second, the consulting fee is fixed from the percentage of construction
value. The consulting fee normally set up to 1.5-5.0 percent of project
construction value [NESDB, 2008]. Therefore, they are no influence to the
procurement between single and multiple contract.


This research methodology is classified as the case study. There are four main
steps of developing the decision support system for selecting the strategic
contract options of project consultant procurement.

The first step involves with the pilot study. It aims to obtain factors related to
the strategic contract options for project consultant procurement. The interview
technique was used to obtain the list of variables from 15 experts who have the
responsibility on the decision.
The second step is questionnaire development. The questionnaire is
developed and designed from the list of variables in the first step. Twelve
variables are initially listed as the important factors for selecting the contract
options for project PCP-SMC. These are; 1) Project construction value, 2)
Overall work checking, 3) Payment process, 4) Quality of consultant services, 5)
Communication, 6) Unity of consultant team, 7) Project construction duration, 8)
Project characteristics, 9) Project's complexity, 10) Owner's experiences, 11)
Types of construction contract, 12) Construction area in each contract.

The third step is the identification of main factors for selecting strategic
contract option. Twelve variables that influence to strategic contract option were
selected by experts. From 15 experts, six from twelve variables are ranked as
important factors. These factors will be used to develop the structure hierarchy
for strategic contract option. The definitions of these six factor were listed as the
1)Project Construction Value (PV); the amount of project construction value
is subjected to civil work and M&E work.
2)Project Characteristics (PC) ; the characteristic of the project is
subjected to the structural work such as elevated track /underground

3)Quality of Consultant Services (QS); the output of consultant services

to fulfil the anticipated need of the client. It may concern effects, if it is
considered on single and multiple contract options.
4)Unity of Consultant Team (UC); the unity of consultant team involves
project management, opinion of consultant team, methodology of work
process and etc.
5)Communication (CO); the communication between client and consultants or
consultant and consultant in the project. Regarding to communicative of work or
documents, such as promptitude, convenient or algorithm.
6)Overall Work Checking (OC); Internal cross checking of work between
consultants team in the project, such as documents cross checking, quality of
work or standard of work cross checking.

The fourth step aims to identify the weight of six factors that influence the
strategic option. This step shall be identified by 15 existing experts. The
analysis of weights is adopted from the AHP technique. Six reasons for
choosing AHP in this research are; 1) Reliability of result more than other
methods, Because of using comparative analysis of twin. The comparison can
be use for qualitative and quantitative data, 2) The structure of AHP (objective,
goal, criterion, sub-criterion and alternatives) is simulated human decision
process which make easily to understand, 3) Decision making can apply for one
person or group, 4) Aviod decision making by prejudice or bias, 5) The
alternative or result can be compromising and referendum, 6) No need of
experts for controlling the process [Sutham A., 2009].
The AHP application steps can be summarized as follows [Saaty, 1980],
[H.S.C.Pereta and W.K.R.Costa, 2008]
1)State the problem.
2)Determine the qualitative factors that need to be evaluated or compared.
3)Identify the factors that influence the problem.
0)Structure the hierarchy of the criterion, sub-criterion (if need), and

1)Set survey questionnaire matrix and state the question for pairwise
comparisons clearly above each matrix. Also, definition the scale values for
pair-wise. For the scale values, Saaty [1994] has given as Table 1. The number
of decision makers that need for judgement and developing the matrix is
n(n-1)/2 , where n is the elements of n x n matrix. For the matrix, let C1, C 2,..., C n
be the set of activities. The Quantified judgement on pairs of activities Ci, C j are
represented by an n x n matrix. A = (aij), (i, j = 1, 2,...,n) . The rules of entries aij
are defined by aij = α and aji = 1/ α , where α ≠ 0 and aii = 1 for all i.

4)Collect data from experts who have experiences on the find objectives by
entering pairwise comparison judgements of factors with respect to their
impacts on the overall objectives. Then each expert enters pairwise comparison
judgements of objectives with respect to all factors, and force their reciprocals in
the survey questionnaire matrix.
Table 1. The comparison scale [Saaty, 1994]
Intensity of Definition
1 Two activities contribute equally to the objective
3 Experience and judgement slightly favor one
factor over another
5 Experience and judgement strongly favor one
factor over another
7 An factor is very strongly favored over another
and its dominance demonstrated in practice
9 The evidence favoring one factor over another is
of the highest possible order of affirmation
Reciprocals of If factor i has one of the above nonzero numbers
above assigned to it when compared to it when
compared with factor j, then j has the reciprocal
value when compared with i

7) The next step is the computation of the vector of priorities from the given
matrix by multiplying the n components in each row and taking the nth root. To
normalize, divide each of the nth root by the sum of the nth roots.

Afterthe vector of priorities is solved, the next step is to determine the

maximum or principal eigenvalue (ëmax).
The solution is obtained by multiplying the
matrix of comparisons on the right by the priority vectors for obtaining a new
vector, then divide the first component of this vector by the first component of
priority vectors, the second component of the new vector by the second
component of priority vectors and so on. Which, it can be obtained the second
new vector. Finally, the result of solving is came from take the sum of the
second new vector components and dividing by the number of components.

The consistency index (C.!.) can be represented by ( max – n)/(n – 1). It shall

apply to determine the consistency ratio (C.R.) which it can be measured for the
goodness of judgements. C.R. can be determined by the ratio between C.!. and
the random index (R.!.). Where, R.!. is depending on the order of the matrix n.
as shown in Table 2. If C.R. is less than or equal to 0.1 judgements are
consistent. If C.R. greater than 0.1 the quality of judgement should be improved.
Table 2. The order of the matrix (n) and the average R.I. [Saaty, 1994]
n 1 2 3 4 5 6 7 8 9 10
R.I. 0 0 0.52 0.89 1.11 1.25 1.35 1.40 1.45 1.49

8) Finally, the overall ranking of alternatives shall be calculated by multiplying

the priorities vector of factors with respect to their impact with the priorities
vector of objective with respect to all factors.

From AHP process, one important thing is that questioner should concern
carefully about the fulfillment of pairwise matrix by experts. Especially, the
fulfillment of scale value, the process should be demonstrated clearly step by


After the six variables were selected as main factors for strategic contract
options of PCP-SMC in MRT construction project and AHP approach has been
illustrated in section 3. It shall be developed a hierarchical structure that shown
in Figure 1. It shall be applied for the comparison of the contract option (Single
and Multiple contract) in each case study and using mathematics for ranking the
results. During the interviews, three case studies were explained to simulate the
projects in stage of consultant procurement strategy. The basic information of
three case studies was described below.

Strategic Contract Options

Figure 1. Hierarchical structure of strategic contract options

1) The M.R. T. Chaloem Ratchamongkhon Line: All of the line is underground

PVa total length
route and PC of 27 km. UCconsists ofCO
QSThe project 18 subwayOC
along the route, 1 depot, 2 parking buildings and 8 ground-level parking lots.
Project value is equal to 116,379 Million Baht. Construction duration is 7 years
(1997-2004). Single(S) Multiple(M)
Contract Contract
2)MRT Purple Line Project : Bang Yai to Bang Sue Section ; The
project consists of 16 elevated stations along the alignment, 1 depot and 4
park & ride buildings. The line total length of 23 km. and project value is
equal to 49,298 Million Baht. The construction period is 5 years (2009-
2014). 1
3)MRT Blue Line Extension1/5 Project : Hua Lamphong - Bang Khae and
Bang Sue - Tha Phra Sections; 3 The project consists of 13 stations
along the alignment, 1 depot 1.32 and 2 park & ride buildings. There
are 5.4 km. of underground 1.19 structure and 21.5 km. of elevated
structure. Project value only PC civil work is equal to 52,460 Million Baht.
Construction period of the project
1 is 66 months (2010-2016).
Data analysis aims to prioritize the 1.44 important factors by using the weight
comparison. In this stage, 15 experts 1.24 had validated pairwise matrix for the
project. The example from one of case 5
studies of solving the priority vector is
shown in Table 3. After averaging from 1 15 experts judgement, the results of the
priority vector for the six factors are shown 3 in Table 4. The highest priority
vector of factor (0.26) is Project Characteristics
405.00 (PC). The second priority vector
of factor (0.25) is Quality of Consultant0.37 Service (QS) and the lowest priority
vector of factor (0.09) is Unity of Consultant
7.00 Team (UC). From all of the pairwair
matrix, the consistency ratio was passed 1/3
0.1 which are acceptable cases. The
results of priority vector for the six factors
1 shall be applied in case studies.
Table 3. An example of comparison of factors
with respect to PCP-SMC from an expert
Factor 1/3


Table 4. Average priority vector of factors for PCP-SMC

Factor Priority Vector Ranking

1) Project Construction Value(PV) 0.18 3

2) Project Characteristics (PC) 0.26 1

3) Quality of Consultant Services (QS) 0.25 2

4) Unity of Consultant Team (UC) 0.09 6

? max =
5) Communication (CO) 6.55 0.11 5
R.I. =
6) Overall Work Checking in Consultant
C.I. = Team(OC) 0.12 4
C.R. =
To get the overall ranking of PCP-SMC, it needs to multiply the weight
indicating the comparison of contract options with respect to the main group
factor by the weight of those factors. The example from one of expert
judgement for the comparison of PCP-SMC with respect to the six factors is
shown in Table 5. The overall ranking results of three case studies are
summarized in Table 6.

Table 5. An example of comparison of PCP-SMC with respect to the six factors from
an expert

1) Project Construction Value(PV) 2) Project Characteristics (PC)

PV S 1 1/7 0.14 0.38 0.13 PC S 1 1/3 0.33 0.58 0.25
M 7 1 7.00 2.65 0.88 M 3 1 3.00 1.73 0.75
Sum 3.02 1.00 Su 2.31 1.00
3) Quality of Consultant Services (QS) 4) Unity of Consultant Team (UC)
QS S 1 1/3 0.33 0.58 0.25 UC S 1 3 3.00 1.73 0.75
M 3 1 3.00 1.73 0.75 M 1/3 1 0.33 0.58 0.25
Sum 2.31 1.00 Su 2.31 1.00

5) Communication ( CO) 6) Overall Work Checking (OC)

CO S 1 5 5.00 2.24 0.83 OC S 1 1/5 0.20 0.45 0.17
M 1/5 1 0.20 0.45 0.17 M 5 1 5.00 2.24 0.83
Sum 2.68 1.00 Su 2.68 1.00
Table 6. The overall ranking results
Strategic Contract Option Actual
Case Study Single Contract
(S) (M)

1.The M.R.T. Chaloem Ratchamongkhon Line M M

0.46 0.54
2.MRT Purple Line Project 0.53 0.47
3.MRT Blue Line Extension Project 0.40 0.60

According to the weight of the six main factors that indicated from Table 4. The
highest relative factor of PCP-SMC is Project Characteristics (PC). Next is
Quality of Consultant Services (QS), Project Construction Value (PV), Overall
Work Checking in Consultant Team (OC), Communication (CO), and the lowest
factor is Unity of Consultant Team (U C).

Based on the results of the six important factors, it can be concluded that
Project Characteristics may effect contract options between single and multiple
contract. Thus, the client shall concern about contract options carefully. Unity of
Consultant Team shows the lowest factor, it could be pointed that the Unity of
Consultant Team may have less influence on selecting contract options. In
addition, the unity among consultant teams may be insinificant.

Regarding to the overall ranking, results of three case studies show in

Table 6. The M.R.T. Chaloem Ratchamongkhon Line and MRT Blue Line
Extension Project prefer to use multiple contract option, which these are 0.54
and 0.60 respectively. MRT Purple Line Project prefers to use single contract
option, which it is 0.53. All of strategic contract options for three case studies
are similar to actual contract option. The conclusion on three case studies can
be confirm that each project should be preferred on different contract options,
depending on the six important factors that had been presented.


Project consultant procurement is one of important processes which the clients

should concern and select at the beginning stage. AHP is one of techniques
that can be adopted for selecting strategic contract option. Which, it is involved
with goal, criterion and alternatives. The comparison can be used for qualitative
and quantitative data. The strategic option of PCP-SMC in MRT construction
project depend on factors for decision making. The results of research show six
main factors that influence selecting contract option. Also, it can be help to
make the decision making on similar projects in future time.

Weight of the six factors in this research could be applied on similar projects.
However, the overall ranking on contract option should be calculated case by


Duangkamol, D. (2010). Railway System. Civil Engineering Magazine. pp 5 – 7.

FIDIC (2003). Selection of Consultants. FIDIC Guidelines for the Selection of

Consultants: First Edition. pp 107 – 123.

H.S.C. Perera and W.K.R. Costa (2009). Analytic Hierarchy Process for the
selection of ERP systems for manufacturing companies.. The Journal of
Business Perspective l Vol. 12 l No. 4. pp 1 – 11.

New South Wales Government (NSWG) (2008). Single and multiple contract
options. Procurement Practice Guide. pp 1 – 2.

Office of Transport and Traffic Policy and Planning(OTP) (2010). Mass Rapid
Transit Master Plan in Bangkok Metropolitan Region. pp 16 – 20.

Office of National Economic and Social Development Board (NESDB) (2007). PPP
Analysis Report of MRT Blue Line Extension Project.. pp 3 – 5.

Sutham A. (2009). Analysis Hierarchy Process: AHP. Process management

(vol.64)j. pp 1 – 7.

Thomas L. Saaty (1980). The Analytic Hierarchy Process. McGraw-hill, Inc.

pp 17 – 34.USA.

Thomas L. Saaty (1994). Fundamentals of Decision Making and Priority Theory:

with the Analytic Hierarchy Process. RWS Publications. pp 69 – 84.Pittsurgh.
Mimi Suriani Mat Daud 1*, Intan Rohani Endut 1, and Harlina Suzana Jaafar 1,
2 3

Malaysian Institute of Transport (MITRANS), 2 Faculty of Civil Engineering,


University Technology MARA, 40450 Shah Alam, Selangor, MALAYSIA, 3

Faculty of
Business Administration, University Technology MARA, 40450 Shah Alam,
Selangor, MALAYSIA
sue, intan@salam .uitm .edu .my,

ABSTRACT: Cargo transportation is the important sector in developing country. After

recover from economic downturn last year, many countries including Malaysia shows
increasing of traffic container in port. The growth of cargo arise from many factors such as
changes in industrial structure, production mix, ways establishments on doing the business
and by population. Because of that, it is essential to get information about the factors that are
important to describe in more detail and where to focus the efforts of the works. The
objective of this paper is to analyze cargo flow in Malaysia from 2005 to 2009. This analysis
will consider the secondary data provided by Road Transport Department and Public Work
Department (PWD) in order to know the trend of the cargo on federal roads. Correlation test
are also done to know the factors that affect the generating and attracting of the cargo flow.
Lastly from the trend, projecting for future cargo is being done to predict the growth for the

Keywords : cargo, truck, planning, federal roads

Cargo transport is one of backbones in any industrial system (Willis, 2010,
Sgouridis, 2003). Despite it support the transportation of commodities and raw
materials to be processed, it is also important for transportation of the end products
to the local and international markets (FHWA, 2010). Malaysian economy also
depends on cargo transport. Between the years of 2007-2011, transport and
communications sectors are expected to achieve an average annual growth of
5.5%, for overall Gross Domestic Product (GDP) and the total value of transport and
communications GDP is expected to rise to US$17.7bn by 2011, representing 7.4%
of Malaysia’s GDP (Statistic department, 2010).

The importance of cargo transport can be seen very clearly; domestic

manufacturers rely on cargo transport for the sources of raw materials to produce
goods; wholesalers and retailers depend on cargo transport to obtain specialized
goods; and households and small businesses depend on cargo transport to deliver
purchases directly to them. With the burst of ‘dotcom bubble’ in 21 st
century, cargo
transport is expected to cover a larger geographic area and be more responsive to
user needs (Hu, 2001; Holguin-Veras and Thorson, 2003). This has being agreed by
leading shipping companies during the World Express & Mail Conference 1996 in
Brussels, where out of 50 shipping participant, more than half cited ‘dotcom bubble’
or e-commerce is the most important factor that driving their growth in business
(Holguin-Veras and Thorson, 2003).

In addition to the economy grows over the next several decades, the
demand for goods will only increase (Holguin-Veras and Thorson, 2003, Sgouridis,
2003). As truck are dominant in cargo transport, the capability of roads and highway
in cater the volume is be wondering. Based on estimates container throughput and
average daily truck trip (Table 1), the transportation system in Malaysia moved an
average of 10013 TEUs of cargo per days from Port Klang and 8241 TEUs per days
from Port Tanjung Pelepas in 2009. More than half of the TEUs moved was an
import from or export to another country and less than 25 percent was moved within
the local areas (Table 2).

Table 1: Containerized throughput and estimated average daily truck trips from 2 top
seaports, 2009
ports Annual Annual one way trips Average daily
TEUs (assuming 2 TEU per one way
truck) truck trips

Port Klang 7,309,779 3,654,890 10013

Tan gjung Pelepas 6,016,451 3,008,226 8241

SOURCE: data compiled by authors from port website, December 2010

Table 2: Statistical data about transportation of cargoes by in Malaysia from 2005 to 2009
Year Types of Traffic Total
National Traffic
Import Export
2005 805,157 1,342,901 1,276,661 3,424,719
2006 836,579 1,403,946 1,367,625 3,608,150
2007 871,234 1,527,893 1,474,193 3,873,320
2008 909,243 1,629,977 1,598,544 4,137,764
2009 936,222 1,515,743 1,478,354 3,930,319

SOURCE : data compiled by author from Road Transport Department and port website,
December 2010
From the result in Table 1 and 2, it is understandable that the role of cargo transport
as the conveyor of goods is increasing. Though there is a pressure from the
community and environmental group regarding to negative impact in cargo’s
activities (air quality problem, noise pollution, pavement damages, accident problem
and etc), but there is no doubt that cargo transportation has makes significant
contribution to the vitality of nation’s economy (Jonnavithula, 2004).

As the objective of this study is to analyze cargo flow in Malaysia, it is

essential to understand the flow that happened in Malaysia and thus the relationship
between the years, time and district as these factors will be the important element in
identifying the area that generate cargo demand.


As the objective of this study is to analysis the cargo flow in Malaysia from 2005 to
2009, the methodology involves are the general analysis of the cargo growth during
the 2005 to 2009, correlation analysis to determine the strength between the factors
of year, time and district in determining the area that generate and attracting the
cargo demand and lastly projection of future growth of cargo based on historical
trend by using compound growth factor to determine the future cargo demand.

Cargo traffic flow can be represented in many different ways, depending on the
mode, type of vehicle, and commodity (Jean-Loup and Kalinowska, 2008). A
common representation that usually being measure is in terms of number of vehicles
where one twenty-foot equivalent units (TEU) represent a standard of 20-foot

In general, there are four controlling factors that impact cargo demand
(FHWA, 2010). These factors can be broadly grouped into the following categories
which were economic structure, industry supply chains and logistics, cargo
infrastructure/modes and cargo traffic flows. Every factor has an impact on the cargo
demand analysis, for example cargo demand has a direct correlation with the type
and amount of economic activity in a region and types of modes that carry cargo
provide different types of service.
From the analysis of cargo flow from 2005 to 2009, it can be seen that
vehicles registration (including van for commercial goods) on average has increased
1.2 times (fig. I). This situation is concurrent with the data publish by Road Transport
Department in Table 2. Majority of these movements are contributing from the
international trade activities (import and export).

Figure I: Lorry and van registration by year

SOURCE: Department of Transportation, December 2010
Lorry & Van registration by year
Based on the product classification, the growth in export activities was
contribute largely from fixed vegetable oils and fats, crude, refined or fractionated
and gas, and manufactured,
850,000 and for import activities, it was contribute largely by
petroleum, petroleum800,000 2005
products & related materials (Statistic department, 2010).
750,000 2007
Intermediate goods contribute about RM331.4 billion to the total imports
700,000 2008
meanwhile capital goods contribute about RM69.1
2009 billion. Consumption goods in
other hand contribute
& van RM31.3 billion 836,579
(Statistic department, 2010). With the
continuing of economy growth in 2011, cargos 871,234
handling in ports are expected to
achieve a 7 percent growth, for both in the import
949,698exports activity, and thus makes
the transportation by truck becomes more important .

Malaysia’s growth in exports was contributed by higher exports to the

People’s Republic of China (RM894.6 million), Japan (RM853.9 million) and Taiwan
(RM444.1 million) which amounted to RM2.2 billion (83.3%) of total increase, year-
on year. As for imports, the growth was attributed mainly to higher imports from the
Republic of Singapore (RM656.9 million), Japan (RM611.1 million), the United
States of America (RM448.3 million) and the Republic of Indonesia (RM424.0
million) which contributed 85.1% or RM2.1 billion of total increase (Statistic
department, 2010).


This analysis is done to know the strength of relationship for each preliminary
element (volume, district, year and time) and this analysis is taking example of
Selangor state. Selangor is choosing because it is where port klang, the bussiest
port in Malaysia, located. The analysis is consisting of correlation between the
volume versus time, volume versus years and volume versus district. This analysis
is done in order to know the factors that influence the demand of cargo truck at one
area. The fist analysis is correlation between the volume and time. This analysis is
based on the data provide by public work department (PWD). This analysis is done
to know either the peak hour suggested in the report is representing the real peak
hour for truck traffic. From the result, R2 for Pearson’s correlation is 0. 286, which
indicates that the peak hour time provide in the report did not representing the peak
hour for truck traffic.

The second analysis is correlation between year and volume. This analysis
is done to know either year (usually related with economic indicators) is related with
the demand for truck cargo. From the result, the value of R 2 is 1.00, indicates that
years and the demand for truck cargo has strong relationship. This result also
proves that economy is one of the factors affecting the cargo demand as more
active the economy, more trucks are needed to transfer the goods.

The last correlation analysis is between the volume and the district. This
analysis is done to find out either district is play an important role in generate the
demand for truck cargo. The R2 for this analysis is 0.521, and from the guideline
provide by Cohen, 1988, this result is consider as medium, which means that some
of the district is responsible for generating truck cargo and some are not.
From the overall result provide by correlation analysis, it can be understand
that years and district play an important role in generating and attracting the cargo


From the secondary data also, the projection for cargo truck in 2020 can be done. In
project the cargo truck growth, there are two methods can be apply which were
either applying growth factor to the baseline cargo traffic data or applying growth
factor to the economic activity. The common method used in project future demand
is based on the baseline cargo traffic data.

In this method, the growth factor is calculated based on the historical traffic
information to the baseline traffic data. This technique are frequently used by the
department of transportation and other planning agencies, to establish rough
estimates of growth for a variety of demand and are certainly applicable to
establishing the cargo traffic for the cargo component of a transportation plan,
program, or project design.

In forecast of truck cargo by trip, compound growth method has been done in
this analysis. By assuming that cargo flow grows in a compound fashion, the annual
growth factor will be the ratio of the flow in the second and first raise to a power
which is the inverse of the number of years between the first and second
observations. The formula of the growth factor is as follow:

AGF = (F 2/F1)1/(Y2-Y1)

Where F 1 is cargo flow in year Y 1 and F 2 is cargo demand in year Y 2. This also can
be expressed as a compound annual growth rate by subtracting 100 percent from
the AGF.

The compound growth factor then be applied to predict future demand (F 3)

for some future year (Y 3) as follows:

F3 = F 2*AGF (Y3-Y2)

If more than two years of historical data are available for the variable to be forecast,
this data can be used to solve a power regression according to the formula:

F(n) = Constant*AGF (n)

Where n is the number of years from the first observation and Constant and AGF
are found from the linear regression. In this case, with an R-square of 0.7972 (Fig.
2) the coefficients therefore are:
F(Y) = 3,424,726*(1.035) (n)
projection of cargo flow from2 2005 to 2020
R = 0.7972

projection of cargo flow from 2005 to 2020

Linear (projection of cargo flow from 2005 to 2020)


Figure 2: projection of cargo flow from 2005 to 2020

Table 3: Statistical data about projection of cargoes transport in Malaysia from 2005 to 2020
Years from Compound
Year Traffic Ln(Traffic) 2005 Regression
2005 3,424,719 15.04653 0 3,424,726
2006 3,608,150 15.09871 1 3,544,591
2007 3,873,320 15.16962 2 3,668,652
2008 4,137,764 15.23567 3 3,797,054
2009 3,930,319 15.18423 4 3,929,951
2010 no data no data 5 4,067,500
2015 no data no data 10 4,830,914
2020 no data no data 15 5,737,610

SOURCE : data compiled by author from Road Transport Department and port website,
December 2010

From the figure 2 and table 3, it can be concluded that the cargo truck growth is
projected will up to more than 5 million in 2020 and since the value of R
2 is 0.7972
(near to 1.0), then the result is consider acceptable.

As currently, the development of cargo planning model in Malaysia have a limitation
in practical and reliable tools. One of the limitations is due to the complexity in cargo
data collection. Because of lacking in data collection, the following problems are
observed: loss of the cargo flow volume, increase of delivery costs, lack of
coordination in usage of modes of transport, and etc.

In conclusion, from the analysis of current status of cargo truck (based on

the secondary data), the following results are observed:

1. Import/export activities contribute major portion to the volume of cargo

truck on the road and with the economy growth in 2011, the numbers will
kept increasing.

2. Majority of the import activities in roadways are from the intermediate

goods and for export activities, it was from the electric and electronic

3. Malaysian top ten export destination are Republic of Singapore, the

People’s Republic of China, the European Union (EU), Japan, the United
States of America, Thailand, Hong Kong, Australia, the Republic of
Korea and India, meanwhile top ten for import activities are People’s
Republic of China, Japan, the Republic of Singapore, the United States
of America, the European Union (EU), Thailand, the Republic of
Indonesia, the Republic of Korea, Taiwan and Hong Kong.

4. District and years plays an important role in generating cargo truck

demand. Years are related with the economy growth and district are
related with the location of the demand either for the origin or destination

5. Based on the result of projected cargo truck demand, the numbers of

truck that will used highway and route ways are up to more than 5
millions in 2020, which will contribute to the traffic problem and others
problem as well if not handle carefully.


The author is grateful to the University Technology Mara, Malaysian Institute of

Transport (MITRANS) and Malaysian Logistic Council (MLC) for funding the
research work. Special thanks also to Dr. Intan Rohani Endut and Dr. Harlina
Suzana Jaafar for their continuous support and encouragement.
Jiang, F., Calzada.C., and Johnson.P., (1999), Cargo Demand
Characteristics and Mode Choice , Journal of Transportation and
Statistics .

Field, M., (2002), Going the Distance- the Challenge of Moving Highway
Cargo From Origin to Destination in the 21st century , CargoTtransportation

Johnson,D., and Ammah-Togue.F., (2004), Understanding Potential Cargo

Bottlenecks in the United States ,7th MTS Research and Technology
Coordination. Washington, D.C.
Jonnavithula, S. S., (2004), Development of Structural Equations Models of
Statewide Cargo Flow , South Florida.

Hesse, M., and Rodrigueb, J.P., (2004), T he Transport Geography of

Logistics and Cargo Distribution , Journal of Transport Geography .

Notteboom, T., and Rodrigue, J.P (2008), Containerisation, Box Logistics

and Global Supply Chains: The Integration of Ports and Liner Shipping
Networks , Maritime Economics & Logistics , 152-174.

Woodburn, A., and Mckinnon, A.C., (1995), Logistical Restructuring and

Road Cargo Traffic Growth, Journal of Transportation .

Holguin-Veras, J., and Jara-Diaz, S. (1999), Optimal Space Allocation And Pricing
For Priority Service At Container Ports , Transport Research Part B 33 (2),

Holguin-Veras, J., and Thorson, E.,(2002), Trip Length Distribution in commodity

Based and Trip-based Cargo Demand Modelling: Investigation of
Relationships , Transportation Research Record 1707, TRB, National
Research Council, Washington, D.C., pp. 37-48.

Statistic Department,20 10. Website:

Date access: 2 January 2011.
Moa, S., and Demetsky, J.M., (2002), Calibration of the Gravity Model for Truck
Freight Distribution , Research Report, No. UVACTS-5-14-14.

1. correlation between volume

versus time Correlations
time 2009
time Pearson Correlation 1 .286
Sig. (2-tailed) .367
N 12 12
2009 Pearson Correlation .286 1
Sig. (2-tailed) .367
N 12 12

2. correlation
volume year
volume Pearson Correlation 1 1.000**
Sig. (2-tailed) .000
N 9 9
year Pearson Correlation 1.000** 1
Sig. (2-tailed) .000
N 9 9
**. Correlation is significant at the 0.01 level

3. Correlation between volume

versus district Correlations
district volume
district Pearson Correlation 1 -.521
Sig. (2-tailed) .083
N 12 12
volume Pearson Correlation -.521 1
Sig. (2-tailed) .083
N 12 12

0. Value of r as suggested by Cohen

(1988) to be guidelines:
r=0.10 to 0.29 or r=-0.10 to r=-0.29 small
r=0.30 to r=0.49 or r=-0.30 to r=-0.49 medium
r=0.50 to 1.0 or r=-0.50 to r=-1 .0 large
Sig. (2tailed) N
volume year
volume Pearson correlation
Kian Ahmadi Azari 1,*, Sulistyo Arintono 2
, Hussain Hamid 3 and Riza Atiq O.K.

Rah mat 4

Department of Civil Engineering, Faculty of Engineering, Universiti Putra


Malaysia, 43400 Serdang, Selangor Darul Ehsan, Malaysia.

Department of Civil Engineering, Faculty of Engineering, Universiti Kebangsaan
Malaysia, 43600 UKM Bangi, Selangor Darul Ehsan, Malaysia.

ABSTRACT: Congestion pricing as a policy to alleviate congestion, environmental and other

problems related to urban areas is still widely considered to be a beneficial and effectual
policy. In this paper, the impacts of cordon pricing in Mashhad city, on traffic volumes, mode
choice, and cordon and parking fees have been investigated using two different surveys.
During charging time period, the high cost sensitivity is fully gave a reason by the fact that
private car travelers reduced about 34% crossing the cordon have the most effects on traffic
congestion. Furthermore, result show that in-vehicle and cordon fee as mode effective
attributes include the least and the most impacts on mode choice.

Keywords: Cordon pricing, Parking cost, Mode choice.


Rapidly growing of population, increasing in real incomes and geographical

expansion of cities in recent years has caused more ownership of cars, fewer
passengers per car and longer trips in many developing cities. Mashhad city as
the centre of Khorasan Razavi province located in the north-east of Iran is not
excluding from this rule (Mashhad, 2004). One of the 163 most populated cities
of the world, Mashhad is the most populated city of Iran with 2. million citizens
in 2007 after Tehran, and is known as a religious city (Mashhad, 2010a). In
addition to its population a great number of guess add to the city, around 20
million tourists, generate some kinds of problems, such as various kinds of
environmental pollutions and inflict a lot of costs by causing heavy traffic and
wasting commuter’s time.

The Mashhad Central Business District (CBD) congestion management

program started late March 2007 as an experimental project. Meanwhile,
according to comprehensive study a seasonal restricted zone defined. The goal
is reducing peak period traffic flows throughout the CBD, noise and air pollution
(Mashhad, 2010b).
Cordon area congestion pricing is a tax or fee paid by users to enter a restricted
area, usually within a city centre, as part of a demand management strategy to
relieve traffic congestion within that area. A few cordon schemes have been
successfully implemented up until now, gained in several cities. The first and
successful implementation of cordon pricing technique can be acquired in
Singapore started in 1997 known as Singapore Area Licensing Scheme.
Upgraded its system in 1998, Singapore have proposed and tested the most
advanced and possibly electronic road pricing schemes worldwide (Goh, 2002,
Olszewski and Xie, 2005, Tuan Seik, 2000).
The practical experience in implementation of this scheme has thrived
rapidly in several European cities. The most famous is London congestion
charge or inner ring road in 2003 with extensions in 2007 (Goodwin, 2004,
Leape, 2006a, Santos and Bhakar, 2006). Moreover, Rome in 2001, Stockholm
after seven month trial in 2006 (Armelius and Hultkrantz, 2006, Gudmundsson et
al., 2009), Oslo, Bergen and Trondheim toll rings in the Norway (Larsen, 1995,
Ieromonachou et al., 2006, Odeck and Bråthen, 2002, Lian, 2008) and more
recently, so-called Ecopass, Milan trial program, in 2008 are the other sample of
the cities successfully implemented.

Many cordon pricing issues are still controversial both at a theoretical and
empirical level. From a theoretical point of view, the vast literatures of cordon
pricing have proposed a number of valuation techniques methods, and models
to separately determine various aspects of the wider effects of cordon pricing
which can be composed in two categories, short-term, long-term changes.

The short-term implications of road pricing has focused on changes in the

travel behavior of users, transport network performance measures (Santos,
2002, Mun et al., 2003, Ho et al., 2005, May et al., 2002, Verhoef, 2002). In
addition, the potential deployment of a number of cordon pricing schemes has
been studied at the grounds of feasibility, revenue generation, efficiency and
acceptability in various urban areas (Schade and Schlag, 2003, Giuliano, 1992).
Long-term impacts of toll welfare-maximizing and change of job or residential
cordon pricing schemes(Santos, 2002, Mun et al., 2003) are implicitly modeled
with the joint consideration of capacity investment (network design) decisions.

At an empirical level, there are some studies and overviews to evaluate

effects on traffic congestion and pollution and changes on residents and road
user's behavior after introduction of cordon charges. In London (Santos and
Bhakar, 2006), Singapore (Tuan Seik, 2000), Stockholm (Eliasson et al., 2009),
Oslo (Odeck and Bråthen, 2002) and Milan (Rotaris et al., 2010). To author's
knowledge all the number of real world implementations happened in developed
cities, yet far too small, there is not any in developing countries. Therefore, for a
better judgment these cases are too specific to allow the scientific community to
draw definite conclusions not in similar society but in other developing urban
areas with various social perspectives.

For a Mashhad city pricing zone involving a flat toll, an estimate of traffic
impact can be made with the knowledge of traffic demand before and after
restricted zone implementation. This article deals with traffic effects after the
cordon was introduced in Mashhad city. Furthermore, the impacts of mode and
parking attributes on mode choice, and cordon and parking charges on
commuters have been investigated.


A. The plan of traffic restrictions in Mashhad city CBD was carried out to improve
traffic and easy access of commuters and pilgrims, reduce air pollution and
traffic congestion in this area (Mashhad, 2010b).

0. During the seasonal implementation, charges were imposed on vehicles

passing a cordon around the central business district of Mashhad as depicted in
the map of Figure 1. Regarding the ministry of interior and the high council of
traffic approvals, the Mashhad Central Business District (CBD) congestion
management program starts every summer since 2007 on six days a week
(Fridays excluded) and would cover the entire streets end at Imam Reza holy

The restricted area of the toll zone is around 530 hectare square meters and
its border length is 8.30 km. From the latest studies on September 2010, 70000
vehicles entering inside the zone with 5 km2 but due to the inadequate road
capacity just 35000 permission was issued (Mashhad, 2010c). The Existence of
business, religious and cultural, educational and service centres is causing the
uptake trip very high in downtown area which is about 20 percent of daily trips
generation in Mashhad city. In addition, the increase average speed of 10 to 20
km per hour of urban traffic and subsequently 40 percent reduction in emissions
of carbon-dioxide produced by vehicles are as a result of cordon changing
scheme (Mashhad, 2010c).
Charges were two categories over the day. The fee for passing a control
point was TN 1 2000 per day (corresponding to US Dollars 2.0) between 7:00 and
21:30 (While London initially charged £5 per day, between 7:00 and 18:30, later
raised to £8 per day and Stockholm was Euro 1.1, 1.6, and 2.2, respectively
depending on the time of day). Fees were not levied in nights, Friday (as a
casual weekend) and public holidays. The drivers that pass the borders into the
zone without any legal justification will be penalized.Various exemptions (for e.g.
taxis, buses, residential commuters who live inside the CBD) were free-of -

Figure 1. The Mashhad traffic restricted area. The solid black line is the charging cordon.


The effects of congestion pricing can be separated into different parts but in this
paper the focus is on:
Decreasing traffic volume for those who travel by car in the charged areas
during charged hours.

Changed travel behavior to avoid the charges—for example by switching mode

and destination or higher travel costs for those who travel by car in the charged
areas during charged hours.

1The toman is a superunit of the official currency, Rial. 1 toman (TN)= 10 rials (IRR) and
1000 tomans =1 U.S. dollar ($)
In our assessment of the proposed cordon pricing scheme, we want to cover
three major objectives which based on two different surveys. In order to collect
information on commuter's travel pattern and their effects on traffic volume, the
first survey was carried out before (September 21) and after (September 28)
implementation of cordon charges on Monday in the summer of 2010. During the
period of 7.30–22.30, traffic counts were done in the entrance section of
restricted zone in four major roads connected to the CBD. Total types of all
vehicles were estimated entering cordon area.
In addition, on 14 to 16 August 2010 a random sample size of 580
respondents was selected to perform a comprehensive survey to identify the
responsiveness of drivers` behavior to mode and parking attributes, and cordon
and parking fees in Mashhad city during the seasonal cordon introduction. This
was a face-to-face interview from the drivers traveling inside the cordon zone.
Data are collected by personal interviews with drivers on their departure from
some parking locations in and outside the CBD.

4.1 Effects on traffic volumes
In the 21st September 2010, after implementation of cordon pricing it was found
that average daily traffic flow during weekday in the restricted zone fell in volume
by 7.6% from over 73069 under 67511 and the number of all vehicle types
dropped by 13.6% from 61446 to 53123.

The traffic volume result has been a significant change in the composition of
Mashhad traffic, as measured by vehicle number driven within the charging zone
(see Table 1). Passenger cars, which was reported for almost more than 47% of
city centre traffic before the charge was imposed, now represent almost 33%, a
drop of 36%. In addition, in the case of private user the congestion levels of vans
and private buses have decreased by 24.8 and 11.9 percent respectively versus
trucks have risen slightly. In relative terms, as other vehicles were free of charge
there have been a sharp rise in motorcycle and bicycles (around 20%), taxis (up
17%) and public buses (around 27%) but the maximum is on minibus with 53.1%

The basic concept of the drop in traffic volume and congestion is that some
commuters have decreased their number of trips on cordon roads; some have
switched the commuting time to free charged hours while others have switched
to public transport (taxis, minibuses, and buses) for some or all of their trips.
These outcomes do indicate that the cordon charge would have a clear negative
impression on car users’ expenditure in the CBD. On the other hand, public
transport users are more increased by the introduction of the cordon charge as

Table1. Impact of the cordon fee on Number of vehicles in the congestion charging zone
in Mashhad
Vehicle Type No. of vehicles Cordon percentage
Before After charge change
Bicycles and
12718 (20.7%) 15206 (28.6%) Free 19.6%
Passenger car 34560 (56.2%) 22050 (41.5%) TN 2000 -36.2%
Taxi 8714 (14.2%) 10227 (19.3%) Free 17.4%
Van 2558 (4.2%) 1923 (3.6%) TN 2000 -24.8%
Minibus 508 (0.8%) 778 (1.5%) Free 53.1%
Public bus 2071 (3.4%) 2629 (4.9%) Free 26.9%
Private bus 101 (0.2%) 89 (0.2%) TN 2000 -11.9%
Trucks 216 (0.4%) 223 (0.4%) TN 2000 3.2%
All vehicles 61446 (1 00.0%) 53125 (1 00.0%) -13.5%

Generally, the decrease in hourly traffic volume was largest in the afternoon
peak period (22.4% between 17:30 and 18:30), and somewhat lower in the
morning peak period (18.7% between 9:30 and 10:30) (see Figure 2). Declining
in equivalent number of passenger cars during the restricted cordon hours have
proofed but around 35% rising between 20:30 to 21:30 indicates shifting in
commuting times in the evening. These reveal that a more portion of arbitrary
trips is made during the afternoon peak rather than in the morning. These kinds
of trips may belong to non-work purpose instead of work trip purpose whose
arrival times fixed more than departure time to work.
Figure 2. Distribution of traffic volume passing across the restricted boundary before and
after implementation of cordon charges
6000 7:30-
5000 9:30-
4000 10:30-
3000 12:30- Before After
2000 17:30-
1000 19:30-
0 20:30-
4.2 Effect of attributes on mode choice
In order to get a realistic picture of the changes in mode shifting, investigation of
driver's responsiveness to mode and parking choice attributes was conducted in
August 2010. On the one hand, the face to face survey was focused on drivers
who have traveled to the CBD including cordon fees with/without their private
cars and so the changes in switching to other modes are likely to happen.
Search time, walk/ egress time, parking fee, in-vehicle and cordon fee are the
five attributes using to estimate their effects on mode and parking choice. They
are defined as:

Search time: The time spent searching and queuing for the parking space,
Walk/ egress time: The time spent walking from parking to the final destination,
Parking cost : The total parking fee and
In-vehicle cost: The total cost paid for using public transport services to access
the CBD

Cordon cost: The entering cordon fee.

The first three attributes straightly but the rest indirectly influence the choice
of parking location. From an overall perspective these constants change mode
choice behavior. Figure 4 shows the responses to the importance of parking and
mode choice variables in changing mode from the survey participant. As it had
been expected, the effects of cordon charge factor on selecting different modes
has generally the highest impact by 64%. Search time, walk time, parking fee
and in-vehicle cost respectively by 56.0, 41.2, 40.1, 28.7 percents have been
placed in the next effective stages.
Figure 4. Effects of parking and mode variables on respondent mode shifting behaviour

Low effect Moderate effectHigh effect

41.2 40.1
64.0 56.0
29.9 30.1
18.0 25.7
28.9 29.8 39.9
18.0 18.3

Cordon fee Search time Walk timeParking fee In-vehicle fee

Further, as it was seen the effects on walk time and parking fee attribute
were nearly the same that it was hard to distinguish the efficient one. This is
likely because the parking cost in Mashhad city is low (around TN 200 per hour)
which confirms that why they are less sensitive to parking charges inside the
cordon area.

Urban cordon pricing is one of the effective solutions to eradicate the traffic
congestion problem excruciating most large cities. Since the Mashhad city had
been facing congestion as well as pollution from vehicle noise and exhaust
fumes, a seasonal restricted zone for implementation of cordon charges
schemes have been defined on March 2007. The aim of this research is not only
to investigate the traffic effect but the behaviour of drivers to mode and parking
choice attributes, and cordon and parking charges in cordon area of Mashhad

Based on two survey result before and after implementation of cordon

pricing, the following conclusion was obtained. These outcomes do indicate that
the cordon charge would have a clear negative impression on car users’
expenditure in the CBD. On the other hand, public transport users are more
increased by the introduction of the cordon charge as expected.

Traffic was sensitive to changes in cordon, and had clear negative effects on
car users. An interesting outcome concerns the magnitude of this change. This
research showed that even though the average charge in Mashhad was
relatively low but the cordon fees resulted in reductions of traffic congestion in
the former periods compared to the expected effects of other measures that are
discussed in other cities.

On an average, there was a decrease in passenger cars over 36% for the hours
of cordon with the charge of TN 2000 per entry in Mashhad, whilst London
initially charged £5 (TN 8000) per day and the number of private cars coming
into central London dropped 34% between 2002 (before the charge) and 2003
(Leape, 2006b). Furthermore, the traffic in Stockholm was reduced around 22%
compared to 12 months before and after the first month of trial implementation in
2005 and 2006 with Euro 1.1, 1.6, and 2.2, respectively depending on the
time of day (Eliasson et al., 2009).

The other result of the study is as follows:

• Cordon charges can contribute to a more efficient utilisation of the road
network, as arbitrary trips can be decreased in the peak hours and shifted to the
off-peak when daily cordon fee is not charged.

• The experiences from Mashhad show that in-vehicle cost affects the least
and cordon fee includes the most impression on mode choice.
There are potential recommendations for economic and environmental
benefits to be gained by introducing Mashhad cordon charge scheme. The first
is establishing fully electronic with non-stop toll lanes from vehicles crossing
cordon area. For better public satisfaction, having time-differentiated charges
supporting traffic management objectives is the second proposal. Finally,
considering maximum limit of one chargeable crossing within specific time in per
hours or month where every single crossing have to be paid for.

ARMELIUS, H. & HULTKRANTZ, L. (2006) The politico-economic link between
public transport and road pricing: An ex-ante study of the Stockholm road-
pricing trial. Transport Policy, 13, 162-172.


Stockholm congestion - charging trial 2006: Overview of effects.
Transportation Research Part A: Policy and Practice, 43, 240-250.

GIULIANO, F. & RAMPIN, O. (2004) Neural control of erection. Physiology &

Behavior, 83, 189-201.

GIULIANO, G. (1992) An assessment of the political acceptability of congestion

pricing. Transportation, 19, 335-358.

GOH, M. (2002) Congestion management and electronic road pricing in Singapore.

Journal of Transport Geography, 10, 29-38.

GOODWIN, P. (2004) Congestion charging in central London: Lessons learned.

Planning Theory & Practice, 5, 501 - 505.


(2009) Framing the role of Decision Support in the case of Stockholm
Congestion Charging Trial. Transportation Research Part A: Policy and
Practice, 43, 258-268.

HO, H. W., WONG, S. C., YANG, H. & LOO, B. P. Y. (2005) Cordon-based

congestion pricing in a continuum traffic equilibrium system. Transportation
Research Part A: Policy and Practice, 39, 8 13-834.

IEROMONACHOU, P., POTTER, S. & WARREN, J. P. (2006) Norway's urban toll

rings: Evolving towards congestion charging? Transport Policy, 13, 367-378.
LARSEN, O. I. (1995) The toll cordons in Norway: an overview. Journal of Transport
Geography, 3, 187-197.
LEAPE, J. (2006a) The London Congestion Charge. Journal of Economic
Perspectives, 20, 157-176.

LEAPE, J. (2006b) The London Congestion Charge. Journal of Economic

Perspectives, 20-Number 4, 157–176.

LIAN, J. I. (2008) The Oslo and Bergen toll rings and road-building investment -
Effect on traffic development and congestion. Journal of Transport
Geography, 16, 174-181.

MASHHAD, R. (2004) Methods to deal with traffic problems during certain date.
Mashhad traffic studies during certain date. Mashhad transportation and
traffic organization, Report from Behin Taradod consultant.

MASHHAD, R. (2010a) Models to estimate population and growth. Updating

comprehensive study of transportation in Mashhad. Mashhad transportation
and traffic organization, Report from Tarhe Haftom consultant engineering.

MASHHAD, R. (2010b) Evaluation report within Mashhad Restricted CBD Zone.

Transportation and Traffic organization of MASHHAD, Report from Specialty
Group Traffic Construction Engineering Organization of Khorasan Razavi

MASHHAD, R. (2010c) Restricted CBD Zone. Mashhad transportation and traffic

organization, Available at
06-41 -09.html.

MAY, A. D., LIU, R., SHEPHERD, S. P. & SUMALEE, A. (2002) The impact of
cordon design on the performance of road pricing schemes. Transport
Policy, 9, 209-220.
MUN, S.-I., KONISHI, K.-J. & YOSHIKAWA, K. (2003) Optimal cordon pricing.
Journal of Urban Economics, 54, 21-38.

ODECK, J. & BRÅTHEN, S. (2002) Toll financing in Norway: The success, the
failures and perspectives for the future. Transport Policy, 9, 253-260.

OLSZEWSKI, P. & XIE, L. (2005) Modelling the effects of road pricing on traffic in
Singapore. Transportation Research Part A: Policy and Practice, 39, 755-

road pricing scheme to curb pollution in Milan, Italy: Description, impacts and
preliminary cost-benefit analysis assessment. Transportation Research Part
A: Policy and Practice, 44, 359-375.

SANTOS, G. (2002) Double Cordon Tolls in Urban Areas to Increase Social

Welfare. Transportation Research Record: Journal of the Transportation
Research Board, 1812, 53-59.

SANTOS, G. & BHAKAR, J. (2006) The impact of the London congestion charging
scheme on the generalised cost of car commuters to the city of London from
a value of travel time savings perspective. Transport Policy, 13, 22-33.
SCHADE, J. & SCHLAG, B. (2003) Acceptability of urban transport pricing
strategies. Transportation Research Part F: Traffic Psychology and
Behaviour, 6, 45-61.

TUAN SEIK, F. (2000) An advanced demand management instrument in urban

transport: Electronic road pricing in Singapore. Cities, 17, 33-45.

VERHOEF, E. T. (2002) Second-best congestion pricing in general networks.

Heuristic algorithms for finding second-best optimal toll levels and toll points.
Transportation Research Part B: Methodological, 36, 707-729.



Khoo Hooi Ling

Universiti Tunku Abdul Rahman, Kuala Lumpur, Malaysia

ABSTRACT: Travel time is a fundamental parameter in traffic engineering studies. It is also

an important criterion that travellers would consider in making their route choice when
travelling between an origin-destination (OD). Travel time between an OD is not a constant
due to dynamic traffic flow. Therefore, it is important to understand and predict such
variability in order to ensure an efficient traffic control and management scheme which could
mitigate the traffic congestion. The objectives of this study are two-fold. First, the travel time
variability with respect to time is analysed by plotting the graphs employing data obtained
from the detectors on the field. This provides some insight on how the travel time on the
highways changes for different time of a day and a week. Second, regression models are
developed to predict the travel time on the highways by relating it to some significant factors.
The study chooses the Middle Ring Road II, located in Klang Valley region, as the case
study. The results show that the travel time on the highway is varying for different time or
day. The regression models show that the variability is mainly due to incident occurrence,
weather, traffic flow and holidays. In the future study, investigation will be carried out to study
on other highways in the region.
Keywords: Multiple Linear Regression, Variability, Travel Time


Travel time is the most fundamental study in traffic engineering. Traffic

management agencies need to predict travel time in order to evaluate traffic
conditions on highways. Drivers are concerned about travel time in their route
choice in which they usually will choose to travel on the roads that they perceived
the least travel time. Nevertheless, the mean travel time between an origin-
destination is not constant at all time. It fluctuates and is affected by various factors
such as traffic flow, occurrence of incidents, the time of a day (peak period and off-
peak periods) and the day of a week (weekdays or weekends).

There are many researches carried out to predict travel time over the past
decades. There are a few prediction techniques adopted, such as statistical
modelling (Chen et al., 2003 and Van Lint et al., 2008), regression analysis (Cohen
and Southworth, 1999, Li, 2004, and MacDonald, 2008), Fourier transform (Ng and
Walker, 2010), and econometric modelling (Peer et al., 2009). In terms of variables
considered in modelling, Cohen and Southworth (1999) investigated the effect of
incidents on travel time in terms of mean and variance of time lost. They estimated
the mean and variance of delays due to freeway incidents, debris and vehicle
breakdown as a function of volume to capacity ratio. Chen et al. (2003) also
investigated the effects of incidents on travel time for weekdays. Results suggested
that incidents induced more variability during off peak periods. Li (2004) investigated
more factors such as incident occurrence, time of day, day of week and weather on
the CityLink freeway in Melbourne, Australia. The findings indicate that around 50%
of variability is influenced by the investigated factors. MacDonald (2008) used five
months data (exclude incident-related data) to develop functions in predicting day to
day variability for United Kingdom’s roadway. He showed that standard deviation of
travel time per km as cubic function of mean travel time per km is appropriate to
describe the day to day variability in highway.

This study makes a novel attempt to investigate the travel time variability on the
highways in Malaysia, specifically in Klang Valley region. The objectives of this
study are two-fold. First, the travel time variability with respect to time is analysed by
plotting the graphs employing data obtained from the detectors on the field. This
provides some insight on how the travel time on the highways changes for different
time of a day and a week. Second, regression models are developed to predict the
travel time on the highways by relating it to some significant contributing factors,
specifically the time of a day, the day of a week, incident occurrence, traffic flow,
weather, and holidays. Middle Ring Road II is adopted as the case study. Results
show that the travel time varies with respect to the time of a day and day of a week.
The factors mentioned above are found to be statistically significant in affecting the
travel time variability.

The study area chosen for the study is Middle Ring Road II (MRRII) which is a 2-
way highway connecting regions near boundary of Federal Territory of Kuala
Lumpur and Selangor. The total length of the highway is 35 km started at Sri
Damansara Interchange at the North and end at Jalan Bandar Tasik Selatan at the
South. The highway penetrates through a few busy towns such as Kepong,
Setapak, Ampang, and Cheras. The stretch of the road where the travel time is
derived is shown in Figure 1 which is about 30 km of the entire highway. The study
period is from 1st February 2006 to 31st July 2006 in which 2 specific periods are
defined for data analysis purposes, namely morning (7am-9am) and evening (4pm-
8pm) peak periods. Weekdays are the working days from Monday to Friday while
Saturday and Sunday are defined as weekends. Whenever there is no public
holiday or school holiday, the day is considered as normal day.
?t m (2)
= =

Figure 1. Study Area: Middle Ring Road II

The traffic flow and speed data is collected from the ITIS Transport Management
Centre (TMC) (CHKL, 2010) who has detectors installed on the road. The detectors
record data at every 3-minute intervals. The mean traffic flow is computed using the
following equation:
∑ ∑
i n t e r v a l
= 1 1

q ˆ ___________________________________(1)

where interval
q j,m indicates the mean traffic flow for interval j and hour m . Since an

interval of 3-minute is used in the study, J 20 ; and M indicates the number of =

hours in the specific period. For example, if morning peak period is considered,
M 3 ; qq N wher e q i is the 3-m inut e int er val tr affic flow f or det ec tor i

j,m i
= ∑

∀ i

and N indicates the total number of detectors on the stretch of the highway.
The mean travel time for the specific period is computed using the following
∑ ∑
i n t e r v a l

where interval
t j,m indicates the mean travel time for interval j and hour m . Since an

interval of 3-minute is used in the study, J 20 ; and M indicates the number of =

hours in the specific period. For example, if morning peak period is considered,

M 3 ;= ? i
t j interval
m =


∑ L vi i

stretch i and N indicates the total number stretches defined for the highway per
direction, i.e. N 18 .

There are a few independent variables considered for the travel time variability
model. The daily (24-hour) rainfall data is collected from Malaysia Meteorological
Department (MMD, 2010). Based on rainfall intensity, the storm severity is
categorized into 4 categories, in which level 0 defines sunny day, level 1 defines
light storm, level 3 defines medium storm, and level 4 defines very heavy fall. The
incident log report is obtained from ITIS TMC in order to understand whether there
are incidents happening on the highways. The number of incidents per day is used
as the input for the travel time variability model by assuming that the effect on the
traffic condition is the same regardless of incident type and severity. The effect of
public or school holiday is surveyed by retrieving the information from the online
calendar. The indicator of public or school holiday will be defined as 1 and 0 for
normal days. For day of week, indicator for weekday is defined as 1 and weekends
are input as 0. The mean travel time computed in eqn. (2) is normalized and used as
the dependent variables in the model.

Linear regression models are statistical models which are helpful in exploring the
relationships between the dependent variables and the independent variables. In
this study, the Best Subsets Regression method is employed to predict the travel
time based on a few variables mentioned above. Hypothesis testing needs to be
carried out to study the significance of the variables considered. For each of the
variable, t-statistic is adopted in the test while for the overall model, F-statistic is
adopted. The setting of the hypothesis testing is as follows:
H 01 2 ?

: 0
? =?

H: β ≠
0 for at least one j
1 j

By comparing the p-value, the significant of the variables could be obtained. For any
variable that has p-value lesser than the pre-determined level of confidence, the null
hypothesis in (3) is rejected and this indicates that the variable is significant.


Figure 2 shows the graphs plotted to demonstrate the travel time variability for
northbound and southbound traffic on MRRII. Figures 2A and 2B show the effect of
the different day of a week on the travel time variability. Both figures show similarity
pattern in which longer travel time occurs on 7am-9am and 5pm-8pm respectively. It
is also observed that drivers who are using the highway on weekdays suffer longer
travel time compared to weekends. Figures 2C and 2D show the effect of school
and public holidays on the travel time variability. For Southbound traffic (Figure 2C),
the travel time is about the same for time period from 12am to 3pm. There is slight
difference for evening peak in which travel time during public holidays is higher
compared to the other days. This indicates that there is “holiday” traffic attracted to
the highway since it is the major highway that connects to other highways at the
South (such as Kesas Expressway, PLUS Expressway at Sg. Besi, and Lekas
Expressway). For the North bound traffic (Figure 2D), it is shown that during school
holidays, there are two significant peak travel time observed on 10am and 7pm.
There might be more traffic attracted to the highway during holiday seasons since it
is the major highway connecting to Genting Highlands, the popular tourism place.
Nevertheless, the peak duration is short and occurs at different peak period
compared to normal days. Accordingly, the extra traffic attracted is causing little
impact on the daily commuter trips.

Figure 2. Travel Time Variability Plots

The travel time variability models developed for southbound traffic during
morning and evening peak period are shown in eqns. (4) and (5):

ln t ˆ = 3.09 + 0.0689 Inc - 0.133 P + 0.123 D

× × × (4)
(4.85, 0) (-3.73,0) (9.37,0)
lnt ˆ 4.45
= − 0.000271 q ˆ × + 0.0565 S 0.108 D× + × (5)
(-5.99, 0) (1.99, 0.048) (4.63, 0)
where tˆ (minutes) and qˆ (veh/hr) are the mean travel time and mean traffic flow

computed using eqn. (2) and eqn. (1) respectively; Inc , P, D, and S represent
incident, public holiday, day of the week and school holiday variables respectively.
The values in the bracket under the variables indicate the t-statistic and p-value
respectively, (t-statistic, p-value), while the bracket at the end of the equations
indicate the R2 and F-statistic respectively, [ 2
R , F-statistic].
The hypothesis testing shows that both models and their independent variables
are statistically significant at 95% confidence level. Eqn. (4) shows that the travel
time on the morning peak periods is most affected by the day of the week. Travel
time increases during weekdays and if there is incidents occur on the highway.
Nevertheless, the travel time reduces during public holidays. In the evening peak
(Eqn. (5) is referred), it is observed that the travel time is longer during weekdays
and school holidays. It is marginally affected by the mean traffic flow on the
highway. It is interesting to note that there is negative relationship between mean
travel time and mean traffic flow. This implies that the traffic condition on the
highway is in the “forced flow” condition which is over capacity. This has restricted
the vehicle flowing smoothly which cause longer travel time.
The travel time variability models for northbound traffic for morning and evening
peak periods are shown in eqns. (6) and (7) respectively:

lnt ˆ 2.98
= + 0.000059 q ˆ × + 0.0530 Inc 0.150 P 0.156 D (6)
×− × + ×

(4.05, 0) (3.22, 0.002) (-3.62, 0) (6.02, 0)


lnt ˆ 4.39
= − 0.000248 q ˆ × − 0.166 P 0.0799 D
× + × (7)

(-4.13, 0) (-2.5, 0.013) (3.3, 0.001)

1 1

where the symbols take the same definition as above-mentioned. The hypothesis
testing shows that the models are statistically significant and all the variables
investigated qre also significant at 95% confidence level. Both models show that
they are sensitive to the day of the week and public holiday indicators, and traffic
flow. It is observed that the travel time is longer for weekdays, which has higher
impact on the morning peak periods compared to evening peak periods. The mean
travel time reduces during public holidays and the impact is about the same for both
peak periods. Eqn. (6) shows that the mean travel time increases with the traffic
flow. Heavier traffic flow induced longer mean travel time if the traffic condition on
the highway is under “balanced” flow condition. The opposite is observed for eqn.
(7) which indicated that the traffic flow condition during the evening peak period is
under “forced flow” situation.

The travel time variability of MRRII is studied for a period of 6 months using the
data obtained from the loop detectors. According to the results obtained, the day of
the week variable has the greatest impact on the travel time variability. In average,
the travel time during weekdays is higher than during weekends. This indicates that
the highway is used by daily commuters who generate work trips. This could also
explain why the mean travel time is shorter during public holidays. The occurrence
of incidents increases the travel time during morning peak period for both sides of
the highway. This is because bottlenecks created from the incidents have worsened
the traffic condition on the highways. It is also observed that the traffic condition
during the evening peak periods is worse compared to morning peak period for both
directions. The negative relationship between the mean travel time and traffic flow
indicates low level of service of the highway. This is supported by Figure 2A and 2B
respectively as the peak of the plots occur in the evening peak periods. Weather or
storm level has no effect on the travel time variability.


Understanding of the travel time variability provides some fundamental studies to

the traffic flow condition on the highways. By investigating how the travel time
changes with respect to time, it gives insight on the traffic condition on the highways.
Using regression modelling, this study found that the contributing factors to the
travel time variability on MRRII are such as: day of the week, public and school
holidays, incident occurrence, and traffic flow. Weather or rainfall intensity is found
to have no impact on the mean travel time for MRRII. The limitation of this study is

that the interaction effect between the variables is not included in the regression
models. For example, the mean traffic flow on the highway might be influenced by
the time of the day, but such interaction is not considered. The traffic management
agencies could adopt the findings in this study to plan for an efficient traffic
management strategies that mitigate traffic congestion on MRRII especially during
evening peak periods. In the future study, the travel time variability with respect to
distant will be investigated to observe the bottleneck points that create congestion.


This participation of this paper in the conference is sponsored by Science Fund,

Ministry of Science and Technology (06-02-1 1-SF0085).


Chen, C., Van Zwet, E., Varaiya, P., & Skabardonis, A. (2003). Travel time reliability
as a measure of service. Transportation Research Record 1855 , pp. 74-79.

Cohen, H., & Southworth (1999). On the measurement and valuation of travel time
variability due to incidents on freeways. Journal of Transportation and Statistics,
2(2), pp. 123-131.

City Hall Kuala Lumpur (CHKL) (2010). Integrated Transport Information System
(ITIS), Transport Management Centre, Bukit Jalil, Kuala Lumpur.

Department of Irrigation and Drainage Malaysia (MMD) (2010). On-Line Rainfall

Data. [online] Available at:
< >
[Assessed 1 August 2010].

Li, R., (2004). Examining travel time variability using AVI data . Technical report,
Institute of Transport Studies, The Australian key Centre in Transport
Management, Monash University. < >
[Accessed 1st Auguest 2010].

MacDonald, M. (2008). Estimation of DTDV functions for motorways . Technical

report, Department for Transport, Great Minster House, 76 Marsham Street,
London, United Kingdom. < > [Assessed 1
August 2010].

Ng, M., & Waller, S. T. (2010). A computationally efficient methodology to

characterize travel time reliability using the fast Fourier transform. Transportation
Research Part B , 44(10), pp. 1202-1219.

Peer, S., Koopmans, C., & Verhoef, E. (2009). Predicting Travel Time Variability for
<http :// www. internationaltransportforum .org/Proceed ings/rel iabil ity/Peer. pdf>
[Assessed 1st August 2010].

Van Lint, J. W. C., van Zuylen, H. J., & Tu H. (2008). Travel time unreliability on
freeways: Why measures based on variance tell only half the story.
Transportation Research Part A , 42 (1), pp. 258-277.

009 YES WE CAN !

Natasa Gojkovic Bukvic 1,2

Management Consultancy, Bari, Italy

MC, Sarajevo, Bosnia and Herzegovina

ABSTRACT : The project starts finding an adequate transport and development of

infrastructure changing policy instrument reducing congestion, making roads safety and
reducing environmental impacts. In the case of implementation of a new corridor,
canalizing transport from EU through short sea shipping and block trains and economic
growth and growth in freight transport, the solution is not in reduction of transport but in
redistribution between modes. The final goal is to develop a common market in the sphere
of transport services reducing greenhouse gas emission and to develop a common transport
policy for south-east Europe. The main goal of the project is to give an opportunity to
demonstrate an increasment of the demand for intermodal/combined logistic chain using
short sea shipping and railways finalizing to a two main objectives: reduce road congestion
and reduce of CO2 emission. Finally, implementation of legal regulations under supervision
could produce different positive consequences on transport, environment, crime control,
protection and defense.

Keywords : EU common transport policy, Balkans, intermodal logistic chain, reduce road
congestion, CO2 reduction

Internalization of external transport costs are seen as an important instrument of
stimulation of intermodal transport observed in the White Paper of the European
Commission (COM 2001). European common transport policy consists in reforming
national policies of member states with an aim to shaping a new European Union
common policy. The goal of the process is to develop a common market in the field
of transport services and to do so, may be a solution should be to develop a
common transport policy for south-east Europe and to the declared aims which
could contribute economic growth, stability and cohesion in this part of Europe. The
EU strategies for the region have a social dimension such as the EU Roadmap on
enlargement for the Balkans (Tilling 2007). “European Commission has prepared
the following steps:

1 .The selection of a limited number transport infrastructure projects of regional

interest (selection of main road and rail axes, selection of seaports etc.)
(Memorandum of Understanding) with a subsequent aim of channelling investment
towards the selected projects.

2.Liberalization of transport operation services and opening access to

Creation of the demand for intermodal logistic chain using short sea shipping and
railways would underline two main objectives such as reduce road congestion and
reduce of CO2 emission. The realization of both objectives consequently will reach a
final aim of increasing profit and give a better contribution on the environment by
reduction of greenhouse gas emission.

1.1 Brief history of a common transport policy for the European Union
A common transport policy starts from 1999 when the Stability Pact for SE
Europe was set up by European Commission. EU today is a promoter of intra-
regional cooperation, with special interest in transport section.

The main objectives for the road transport are to improve quality, apply existing
regulations more effectively by tightening up controls and penalties. The World Bank
document “Railway Reform in the western Balkans (2005)” contains a list of
recommended railway reform measures to be implemented by each western Balkan
country such as staff reductions, privatization of freight operator and closing loss
making local lines .

Objectives for the sea and inland waterway transport are to develop the
infrastructure, simplify the regulatory framework by creating one-stop offices and
integrate the social legislation in order to build veritable “motorways of the sea”.
Proposed measures are a key part of intermodality, which allow a way round
bottlenecks between France and Spain in the Pyrenees or between Italy and the
rest of Europe in the Alps, as well as between France and the United Kingdom and
looking ahead between Germany and Poland. The Commission has proposed a new
legislative framework for the ports which is designed to lay down new, clearer rules
on pilotage, cargo-handing, stevedoring etc; to simplify the rules governing
operation of ports themselves and bring together all the links in the logistics chain
(Pilsoo 2003).

Objectives for the intermodality are to shift the balance between modes of transport
by means of a proactive policy to promote intermodality/combined transport by rail,
sea and inland waterway. In these connection major initiatives is the “Marco Polo”
community support program to replace the current program. Proposed measure is to
open all appropriate proposals to shift freight from road to other more

environmentally friendly modes. The aim is to turn intermodality into a competitive,
economically viable reality, particularly by promoting motorways of the sea (COM

1.2 What does European Union plan for the region?

The Plan of the European Union has hard and soft measures. The hard
measures are related to infrastructures and soft measures are harmonization and
reforms (technical standards and border crossing procedures). The soft projects
consider rail and ports , which both are ably affected by “ regionalization” and so
Intergovernmental Working Group on Railway and Intermodal Policy were set up.
One of the main job of the Working Group is to make an inventory of rail reforms
and further recommend measures that ensure the regional integration and
harmonization of the reforms for every country and to open access to transport
infrastructure. Governments have usually denied railways enterprises the freedom of
a commercial business. This must change. Some railways may focus entirely on
their core business of operating trains while others may choose to enter into
partnership for example with road haulers or logistics companies and offer door to
door intermodal services. Some may operate across Europe, while others may
concentrate on local services. What is in common for all railways in Region? The
simplest answer should be that they must focus on customers necessity and the
best way how to satisfy these needs. The imperative is to establish common traffic
management whom focus would be on planning, monitoring, control and/or influence
of traffic. As well reported in The World Bank document (2005) where the aim has
been discussed in order to maximize the effectiveness of the use of existing
infrastructure, ensure reliable and safe operation of transport, address
environmental goals and ensure fair allocation of infrastructure space (road space,
rail slots etc.) among competing users. Regarding seaports intention is to identify
which regional port (out of total number of seven ports in the regional core network)
provides the best long-term solution for Adriatic shipping.

“REBIS” is a network and a study that includes the mail rail and road connections
between the five capitals of the region, as well as the cities of Banja Luka,
Podgorica and Pristina; the linking of these cities with the capitals of the neighboring
count ries and t o the ports of the Adr iatic Sea and t he river Danube

1.3 Social impact
The European Union found out that transport has a determinant impact on
regional development and regional cohesion. South-East Europe countries are
more sensitive because facing more or less critical levels of socio-economic,
stability and having a rich history of ethno-political problems.

Most countries in the region have very high (30%) rate of unemployment reached
also as a result for transiting, the Balkan conflicts etc. For example: a sector as
railways transport where 50% of labor force has left their jobs within last ten years.
The current plans to liberalize rail transport in the region will have a further impact
on jobs – and with no chance of resorting to voluntary departures or early retirement

Regarding seaports, the intention is to select a few ports of regional interest for
future investments. This is another critical impact on port-dependent communities
and domestic economies. Here are some key questions such as:

What are the job opportunities for the people who leave the transport industry?
What are the measures to retain the active population in the labor market-training
and job opportunities?

The trade unions should be involved as a social partners in the process of reform
and some other aspects must be pointed out: South-East Europe countries have
young democracies and so less transparency and social dialogue. Assistance of
European Union on South-East Europe countries also as a pressure to implement
social dialogue, consultation and information mechanisms. International financial
institutions – often a barrier for the reform process because of “chain of blame”
social dialogue at national level Consultation of trade union on the European Union
regional transport policy for south-east Europe (De Placio, 2005).

1.4 Environmental issues aspect

The following research targets in the areas of environment, energy and
resources has been identified by ERTRAC (The European Road Transport
Research Advisory Council, 2010):

-Improvements in vehicle efficiency should deliver as much as a 40% reduction in

CO2 emissions for cars and 10% for heavy vehicles for the new vehicle fleet in 2020

- Fuel consumption and CO2 emissions should fall by at least 10% for cars and 5%
for heavy vehicles as a result of better vehicle maintenance and driving for fuel
- Further reductions in fuel consumption of 10-20% should result from improvements
to road infrastructure, better use of transport modes, IT systems, higher car
occupancy rates and freight loading factors
- Further reduction of carbon emissions associated with fuel production should be

- By 2020, fuel cell vehicles and low carbon or hydrogen fuels should start
contributing to carbon reduction, provided sustained research efforts are begun
now - By 2020, Euro 5 and 6 vehicles should be well established in the vehicle fleet

- transport noise should be reduced by up to 10 dB(A) through a system approach

including better indicators and improvements to vehicle and infrastructure (Schade
- Sustainable use of resources and recycling of vehicles and road infrastructure
materials should contribute to the preservations of the environment (ERT 2010).

Research has been conducted and there are some measures which have
environmental impact of transports: regulations on the use of vehicles and the
effective enforcement of those regulations; demand management approaches which
reduce the need to travel or the use of less sustainable modes: pricing measures;
soft options such as travel plans and informational and behavior change campaigns
and land use planning (EEA 2003).

Research on climate change for the United Kingdom Government From the Tyndall
Center for Climate Change Research (Bows at al. 2006; Banister et al. 2006) has
shown that limiting carbon emissions from transport in order to achieve sustainability
targets will be extremely difficult to achieve. These environmental aspects of
transport cover the full life cycle of transport.(Banister et al. 2000). The largest
impacts come from transport use, but the effects from development and construction
of infrastructure and vehicles, as well as the waste from the disposal, and to the
environmental costs of transport.

The Transport and Environment Reporting Mechanism report for 2008 from the
European Environment Agency , concludes that “ the current economic turmoil may
lessen the demand for transport, but the transport sector still contributes significantly

to rising emissions of greenhouse gases, noise exposure, air pollution,
fragmentation of habitat and impacts on wildlife.

There is growing awareness of the transport sector’s disproportionate impact on the

environment, the report shows that there is little evidence of improve performance of
a shift to sustainable transport across Europe” (EEA 2004; EEA 2009). The
increasing volume of transport is challenging the EU transport policy of decoupling
transport usage from economic growth.

Growing transport volumes have driven emissions up by 27% between 1990 and
2006 (excluding the international aviation and marine sector), so transport issue is
moving up into many political agenda and more research in this area is being
conducted. The world’s total energy consumption is expected to increase at an
average annual rate of 1.7% to 2.0% (EEA 2005; EEA 2006). The transport sector
will represent 63% of the increase in global oil demand over the period 2004 to 2030
and in non-OECD countries transport will be the biggest contributor to oil demand
growth. (Transplus).

The principal topics could be recognized in environmental benefits which are:

decrease CO 2 emission,
avoid accidents and
reduce traffic congestion
The main goal of the project is to give an opportunity to demonstrate an increasment
of the demand for intermodal logistic chain using short sea shipping and railways
finalizing to a two main objectives: reduce road congestion and reduce of CO 2
emission. Connecting Southern part of Europe using intermodal transport chain
(short sea shipping (Ro/Ro) and railways and in that case eliminate road transport
means directly reduction of road traffic and of course enormous reduction of CO2.
Kreutzberger et al. (2003) reported accidents, noise, air pollution, climate change
and the most important external costs of transport. The external costs
of long distance road haulage are twice as high as those of rail haulage, and 5 to 6
times that of barge and short-sea shipping. The largest external costs of road
transport are local emissions (33%), congestion (23%) and accidents (22%) The
largest one of rail transport are local emission (31%), noise (28%) and infrastructure
(23%). On the base of what has been demonstrated above congestion and
accidents costs are much more than 50% of external costs in road transport and

those could be easily improved switching in a rail transport where that is possible or
still better switching into a intermodal logistic chain.

The new modal shifted route, as recently showed by researches of European

Environmental Agency that emissions of 51% of nitrogen oxide, 34% of volatile
organic compounds and 65% of carbon monoxide are imputable to road traffic,
could decrease all the above mentioned elements and help in global environment
situation of South-East Europe. The most part of motors pollute are caused by the
diesels engine which are equipped mostly in used in commercial vehicles (EEA
2007; EEA 2008). Improving this multimodal transport chain it also give social
benefits linked to the road safety. In fact, the foreseen reduction of the freight
transport by road, especially on long distance transport (with high safety risks) and it
implies a consistent reduction of the driving kilometres The quantitative
environmental and social benefits have been calculated with the comparison of the
difference between the relevant external costs for the old route entirely done by
trucks and the new intermodal road short sea shipping and railways .

The project idea is to create unaccompanied combined transport chain of
intermodal transport units in South East Europe between Bari Logistic Center and
Logistic Railways Terminals in Bosnia and Herzegovina, Serbia, Romania,
Montenegro, Croatia and Bulgaria avoiding the road traffic and reduction of CO2
using short sea shipping by Ro/Ro vessels and block trains. The European
Commission has developed policy measures to shift the balance between transport
modes with special focus on promoting intermodal transport. The type of transport
has been strongly advocated because of environmental concerns and safety
reasons to avoid road congestions.

The thing to do is to organize railways practice in all Balkan Peninsula countries

mixing private and public consortium which could be able to move merchandise
from/ to Southern Europe to/from Eastern Europe. To start up a common railways
practice it is necessary to create a Intergovernmental Working Group on Railways-
new railway management model able to take care of the opportunities given for
intermodal transport sector - which will include all countries interested in a project
start up. The aim of EU policy has been to reduce and in the future to eliminate
technical and operational differences among national railway systems and achieve

harmonization in terms of technical specifications for infrastructure, signaling,
telecommunications and rolling stock as well as certain operational rules (CEC
2001; CEC 2006) This group should create common intermodal policy.

It is necessary to recognize that the environmental and social implications of the
transport need to be constantly and carefully monitored.

The starting point is to find sustainable transport and welcome the development of
infrastructure changing as a policy instrument to contain and reduce congestion and
reduce environmental impacts. Kreutzberger (Kreuntzberger et al 2003) state that
the environmental performance of intermodal transport is substantially better than
that of unimodal road transport when looking at every use and CO2 emission and
this is even more outspoken when also local emissions, accidents, congestion and
noise are integrated. As regards of the automatic link between economic growth and
growth in freight transport, the solution is not in reduction of transport but in
redistribution between modes. This is a reason why a project idea could have
success. In this case we are not only talking about redistribution between modes
(Ogorelc 2003) of transport but also implementing a new corridor. Fair and efficient
pricing, better investments and some of policy tools that can help to achieve this.
Enlargement of the European Union is set to trigger larger exchanges of goods and
so need for additional investments in transport infrastructures. It’s well known south-
east Europe transport system distinguishes itself by extremely fragmented transport,
i.e. 5000 km of border. Italy, especially South Italy, with its geographical position and
also cultural, political, humanitarian, historical connections could have prestige and
favorable role between European Union and Balkans. Furthermore, implementation
of legal regulations under supervision could produce different positive
consequences on transport, environment, production, crime control, protection and
defense etc. The European Commission is supporting the idea that transport costs
should reflect the true impact on environment and society and is relentlessly pushing
towards the so called internalization of external costs as a policy instrument in order
to establish fair and efficient pricing of different transport modes.

Banister.D and Hickman.R (2006) “How to design a more sustainable and fairer built
environment ” IEEE Proceedings of the Intelligent Transport System 153(4). 276-291

Banister.D , Stead.D, Steen.P, Akerman.J, Dreborg.K, Nijkamp.P and Scheicher-
Tappeser.R (2000) “European Transport Policy and sustainable mobility ”; London
E&FN Spon

Bows.A, Anderson K. and Upham P. (2006) “Contraction & Convergence:UK carbon

emissions and the implications for UK air traffic ” Tyndall Centre Technical Report 40

Commission of the European Communities (CEC) (2001) , “White paper: European

transport policy for 2010: time to decide ” Luxemburg: Office for Official Publications
of the European Communities

Commission of the European Communities (CEC)( (2006) “Keep Europe Moving-

Sustainable Mobility for Our Continent , Mid-term review of the European
Commission’s 2001 Transport White Paper” Luxemburg; Office for Offical
Publications of the European Communities

Commission of the European Communities (CEC)(2007a) “Towards a new culture

for urban mobility ”: DG TREN , Brussels

COM (2001) White paper:European transport policy for 2010:time to decide. 370

De Placio (2005) See the 2005 report of the High level Group co-ordinated by
Loyola de Placio “Networks for peace and development-extension of the major
trans- European transport axes to the neighbouring countries and regions , on the
basis of which the European Commission will issue a Communication in autumn

European Environmental Agency (EEA) (2003) Term report indicators ,

Luxembourg: Office for Official Publications of the European Communities

Environmental Agency (EEA) (2004) Transport and Environment in Europe ,

Luxembourg: Office for Official Publications of the European Communities

European Environmental Agency (EEA) (2005) Core set of Indicators , Luxembourg:

Office for Official Publications of the European Communities

European Environmental Agency (EEA) (2006) Transport and Environment:

facing a dilema, TERM 2005 Indicators tracking transport en environment in
European Union. Report 03, European Environment Agency, Copenhagen (May
2009). report 2006 3/en/eea report 1 2005.pdf

European Environmental Agency (EEA) (2007) Transport and Environment: on the

way to a new common transport policy , TERM 2006 :Indicators tracking transport
and environment in the European Environment Agency, Copenhagen (May 2009) report 2007 1/en/term 2007.pdf

European Environmental Agency (EEA) (2008), Climate for a transport change .

TERM 2007: indicators tracking transport and environment in the European Union.
Report 01, European Environment Agency, Copenhagen (may 2009) report 2008 1/en/EEA report 1 2008 TERM pdf

European Environmental Agency (EEA) (2009), Transport at a crossroads . TERM

2009: indicators tracking transport and environment in the European Union.

Report 03, European Environment Agency, Copenhagen (June 2009)

ETRAC (2010), “European road transport 2020:a vision and strategic research
agenda ”, Sixth Framework Programme Research Advisory Council.

Kreutzberger, E., Macharis, C. ,Vereecken, L. ,Woxen ius, L., 2003 Is intermodal freight
transport more environmental friendly than all-road freight transport ? A review
in:7th Nectar Conference “A New Millennium. Are things the Same? Umea
Sweden 13-14 June 2003

Memorandum of Understanding on the development of the South East Europe Core

Regional Transport Network

Memorandum on European Transport policy

Ogorelc A. (2003)“ European Union Common Transport Policy “; Nase more” 50. 5-6

Pilsoo Jung (2003) ” Annex III EU Common Transport Policy : Trans-European

Networks 67-74.

Schade W.(2003) “Noise :a challenge for sustainable mobility” UNESCO

International Social Science Journal 179, p279-294

Tilling C. (2007)“The EU common transport policy for south-east Europe-what

makes it a factor of cohesion and sustainability? “ , Central and Eastern European
Online Library Transport & Environment statement on publication of EU Common
Transport Policy

The World Bank document “Railway Reform in the western Balkans (Decembe2005)”

TRANSPLUS, “Transport Planning, Land Use and Sustainability ” Fifth Framework

Programme Research Project
al.pdf (01/02/2010)


J. Ben-Edigbe 1, A. Pakshir 2
1, 2 Universiti Teknologi Malaysia, 81300 Skudai, Johor, MALAYSIA
Corresponding Author email:

ABSTRACT: The estimation of future traffic flows in Skudai town hinges on trip distribution
as well as trip generation, modal choice and network assignment. Growth factors and
synthetic methods are often employed in the matrix of elements in inter-zone and intra-zone
movements. Growth factor methods assume that future trip-makings will remain substantially
same even though the volume of trips may increase in line with prevalent growth rates.
However, synthetic models argue that socio-economic factor as well as travel cost would
affect future trip makings. Given the disparity, the study carried out in Skudai Town Malaysia,
explored the hypothesis that future transport network will not have travel resistance as the
present network as suggested in growth factor methods. In essence, year 2009 home to
work trips produced by the residents of Skudai Towns were evaluated and distributed into
zones accordingly, then future trips for 2014 were approximated using growth factor and
gravity (synthetic) models. Trip matrices results were analysed and compared. In the growth
factor method, random irregularities in the observed travel pattern are enlarged in the
predicted pattern. In the gravity model, the common impendence function for Skudai Town is
unsatisfactory even though the model takes into account competition for trips. The study
concluded that notwithstanding its sensitivity to changes in the separation between zones,
gravity model is best suited for responsive trip distribution, it can be argued.

Keywords : trip distribution, growth factor, gravity model, friction factors

Travel forecasting models are used to predict changes in travel patterns and the
utilization of the transportation system in response to changes in regional
development, demographics, and transportation supply. Modelling travel
demand is a challenging task, because Different modeling technique often
results in different outcomes, so it is important to assess the merits and demerits
of the two well established modeling techniques for trip distribution as the
second steps from four steps of transportation planning process (average growth
factor and gravity model) and one that is required for rational planning and
evaluation of transportation systems.

Trip distribution is an iterative procedure used in estimating the number of

trips per unit time between a pair of zones. Growth factor and gravity model
techniques are useful in sequential forecasting of travel demand on any road
network. However, the assumption that transport network of the future will have
the same travel resistance as the existing network is a basic weakness of growth
factor methods. The gravity model on the other hand distributes trips from the
generating zone to the attraction zone. Growth factor methods assume that
future trip-makings will remain substantially same even though the volume of

trips may increase in line with prevalent growth rates. However, synthetic models
argue that socio-economic factor as well as travel cost would affect future trip
makings. Both have commonalities in terms of independent parameters,
however, growth factors relies on historic growth rate for prediction while gravity
model relies on socio-economic variables or friction factor. Whilst the
advantages and disadvantages of these methods have been discussed by many
scholars, it is unclear which approach would assess future trips in a growing
town that is experiencing substantial changes in land use patterns.

Given the disparity between the growth factor and gravity model methods,
the study carried out in Skudai Town Malaysia, explored the hypothesis that
future transport network will not have travel resistance as the present network as
suggested in growth factor methods. Skudai Town, Johor is a rapidly expanding
suburb of Johor Bahru, the capital of the state of Johor in Malaysia.

To Kuala Lumpur

To Kluang
To Batu Pahat
To K

Bukit Batu To Kota Tinggi To Mersing

Ayer Bemban Sengkang

Kelapa Sawit

Senai Ulu Tiram
To Pontian


To Pontian

Johor Bahru
Gelang Patah Pasir Gudang


Tg. Pelepas

To Singapore

Figure 1: Location of Skudai Town in Iskandar Malaysia Region


Skudai Town shown above in figure 1 is included in the new development of

Iskandar Malaysia. Comprehensive Development Plan (CDP) of

Malaysia has been identified Skudai Town and Senai as a logistics hub for the
future development growth. The accessibility to these areas is important to the
mobility of people and transportation of goods and also for future expansions of
urban communities. Skudai town has forty local districts. The total land area is
18,957 acres with a population of is 40,566 (2000), assuming a generalised
growth rate of 3% over 10 years along national line, Skudai population for the
year 2010 is estimated at 55000. It is home to the Universiti Teknologi Malaysia


Trip distribution models strive to predict the number of trips that will be made
between a pair of zones and describe the destination-choice phase of the
sequential demand analysis procedure. Synthetic and growth factor methods are
basic techniques used to determine trip distribution. The more general trip
distribution model can be expressed as:

Where; =Trip from zone i – j; =Production; =Attraction;
= impedance function
Production and attraction are the basic commonality factor in the various models
of trip distribution, model interpretation of impedance function differ in their
characterisation and the way they are assumed to affect trip distribution.

2.1 Applications of Growth Factor Models

Application of growth factors is dependent largely on the accurate calculation
of the growth rate itself and by default makes the model susceptible to gross
inaccuracy. Even then, it is however, the lack of any measure of travel
impedance that makes the growth factor unreliable. Growth factor methods
include constant and average factors, Fratar and Furness methods. As
expected constant factor method assumes that all zones will increase
uniformly and traffic pattern will remain same for the future. The average
factor takes into account varying rates of growth of trip making which can be
expected in different zones. Fratar method assumes that existing trips will
increase in proportion to growth rate, so that attractions of flows to zone are
first balanced and then production whereas in the Furness method,

productions of flows from zone are first balanced and then attraction. The
relative growth factor rate can be estimated as follows:

Percentage of Relative Growth Index zone = [(difference in total growth in

zone base year-projected year / difference in total growth in town zone base
year-projected year)] x 100

Constant Factor, (2)

Average Factor (3)

Fratar, (4)




= future trips between zone i and j ;

= present trips = constant;

2.2 Application of Gravity Model

The use of gravity model allows for inclusion of travel constrains and
transport strategies to be include in the trip distribution matrices. In the
gravity technique, the number of travel demand between two zones
represented on the attraction in the area that this is depended on the
dimension of attractor and vice versa in the spatial separation between the
areas. Gravity model establish upon from four independent parameters. The
below formula use to estimate trip distribution based on the gravity model.













( )












from four parameters, friction factor (fij) and socio-economic factor ( Kij) are
others parameters. To make predictions the forecast zone trip end totals Pi ,

Aj, Fij
factors are input into the relevant models and the appropriate
iterative procedure to satisfy the trip end constraints carried out. Note that
travel patterns are determined by the calibration of gravity model that is
directly related to the friction factor and socio-economic factor.

Friction factor is the function of and inversely proportional to travel

time, it can be argued. Now if it is assumed that the observed distribution tij
a randomly chosen sample of size N from a multinomial distribution, where N
is the total number of trips observed, then the sum of the estimated trip
frequencies must equal the sample size N. Thus, the likelihood of the
observed distribution is given by:




Where, is a Lagrangian multiplier

is to be maximised
A = sum of terms not containing Tij
Assuming a Poisson distribution for tij;

The likelihood of observed matrix;



In any case, gravity model is based on the trial-and-adjustment process to

get the best result. It is usual to assume that if the zone totals of trip
attractions and/or generations are known, they will be used as the
appropriate constrained model.

As mentioned earlier it is common practice nowadays to convert

production constrained model to doubly constrained models by sharing out
the total trips generated between the attraction zones in proportion to their
relative attraction. The measure of attraction used in a model depends on
the trip purpose for which the model is to be used.

Several different measures of separation between zones have been and

are being used. Early studies used the shortest distance with an allowance
being made for any physical barriers, or the shortest road distance.

Most studies in the past have used travel time as the measure of
separation of ten with some questions unresolved. Travel time depends on
traffic conditions and modes of transport and so it is not a unique figure for
any pair of zones. If trip frequencies for different modes of transport are
calculated separately, different travel times may be used in the gravity

In any case, since, gravity model is driven by attraction rates; the choice
of Furness Method is ruled out because it is driven by production rates,
hence Fratar method is the preferred method for comparative assessment
because it is also driven by attraction rates. In sum equations 4 and 6 were
used to assess trip distribution in Skudai Town.


While it is of great interest to give a detailed study setting, a more pragmatic

approach with relevant information will suffice, but not without mentioning that
external zones are often a function of national zoning system. A zoning system
is used to aggregate the individual premises into manageable chunks for
modeling purposes. So, for the purpose of the study, the cordon lines as shown
below in Figure 2 include the land area within the boundary of Jalan Skudai,
Second-Link Highway, and Pasir Gudang Highway.
Figure 2: Zoning Locations in Skudai Town

Further, Skudai town was divided into 6 zones which represent one group
per zone. Particularly Zone 1 is Universiti Teknologi Malaysia. Zone 2 consist of
Taman Jaya,Taman Nesa, Taman Sri Putri, and Taman Sri Skudai. Zone 3
consists of Taman Bukit Gemilang, Taman Damai Jaya, Taman Harmoni 1,
Taman Harmoni 2, Taman Sutera Utama, Taman Ungku Tun Aminah, Taman
Seri Orkid and Taman Mutiara Rini.

Zone 4 is the largest area including Taman Seri Orkid, Taman Dato’ Yunus
Sulaiman, Taman Lima Kedai, Bandar Selesa Jaya, Taman Jaya Mas, Taman
Melawati, Taman Nusa Bestari 2, Taman Nusa Bestari Jaya, Taman Nusa Jaya
Mas, Taman Timur, and Taman Industri Jaya.

While for zone 5 also including Taman Mutiara Rini, Taman Berjaya, Taman
Pulai Bayu, Kampung Baru Skudai, Taman Desa Skudai, Taman Desa Skudai
Fasa 3, Taman Pulai Flora, Taman Pulai Utama, Taman Skudai, Taman Skudai
Baru, Taman Skudai Indah, Taman Skudai Jaya, Taman Skudai Ria, and Taman
Universiti. Finally is zone 6 which consist of Taman Teratai, Taman Sri Pulai and
Taman Sri Pulai Perdana. A summary of estimated zone population and
employment for the 2010 is shown below in Tables 1 and 2. In circumstances
where a zone is described as 4, 4a and 4b say; it’s merely indicative of the
multiple roles of the zone. There are no exclusively residential zones in Skudai

Table 1: Zone Population and Employment in Year 2010

Zone Ultimate Area
2010 Growth Employment Employment
No. (Ha.)
Population Factor Zone No.
2 396.34 6,309 2.3 1 4,359
3 986.63 15,721 6.2 3a 2,785
4 693.27 11,041 4.3 4a 4,074
5 825.08 13,177 5.1 4b 3,407
6 287.76 4,579 1.6 5a 4,598


Data obtained from local authority Johor Bahru Central Municipal Council (MPJBT)
were examined and verified by way test surveys. Other relevant data were collected
by way of home interview surveys. As shown below in Table 2, travel time from
residential zone to working zone is computed as a function of distance and speed. It
is pertinent to bear in mind that the computed travel times do not take into account
intersection delays as to be expected. Travel times shown in Table 2 are link based.
Estimated trip generation for the year 2009 and their corresponding growth factors
are shown in Table 3 below.

Table 2 Travel Time (Mins

Zone 1 3a 4a 4b 5a
2 6 5 12 11 8
3 10 4 8 10.5 8
4 14 7 4 7 10
5 6 8 9.2 9 5
6 6 14 21 15 8

Trip generation data presented in Table 3 and link travel times are analysed
using fratar and gravity methods. By using fratar method the estimated growth
factors are relied on, however, caution should be exercised when interpreting
growth factor findings because of the conflicting figures. Historic population and
employment growth trends were used in fusion with literature obtained from
library and government sources. Nonetheless, the averaging techniques used in
deriving the figures make them acceptable with some degree of acceptable
Table 3 Trip Generation in the year 2009 (100
Zone 1 3a 4a 4b 5a ∑Pj Gf
2 3 2 3 2 3 13 2.3
3 25 16 23 19 26 109 6.2
4 8 5 7 6 8 34 4.3
5 16 10 15 12 17 70 5.1
6 5 3 5 4 5 22 1.6
∑A i 57 36 53 43 59 248
Gf 5.03 5.05 5 5.02 5.02

Care should also be taken when using the friction factors presented in Table 4
below, because delay at intersections that should otherwise be were not taken
into account. The table mere assumes that travel time is inversely proportional to
friction factor, this assumption is debatable. Take note that socio-economic
impedance factors that represent the socio-ecomic constraints associated with
trip making was assumed to be same overtime, hence taken as 1.
Table 4 Friction Factors versus Travel Time
TT 4 5 6 7 8 9 10 11 12 14 15 21
FF 26 21 18 15 14 12 11 10 9 8 7 5
Note: TT – travel time; FF = friction factor


Year 2009 home to work trips produced by the residents of Skudai Towns
were evaluated and distributed into zones accordingly, then future trips for
2014 were approximated using growth factor and gravity (synthetic) models.
Trip matrices results were analysed and compared. In the growth factor
method, random irregularities in the observed travel pattern are enlarged in
the predicted pattern. In the gravity model, the common impendence function
for Skudai Town is unsatisfactory even though the model takes into account
competition for trips.

4.1 Analysis Using Fratar Method

Essentially each complete iterative cycle in fratar method consists of two
steps, one of which guarantees satisfaction of the general constraint. The
iterative process commence with row balancing then column balancing.
Consequently, denote the estimate of Tij
obtained in the first part of the n+1 th
iterative cycle by Tij (n+1/2) and that derived in the second iteration by Tij (n+1)

Note that

Hence and

The resultant trip distributions are shown below in table 5. The most
advantage of the growth factor techniques is that to express the numerous
particular travel relationships that exist in any urban area. In addition, the
Fratar method shows defect in territories with land use change by applying
compounded growth rate generously to all zones. Although the model is
simple and easy to apply, the assumption that travel resistance is unchanged
Table 5 Predicted 2014isTD
with time Based on the
questionable. Fratar
Thus, Method
this model(100)
is good for zone snap shot picture
of what’s happening not prediction that would be relied on for town planning
or future land use. The five year forecast presented in table 5 would have to
be compared with results obtained using synthetic method for validity. Given
that Skudai Town is small, with nearly precise information; one would expect
the results from Gravity model to have a close resemblance to that shown
here in Table 5. However, in an urban area like Johor Bahru or Kuala
Lumpur, a large disparity in outcome would occur.
Zone 1 3a 4a 4b 5a

2 7 5 7 4 7
3 155 100 142 118 161
4 35 22 30 26 35
5 82 52 76 61 86
6 8 5 8 6 8

4.2 Gravity Model Method

Travel patterns for Skudai Town predicted for year 2014 are shown below in

Table 8.The friction factor derived for the calibrated model (see table 6) are
assumed to remain unchanged with time, however if Kij
factors are found
necessary they should be altered to take into account changes in the socio-
economic relationships which gave rise to their original use. In the study,
socio-economic factors are taken as 1. The higher the friction factor, the
smaller the travel time and vice versa. In table 7, matrix figures for the first
iteration are presented. After three iterations the differences between the
given and estimated parameter lie within 95 % to 105% range, hence they
are acceptable. The summary of predicted trip distribution for year 2014 is
shown below in Table 8.

Table 6 Skudai Town Friction Factors

Zone 1 3a 4a 4b 5a
2 18 21 9 10 14
3 11 26 14 10 14
4 8 15 26 15 11
5 18 14 12 12 21
18 8 5 7 14

Table 7 Gravity Model for the first iteration

zone 1 3a 4a 4b 5a ∑
2 8.8 6.5 4.1 3.7 7.1 30
3 119.2 178.7 140.1 81.6 156.5 676
4 18.5 21.9 55.4 26.1 26.2 148
5 93.6 46.2 57.6 46.9 112.6 357
6 13.3 3.7 3.4 3.9 10.7 35
∑ Aj1 253.3 257.1 260.5 162.1 312.9
Given Aj 287 182 265 216 296

Table 8 2014 Predicted TD Based on Gravity Model (100

zone 1 3a 4a 4b 5a
2 10 5 4 5 7
3 138 129 146 111 151
4 20 15 55 34 24
5 103 32 57 61 104
6 14 3 3 5 10

Trip distributions in zones 2, 4 and 6 do not differ significantly for the two
estimation techniques employed probably because of small traffic volume
between the production and attraction zone as to be expected in suburban
areas. However, when there is a surge in traffic volume, significant differences
become obvious as clearly demonstrated in zones 3 and 5. Although gravity
model tends to overvalue near trips and undervalue far trips, nevertheless, there
calibrating ingredients contained within allow for considerable manipulation in
other to obtain a good fit to existing conditions.
Based on the hypothesis that future transport network will not have travel
resistance as the present network as suggested in growth factor methods. The
study concluded that notwithstanding its sensitivity to changes in the separation
between zones, gravity model is best suited for responsive trip distribution, it can
be argued. Further, that the wide use of a common impedance function for an
entire suburban area is unsatisfactory because the effect of separation probably
differs for zones having different economic or social characteristics. The use of
adjustment factors is intended to overcome this problem but for most trip
purposes a greater degree of stratification seems to be the best solution

Nicolas J. Garber, Lester A. Hoel (2009) Traffic and Highway Engineering 4th
Edition, Thomson Learning Academic, Ontario Canada
John W. Dickey. (1975) Metropolitan Transportation Planning, 2nd edition. Publisher:
McGraw-Hill Education, North-Holland, New York USA
Martin Rogers, (2003). Highway Engineering Publisher: Wiley-Blackwell | ISBN:
0632059931 1st edition, Maiden Massachusetts USA
U.S. Department of Transportation, (1977). Federal Highway Administration Urban
Mass Transportation Administration
Coombe, D. (1996). “Induced Traffic: What Do Transportation Models Tell Us?”
Kluwer Academic Publishers Printed in the Netherlands
Cochran, W.G (1977). “Sampling Techniques” Third ed. Canada: John Willey & sons
Inc. New York USA
Burton, M. J. (1985). “Introduction to Transportation Planning”3rd Edition
Hutchinson & CO. Ltd London England
Institute of Transportation Engineering (1991) “Trip distribution” 5th edition
Washington D.C. Institute of Transportation Engineering

ZAREDA Abu Bakar 1, ISMAIL Yusof 2, MOHD ERWAN Sanik 3

1, 2, Universiti Tun Hussein Onn Malaysia, Johor, Malaysia


ABSTRACT: The transportation systems are getting advanced every year. It is frequently
observed in a rapidly growing city such as in Malaysia, that high traffic flow and long queues
at intersections occur during peak hour cause by traffic growth year by year. Signalised
intersections are an essential part of a road network, particularly in urban areas where traffic
congestion has always been a major problem. When it is properly timed, the traffic signal
increases the traffic handling capacity of an intersection. This study for the coordination at
signalised intersection considers the traffic flow improvement at the approaches. The
capacities of signalised intersection depend on the timing plan (cycle length, phase split and
offset) and travel time. The study is focused at Jalan Bakau Condong signalised intersection
which located along FT005 Jalan Batu Pahat-Muar, and to determine the Level of Service of
the intersection using Sidra analysis. Result of this study is based on the SIDRA Intersection
software shows the best performance of time- space diagram for one way street progression
by existing phasing sequence and proposed data to build the synchronization of the study
location. Therefore, cycle length of 100 seconds (daily traffic volume) and 140 seconds
(traffic volume based on the traffic growth) is the accurate time for a platoon of vehicle to
move without stopping along the synchronise intersection with different minimum travel time
and average speed according to the peak hours. Other than that, the operating cost and
pollutant emission is much lower for the coordinated cycle than uncoordinated cycle. To
some extent, suggestion can be made to the authority to rectify traffic signal cycle at
selected intersections.

Keywords : Cycle Time, Level of Service, Signalised Intersection

The traffic growth has been recognized as a serious problem in all large
metropolitan areas in the country, with significant effect on the economy, travel
behaviour, land use and a cause of discomfort for millions of motorists. Although
traffic congestion is not a new problem in urban areas, it has been extended to
suburban areas sooner than expected. The number of vehicles in Malaysia has
been seen to have increased tremendously at the rate of 8% annually over the
past few decades without sign of slowing down. Table 1 shows the statistics of
vehicles registered in Malaysia from 1996 to 2007.
Table 1. Statistics of New Vehicles Registration in Malaysia (RTD, 2007)

To achieve this aim, the study was carried out to perform traffic flow
characteristics and LOS analysis at four signalised intersection in the study area,
to analyse travel time to cross through the four intersections, to obtain the
efficient cycle length of traffic light for each intersection from the SIDRA analysis,
to develop time space diagram on one way street for the signal progression for
the purpose of traffic signal coordination.

Traffic signal coordination is a method of timing groups of traffic signals

along a major roadway to provide for a smooth flow of traffic with minimal stops.
The goal of coordination is to get the greatest number of vehicles through a
system a group of coordinated traffic signals with the fewest number of stops.
While it would be ideal if every vehicle entering the system could proceed
through without stopping, this is not possible even in a well-spaced, well-
designed system. Coordinated traffic signals also result in less stop-and-go
traffic. This can reduce driver frustration and stress levels, and may reduce a
driver’s potential to take risks on the road (McShane et al., 1998).

The scope will focus on the objectives of the study of traffic flows based on the
coordination of the signalised intersection. The initial approach will be the
collection of geographical and traffic data of all locations. Preliminary data such
as the number of lanes, distances between intersections, phasing sequence,
signal timing, traffic light cycle time and traffic volumes, and travel time be
collected so as to assist in the planning for further work. Figure 1 through 3
show the equipment used in this study.
Figure 1. CCTV Camera

Figure 2. Stop Watch Figure 3. Road Measuring Wheel

Manually traffic flow counting method will be used to define the traffic flow at
the arterial road and peak hour for this arterial road. Camera video or CCTV was
used to record each vehicles passing through the intersection. Travel time along
several segments from the arterial was estimated using Chasing Car Method. All
the data obtain at the study location will be transfer in SIDRA Intersection
software. The test site consists of four signalized intersections spaced 0.5 km,
0.7 km and 0.8 km for each segment from KM 127-129 at federal route (FT005)
Jalan Batu Pahat- Muar. The layout of Jalan Bakau Condong is as shown in
Figure 4.
Figure 4. Segment length of Jalan Bakau


3.1 Traffic Flow

Jalan Bakau Condong is expected to experience approximately 3.51% (JKR,

2008) of the annual traffic growth for major roads and minor roads even
without the proposed development in the Batu Pahat. Of this total traffic, the
traffic volume increase year by year due to the traffic growth below (Figure
5). The study area is analyzed for existing conditions observed during weekday
which is most critical peak hours.
Figure 5. Traffic Growth for each intersection

Table 2. Level of Service based on Traffic Growth

8000 Jalan Balau/ Jalan Seraya/ Jalan Melunak/ Jalan Cengal/
LOS Jalan Mat Akil Jalan Mustafa Jalan BHP Jalan Rugayah
6000 Intersection Intersection Intersection Intersection
4000 Daily E E D E
2000 Year 1 E E D E
Year 2 E E D E
Year 3 E E D E
Year 4 F Year 1 E Year 2 DYear 3 Year
E4 Year 5

Year 5 Intersection
F 1 IntersectionE2 IntersectionD3Intersection 4 E
Based on the result in Table 2, the worst level of service happened at the
fourth year (2014). It is use to define forced or breakdown flow. This condition
exists wherever the amount of traffic approaching a point exceeds the amounts
which can transverse the point. Further studies will focused on the worst case for
all intersection at the forth year (2014).

3.2 Cycle Time

The best solution for this matter was using the 100 second cycle time because of
reduce level of service of each intersection (Table 3).

Table 3. Level of Service for different cycle

Jalan Balau/ Jalan Seraya/ Jalan Melunak/ Jalan Cengal/
Jalan Mat Akil Jalan Mustafa Jalan BHP Jalan Rugayah
Time (s)
Intersection Intersection Intersection Intersection
100 E D C D
120 E D D E
140 E E D E
160 E E D E
180 E E E E

The best solution for this matter was using the 140 second cycle time
because of reduce level of service of each intersection (Table 4).
Table 4. Level of Service for different cycle at the 4 th year (2014)
Jalan Balau/ Jalan Seraya/ Jalan Melunak/ Jalan Cengal/
Jalan Mat Akil Jalan Mustafa Jalan BHP Jalan Rugayah
Time (s)
Intersection Intersection Intersection Intersection
100 F E C E
120 F E D E
140 E E D E
160 F E D E
180 F E D E

3.3 Delay
In SIDRA intersection software, the output for stop and delay are measures in
term of time. Table 5 presents different time for uncoordinated and coordinated
cycle. The delay for uncoordinated cycle is much higher than coordinated cycle.
Table 5. Delay
Intersection Uncoordinated Coordinated
Uncoordinated Coordinated
cycle (next 4 th cycle (next 4 th
cycle (sec) cycle (sec)
year) (sec) year) (sec)
Jalan Balau/
50.17 39.02 66.82 63.96
Jalan Mat Akil
Jalan Seraya/
28.28 22.18 35.84 34.90
Jalan Mustafa
Jalan Melunak/
14.88 14.01 20.54 25.26
Jalan BHP
Jalan Cengal/
37.96 31.48 50.37 49.51
Jalan Rugayah

3.4 Travel time analysis

According to data in Table 6, it shows that the driver have to spend more of their
time to travel along this area on evening peak hours than the morning peak hour.
It is occurred because the nearby residential committee to the arterial road
would like to go to the shopping complex which is located in the middle in this
area. Besides that, this arterial road is one of their short cut routes to travel to
Muar or Tongkang Pechah which the place that the committee in the Batu Pahat
live mostly.

Table 6. Average travel time for morning and evening peak hour (Batu Pahat/ Muar)
Travel time (s) Travel time (s)
Intersection Segment length (km) (average) (average)
(morning peak hour) (evening peak hour)
0-1 1.2 135.46 143.95
1-2 0.7 74.54 85.65
2-3 0.5 56.63 65.21
3-4 0.8 100.23 104.48

Referring to Table 7, it shown that there must be lower traffic flow in this road
by this time than the way to Muar because the travel time to pass the entire
segment was lower than the travel time to pass the segment in opposite ways. It
is occurred because the committee whose stayed near this road were preferred
to take the others road to go to their office at Batu Pahat or to Kluang.

Table 7. Average travel time for morning and evening peak hours (Muar/ Batu Pahat)
Segment Travel time (s) (average) Travel time (s) (average)
length (km) (Morning peak hour) (evening peak hour)
4-3 0.8 75.12 78.28
3-2 0.5 45.22 50.14
2-1 0.7 65.16 83.53
1-0 1.2 116.25 121.15
3.5 Fuel Consumption and Emission

In estimating the fuel consumption, current cost of fuel in Malaysia presently was
RM1 .85 per litre (currently taken the latest update, 5 November 2010). Using the
fuel consumption and emission value calculated for light and heavy vehicle in
each lane of movement or approach road are calculated by aggregating the
value for the lanes that belong to the movement or approach. The fuel
consumption, fuel and pollutant emissions result for all movement are given in
Table 4.14 through Table 4.17. The light and heavy vehicle mass are assumed
as 1400kg and 11000kg (Akcelik, 2002).

Table 8. Fuel consumption, cost and pollutant emissions result for all approaches (cycle
Operating Carbon Carbon
Fuel NOX
Cost Dioxide Hydrocarbons Monoxide
Intersection (total) Consumption (total)
(total) (total) (kg/h) (total)
(RM/hr) (total) (L/h) (kg/h)
(kg/h) (kg/h)
Jalan Balau/
421.39 236.2 591/1 1.060 42.32 1.265
Jalan Mat Akil
Jalan Seraya/
257.95 153.0 383.0 0.666 27.65 0.840
Jalan Mustafa
Jalan Melunak/
196.83 128.8 322.8 0.527 22.90 0.721
Jalan BHP
Jalan Cengal/
357.57 214.1 536.5 0.929 39.31 1.177
Jalan Rugayah
*latest updated for Ringgit Malaysia (RM) currency, 5 November 2010

Table 9. Fuel consumption, cost and pollutant emissions result for all approaches (cycle
time-coordinated) at 4 th
Operating Carbon Carbon
Fuel NOX
Cost Dioxide Hydrocarbons Monoxide
Intersection (total) Consumption (total)
(total) (total) (kg/h) (total)
(RM/hr) (total) (L/h) (kg/h)
(kg/h) (kg/h)
Jalan Balau/
357.57 214.1 536.5 0.929 39.31 1.177
Jalan Mat Akil
Jalan Seraya/
350.50 190.2 476.3 0.847 32.73 0.995
Jalan Mustafa
Jalan Melunak/
302.76 178.5 447.0 0.762 30.31 0.956
Jalan BHP
Jalan Cengal/
488.69 265.6 665.5 1.181 46.17 1.384
Jalan Rugayah
*latest updated for Ringgit Malaysia (RM) currency, 5 November 2010

The different value shows that the proposed data including the coordination
parameter can save the operating cost including tyre, oil, repair and
maintenance as a factor of the cost of fuel. The operating cost and pollutant
emissions results between the existing data (non coordinate) and proposed data
(coordinate) shows in Table 8 and Table 9. The different amount between
existing data and proposed data of pollutant release in air shows that, the
signalised coordination was a better way to help in decreasing the toxic gas in
the air. Hence, it helps to reduce the green house effect to the environment.

3.6 Analysis in Signal Controller on Site

In coordinated systems, all signals must have the same cycle length. This is
necessary to ensure that the beginning of green occurs at the same time relative
to the green at the upstream and downstream intersections. There are some
exceptions, where a critical intersection has such a high volume that it may
require a double cycle length, for instance, but this done rarely and only when no
other solution is feasible.

3.7 The Time-Space Diagram and Ideal Offset

The time-space diagram is a plot signal indications as a function as a function of

time for two or more signals. The diagram is scaled with respect to distance, so
that one may easily plot vehicle positions as a function of time. The time-space
diagram can be design based on the timing plan (cycle length, phase spilt and
offset) and the travel time. All data were based on collection and observation

All the existing signal controller are designed for an isolated and working on
fixed system operation and the offset will added in the controller. Table 10 and
11 show the offset time which calculated based on the ideal offset formula.
Table 10. Ideal offset for peak hour

Q (veh) L (m) S (km/hr) h (sec) Loss (sec)
1 21 - 33.48 2 2
2 13 700 33.48 2 2 47.27
3 17 500 33.48 2 2 17.76
4 19 800 33.48 2 2 46.02

Table 11. Ideal offset for peak hour for 4 th

Q (veh) L (m) S (km/hr) h (sec) Loss (sec)
1 23 - 33.48 2 2
2 15 700 33.48 2 2 43.27
3 17 500 33.48 2 2 17.76
4 19 800 33.48 2 2 46.02
3.8 Bandwidth Efficiency

For the data obtained, the bandwidth efficiency can be calculated by our own
since all the value needed is obtained. From the time-space diagram above, the
bandwidth efficiency for this area are 15% (bandwidth-15 seconds) for 100
seconds cycle time and 13% (bandwidth-18 seconds) for 140 seconds cycle
time. According to McShane et al. (1998), the bandwidth efficiency of 40% to
50% is considered good. In this study, the bandwidth obtained was half
comparing to the good efficiency. So, we can conclude that this segment area
able to brought half from the full platoon to drive through by this area without any
stopping and delayed.

3.9 Bandwidth Capacity

In considering the bandwidth value is 15 seconds (cycle time-100 seconds) and

10 seconds (140 seconds), the results for bandwidth capacity can be calculated
with the equation above. The result obtained for the bandwidth capacity is 540
veh/hr and 470 veh/h. That means this segment area able to bring more than
hundred vehicles to travel down the streams per hour without stopping and

The proposed data (100 seconds and 140 seconds) for the cycle length was
success time to implement the coordination between the intersections. Based on
the result and analysis, the design of time- space diagram shows a platoon of
vehicle can have a movement from first intersection to the forth intersections
without stopping. This shows that the effect if coordination between the
intersections can safe several time to accelerate at the minimum speed without
having stop and reduce delay at the intersection. Therefore, the benefit of
synchronization can help to improve a very effective at reducing congestion. In
fact, it also can avoid queuing at the approach delays.


Akcelik, R. (2002), aaSIDRA Traffic Model Reference Guide , Akcelik & Associates

Dewan Bandaraya Kuala Lumpur (2003). KL Structural Plan 2020. Online: 5

December 2007,
JKR (2008), Road Traffic Volume Malaysia. Highway Planning Unit Ministry of Work

McShane, W.R., Roess, R. P., and Prassas, E. S (1998). Traffic Engineering , 2nd
Edition, Prentice Hall, Upper Saddle River, New Jersey.

Road Transport Department, Transport Statistics. Online: 26 July 2007

M. Z. Maleki. & M. F. M. Zain.

Department of Architecture

Universiti Kebangsaan Malaysia(UKM)

Address:Maleki. Zadeh. Mohammad. H1 18 Kolje Zaaba UKM bangi 43600 Kuala


ABSTRACT: Sustainability is influenced by various ranges of human activities which offers

multidimensional impact on human’s life aspects and environment. An integrated systematic
view to events can lead us to a better understanding in the interaction between systems that
may work along no across of each other. Mostly cities growth and we plan for their
sustainable transportation such as railway after we feel needs for mobility. Travelling is
essential for daily needs of people that are possible via consumption energy mainly fossil
fuel and widespread pollution. We need to give attention to the future condition of availability
and cost of energy; it is necessary to design a city and plan its transportation system in
integrated view to minimize the use of energy and environmental pollution. How can we
connect origin to destination to decrease car based travelling? How can we plan land use
with relation to transportation points to support sustainable urban environment? The findings
in this paper can support energy saving and reduce Green house gases emission. The data
from world and local case studies was used by author to describe and find out the answer to
the research questions. This paper tries to answer the basic questions of urban designers
and urban transportation planners as a whole. The government including health sectors,
environment sectors, energy sectors, and municipals can use of this research’s finding.
Keywords : Sustainability, Urban Planning and Design, Green house gases, Energy
efficiency, Transportation System

Sustainability is on top important issue of 21th century due of its extension and
centrality to global environment, energy, economic, and security involved. Increase
average of temperature of earth surface by 0.74 ± 0.18ºC during years 1906-2005
made north hemisphere and land warming faster than other area and ocean, and
caused the forest decrease throughout the world, sea level to rise up by 3.1± 0.07
mm from 1993 to 2003. The world which faces with increases in average global
temperature from 1.4 to 5.8°C by the year 2100 and increase of 26% in carbon
dioxide (CO2) concentrations in the atmosphere since the industrial revolution; If
these activities continue at the present manner, this concentration will increase three
times by the year 2100. Indicator shows changes in plants, insects, birds and fish
have been influenced by green house gases and weather conditions. After
increasing world green house gases, societies would think about review in activities
which result into this situation. Rio conference convened with aim of rethinking on
main issues of our only earth. Researches indicate that there is a positive correlation
between increases in atmospheric concentrations of green house gases and a
growth in the earth's average surface temperature (Moniz 2008).

During the last two decades (1984–2004) primary energy has grown by 49% and
CO2 emissions by 43%, in an average annual increase of 2% and 1.8% respectively
(Pérez-Lombard et al. 2008). Urban area complex with its built environment,
transportation and services use up 75% of the world’s energy resources and sent
out the waste material, pollution, green house gases and climate-changing (Al-
Hosany & Elkadi 2002). New trend in built environment is toward paying attention to
sustainability in design, construction, and maintenance. Refinement and energy
saving are about main part of sustainability. This motives for urban compaction,
shorter journeys and walking and cycling, support economic viability, enhance social
sustainability; and encourage social interaction (Carmona et al. 2003). Since 2007,
the US Green Building Council (UGBC) has started to consider neighborhoods as
sustainable communities instead of focusing on individual buildings for conservation
and efficiency movement; this trend includes colleges and universities as well (Clark
Ii & Eisenberg 2008). They are developing criteria everywhere for certifying project
in new Neighborhood program (LEED 2008).

A research on fossil fuel emissions with urban ecosystem indicates that
developmental density coupled with investments in transportation systems has a
large impact on CO2 emissions, and a great potential for decreasing future CO2
emissions with policies that promote efficient urban form. This study demonstrates
there is a relation between urban density and road extension and use of energy. The
finding can be used to complement land use and transportation policy improvements
(Pataki et al. 2009). This means cities can reduce environment impacts by short
journey through compact transportation systems, so today the arguments are well
known for increasing the density of development and creating more compact city
forms, consequently ensuring a mix of uses, the containment of urban ‘sprawl’ and
achieving social and economic diversity and vitality. Effectively, most of the
arguments now in popular had been made decades ago, namely for urban
containment, compact forms, efficient use of land, a mix of building types, and
proximity to facilities, transport and work. Yet, despite the familiarity, there are
something new. The arguments for compact forms and higher density in the 1990
and present decade are promoted in terms of sustainability. Basic factors for
compact urban forms are: convenience, connectivity, and comfort. Sustainable cells
of city require the creation and/or regeneration of clustered, pedestrian, and transit-
oriented communities of variety of sizes (large to small) defined by areas that
conserve green or amenity land. In the way of compact city form, multi level covered
walkway network connect work, home and business places together during 24 hours
accessibility in city with high density (Jenks & Dempsey 2005). Today trend is
looking for residents of pedestrian-friendly, mixed-use, high-density communities,
with short and fast accessibility to public transportation instead of driving several
miles from far suburbs. “One of the best ways to reduce vehicle travel is to build
places where people can do more with less driving,” says lead author Reid Ewing,
associate and research professor at the National Center of Smart Growth at the
University of Maryland in College Park, Md. “Customer is looking for smaller
housing units and mixed-use communities projected in compact living environments”
(Binsacca 2008).

Ecological issues that creates global warming, energy and renewable energy since
long has been cited by government throughout the word; now acts is strongly
needed, specially with related to cities, that are mentioned as main contributor to
sustainability and focal point within which significant solution can be found (Jenks et
al. 1996).


In recent decades, the world has been facing an energy crisis, and energy
resources have became scarcer. With increasing growth in various areas of human
life, the need for holistic approach is increasing day by day. Societies need to
rethink sustainable developments that fulfill present and future generation needs
with rational use of resources.

Efficiency is the highest aspect of sustainability related to competence and

productive utilization of resources without wasting them (Ferriter 2008). The key
consideration in cost and energy efficiency, is location and intensity of used areas
(Chapin & Kaiser 1979). Urban forms provide possibility for energy used in transport
sector. Researches illustrate that urban density is correlated with travelling and per
capita fuel use. Urban development that produces complex travel pattern generates
more fuel demand. Risk valuation on travel needs in given urban forms in planning
horizon framework is an important part of urban policy development investment
planning in urban area. Integration connectivity to improve non-motorized mode of
transportation are important factors in flexibility and adaptability of urban forms. A
new research is needed on travel demand and transport mode adaptability to keep
participation in activities to optimize investment in urban development and re-
development. The essential need of design for today’s new development is to
provide accessibility with significantly less fuel than current fuel use (Krumdieck et
al. 2010).

It is recommended that urbanism and policy makers should start make land use
have acceptably lower impact on environment. They must focus on human scale in
urban land use development. This can support the promoting of transportation that
uses energy resources other than fossil fuel with no more waste than plant’s
restorative ability and will cause less problem for public health and well-being
(Heberle & Opp 2008).


Data which shaped indicators that contribute to accessibility in neighborhood was

prepared by author. These indicators can be split into two groups; the first group
includes street system, mixed land use and density diversity, apartment proportion,
car ownership, motorcycle ownership, and employment; the second group are
average distance to facilities including primary school, health centre, mosque, High
school, police station, green space and park, commercial above 250m2, and
industrial centre in neighborhoods. In the next stage influence of distance to train
station on these indicators was measured by author with SEM WARPPLS software
regression method.

The study area for this research is 155 neighbourhoods of Subang Jaya. The
average land area of the neighborhoods is 75 Hectare, with 0.28 Hectare (minimum)
and 766 Hectare (maximum). Demographic data including residents, workers,
students and gender characteristics and economic conditions (vehicle ownership)
were based on published data from official Malaysia population census of year
(2000). Spatial data including distance and area were obtained from query from land
use and streets map of municipal. Data was gathered from survey and query via
GIS map figure 1.

Figure 1 Map of Subang Jaya study area

Recent data of 211 countries throughout the world was used by author. These data
are 40 dependent and independent indicators including CO 2 emissions, SO 2
emissions, urban SO2 concentration, urbanization, percentage of population living in
urban area and rural area, urban and rural population growth annual, population
density, rural population density of people/km2 of arable land, percent of forest and
agricultural land, oil consumption (bbl/day per 1,000 people), expenditure and
revenues budget, largest city population per capita, areas under protection, Forest
area, agriculture land, and traditional fuel consumption. Variables including CO2
emission, urban SO 2 concentration, and oil consumption were selected as
dependent variables.

The influence of distance to train station on density, diversity, apartment proportion,
car ownership, motorcycle ownership, street system design, and employment was
measured by author figure 2.
Figure 2 Influence and coefficient of train station on factors of sustainable design

The influence of distance to Train station on distance to facilities including primary

school, health centre, mosque, High school, police station, green space and park,
commercial above 250m2, and industrial centre in neighborhoods was measured by
author figure 3.

Figure 3 coefficients between distance from train station and accessibility to facilities
in neighborhoods

An analysis by linear regression on Data from 211 countries shows population

concentration in urban area influence oil consumption negatively; Highways and
2.5 2 1.5 1 0.5 0
railways good transported have negative effect on oil consumption and SO 2
Oil Consumption
-1.5 SO2

goods transported-2increase oil consumption throughout the world

emission. RoadCO2
figure 4. -3


Oil Consumption

Urbanization Highways Railways goods transported Roads goods transported

independent variables
Figure 4 coefficient of urbanization, highways, railways goods transported, and road transported as
independent variables with relation to oil cunsumption and co 2as dependent variable

The analysis by linear regression on Data from 211 countries shows population
density has negative effect on CO2 emission and rural population influence s CO2
emission positively figure 5.

Figure 5 impacts of population density, highways, and rural population on oil

consumption, So 2 and Co 2 emission


By efficient land use planning that reduces trip distance, increases vehicle
efficiency, transport system changes that provides public with low energy
consumption and support active modes, and finally reduce people activities
participation by use of technology and changes in behaviours can reduce energy
use considerably (Krumdieck et al. 2010). Cameron (2004) says model mobility as
the vehicle kilometres travelled (VKT) is a strong function of vehicle ownership
(Krumdieck et al. 2010). The current study shows:
1-Car ownership and motor cycle ownership increase if distance of neighborhood
from train station increase.

2-Density, employment, streets, and apartment proportion shows negative

coefficient with distance from train station. This means that density, employment,
street, and apartment increase in shorter distance to train station and decrease in
farer distance.

3-Distance to facilities including health centre, mosque, High school, and

commercial above 250m 2 increase positively and significantly if distance of
neighbourhood from train station increased.
4-Accessibility distance to police station decreases if distance to train station
increase significantly.
5-Population concentration in urban districts of countries shows oil consumption
reduction significantly.

6-Lengths of highways and railways good transported make oil consumption and
SO 2 decrease and road goods transported has opposite coefficient as well. 7-
Population density generally shows decrease influence on CO 2 but rural
population has positive effect on CO2 emission.

Al-Hosany, N. & Elkadi, H. 2002. Sustainability approaches for incarceration
architecture. Renewable and Sustainable Energy Reviews 6, 5, 457-470.
Binsacca, R. 2008. Builder March availble at: -
cure.aspx?printerfriendly=true , 1-3.
Carmona, M., Health, T., Oc, T. & Tiedell, S. 2003. Public Places – Urban Spaces.
Chapin, F. S. & Kaiser, E. J. 1979. Urban Land Use Planning.
Clark Ii, W. W. & Eisenberg, L. 2008. Agile sustainable communities: On-site
renewable energy generation. Utilities Policy 16, 4, 262-274.
Ferriter, E. (2008). The sustainability of New Urbanism: Case studies in Maryland.
United States -- Delaware, University of Delaware. Ph.D.: 257.
Heberle, L. & Opp, S. M. 2008. Local Sustainable Urban Development in a
Globalized Word. 14, 86,.
Jenks, M., Burton, E. & Williams, K. 1996. The Compact City A Sustainable Urban
Form? . Oxford Bookes University, Oxford UK, 57, 84, 172, 173, 208,.
Jenks, M. & Dempsey, N. 2005. Future Forms and Design for Sustainable Cities.
ISBN 0 7506 6309 X, P24 pp163 -300 p p 113- 226 p 113.
Krumdieck, S., Page, S. & Dantas, A. 2010. Urban form and long-term fuel supply
decline: A method to investigate the peak oil risks to essential activities.
Transportation Research Part A: Policy and Practice 44, 5, 306-322.
LEED 2008. LEED for Neighborhood Development Rating System. 1st available at;,
Moniz, J. E. 2008. Climate Change In Energy Pathways for the Mediterranean. P3.
Pataki, D. E., Emmi, P. C., Forster, C. B., Mills, J. I., Pardyjak, E. R., Peterson, T.
R., Thompson, J. D. & Dudley-Murphy, E. 2009. An integrated approach to
improving fossil fuel emissions scenarios with urban ecosystem studies.
Ecological Complexity 6, 1, 1-14.
Pérez-Lombard, L., Ortiz, J. & Pout, C. 2008. A review on buildings energy
consumption information. Energy and Buildings 40, 3, 394-398.
054 The Assessment of Significant Aspects and Impacts at Highway
Construction towards Sustainable Development

Alea Wahida Ismail, Sumiani Yusoff, M. Rehan Karim,

Dept. Of Civil Eng, University Malaya, Kuala Lumpur. , ,


At dense rural areas, the environmental impacts resulting from highway construction
activities need a proper planning and management to mitigate the impact to the environment.
Awareness or knowledge in minimizing the water, air and noise pollution which effect
surrounding flora, fauna and human beings for transport improvements underlies any
behavioral response. For this study, the Trans Eastern Kedah Interland highway is chosen
which involves the construction of 115km highway that compromises new alignment and
upgrading of existing road. The aim of study is to determine the key potential significant
aspects and impacts of the activities at the highway construction as well as implementing
“good green environmental practice” towards sustainable development. The ISO 14001
Environmental Management System (EMS) is providing a framework for organizations to
systematically review and improve operations for better environmental performance. By
assisting an organization in achieving economic goals, an EMS can also be integrated with
other management functions. An assessment of environmental impacts related to the
highway construction activities was carried out to determine the significant aspects
associated with the activities at highway construction site and their impact to the
environment. In assessing the significance of environmental impacts in this research study
where the technique used is numerical scoring. There will be discussion to establish action
plans and guidelines for mitigation of the environmental impacts at highway construction
towards sustainable development. The results are analyzed quantitatively and presented in
statistical diagram. Minimizing the environmental pollutions are a beginning to adopt new
paradigm and advanced integrated environmental, health and safety (EHS) practices to
support the project team commitment to a better environment.

Keywords : Highway construction, ISO 14001, EMS, significant impact, significant aspects,
sustainable development.


In achieving sustainable development, the highway construction

project has a crucial to play in assisting the efforts of the government where
there is a balance between social environmental protection and economic.
Highway construction activities are environmental nuisance in the form of
noise, dust, muddy runoffs and improper disposal waste. Many highway
construction firms have undertaken measures in facing the growing pressure
of environmental protection to reduce their damage to the environment. The
aim of the research is to determine the key potential significant aspects and

impacts of the activities at the highway construction as well as implementing
“good green environmental practice” towards sustainable development.

Environmental management has become an important worldwide

issue due to the growing concern of environmental problems such as
pollution, ozone depletion, global warming and waste disposal that are
caused by resource consumption as well as environmental emissions and
wastes originating from the construction activities (OECD, 2001).

Over the last decade there has been an increase in involvement of

highway construction activities in environmental issues which is related in a
change of the role of construction from complying with environmental
legislation to actively implementing voluntary EMS that help to manage and
minimize the environmental impact (Ahmad Husseini, 2001).

The site chosen for the research study is the Trans Eastern Kedah
Interland Highway involves the construction of a 115km highway that
compromises new alignment and upgrading of existing road. The project
route traverse through jungle and forest areas and the terrain is generally
mountains. The roadworks include highway construction of flexible pavement
for carriageway and lateritic/ graveled for shoulder.

Figure 1: Slope had been left out without any protection after site clearing

Many environmental problems caused by highway construction site
activities are reported. Although the highway construction activities provide
spaces for human activities and social functions contribute impacts to the
environment. Most highway construction involves a drastic change in both
physical and biological components of the natural environment. Land clearing
results in the exposure of topsoil to erosive rain and earthwork activities
enhance formation of loose soils.

1.1 ISO 14001

Today it is important to identify where these impacts originate from in
order to minimize environmental impacts. The key requirements of an EMS is
that a highway construction project need to go through a process that
identifies and priorities the aspects of the project activities that have the most
significant negative impact. The EMS such as ISO 14001 is globally used
which provides guidance for highway construction to identify and evaluate
environmental aspects of its activities and their effects on the environment.

The ISO 14001 standard is part of the ISO 14000 series that was
developed and released by the International Organization of Standardization
(ISO) to provide organizations worldwide with a common approach to
environmental management (Tibor & Feldman, 1996). An integral part of the
ISO 14001 EMS is through identifying significant environmental aspects and
impacts. Becoming proactive in managing the environmental aspects of a
highway construction project will make a good business sense, as it will bring
benefits to an organization.

1.2 Malaysia and Environmental Management

Malaysia’s commitment is also seen in its involvement as a
participating member of the ISO/TC207, which was formed to manage the
development of the ISO 14000 environmental management system (Cascio,
1996). The government can only do so much unless there is a total
commitment of all Malaysians construction and a change of their mindset
regarding the environment aspects and its impacts. As Malaysian highway
construction activities started to realize the importance and implications of
integrating environmental values into their operations, the need for the ISO
based EMS was slowly gaining popularity.

An environmental aspect is an element of an activity that can interact

with the environment. The environmental aspect of an activity is that part of it
that creates a possibility for an environmental impact such as it is equivalent
to the concept of “hazard” in safety that is also defined as the mere possibility
of a negative event. An environmental aspect that has or can have a
significant environmental or business impact can potentially cause a
significant environmental or operational impact that has applicable regulatory
or other requirements.

An environmental impact is any change to the environment

whether adverse or beneficial wholly or partially resulting from the facility
activities. A potential environmental impact is equivalent to the concept of
“risk” in safety that assigns a probability and consequence to the possible
negative event that may result from a hazard. The highway construction itself
must be sustainable to minimize the pollution causes at the work site.

1.3 Highway construction and Environmental

The highway construction generated impetus to the Malaysian
economy for many years it has created an important role in improving the
quality of life for Malaysians through multiplier effects to other development.
Physical development solely would not give guarantee to the quality of life for
future generations.

Understanding the concept of sustainability and exchanging it into

practice as sustainable development is a key challenge for today’s highway
construction environment professionals. The skills and vision of those who
shape our cities is vital to achieve sustainable solutions to the many
environmental, economic and social problems we faced on a national, local
and global scale.

1.4 Issues and challenges in Malaysia Highway Construction
Developing country like Malaysia, sustainable highway construction
trend tends to focus on the relationship between highway construction and
human development while marginalizing environmental aspects. Environment
encompasses physical and non-physical medium such as air, water, noise
pollution and solid waste.

The environmental issues that happened in Malaysia are due to lack

of environmental consideration in the exploitation, development and
management of resources as well as lack of control of the resulting pollution.
These can be done by education, site planning, management and design
practices to adopt new technologies.

The control of environmental impacts from highway construction has

become a major issue to the public (Shan and Tam, 2002). The promotion of
environmental management and the mission of sustainable development
have resulted in pressure demanding the adoption of proper ways to improve
the environmental performance across highway construction.

Figure 2: Cutting work in progress and air pollution is kept to acceptable limit.

In the awakening of environmental issues, environmental guidelines

and standards are produced with the general aim of helping constructions
manage and improve their environmental performance. Highway
constructions are beginning to develop a new technologies and techniques
that may help to move towards sustainable world (Welford, 2000).

Greater adoption and use of environment-friendly planning techniques
and designs in property projects will go a long way towards promoting green
practices in the country. It will be more effective if highway constructions
voluntarily adopt green and environment friendly designs and concepts in
their projects rather than depend on legislation to make it mandatory for them
to incorporate pro-environment design features in the projects (Stars 23rd
May, 2009).

1.5 Issues and Challenges of Sustainable Highway Construction

Making the highway construction process environmentally friendly and
sustainable is a complex process that needs different inputs and skills or
expertise from different stakeholders. Different actions were taken at different
stages of the life cycle of a planner, civil engineers, surveyors, contractors,
suppliers which has a role to play. The built environment is tremendously
huge in terms of money as well as all other aspects and it represents a prime
driver for development in any country.

Sustainable highway construction is seen as a way for the highway

project to respond towards achieving sustainable development on the various
environmental, socio-economic and cultural facets (CIB, 2002).

• Construction Impacts on highway construction

The impacts of construction on sustainable highway constructions are

important as far as continuing transportation reinforces the importance
of creating a built environment that is sustainable for future
generations. The provision of infrastructure and utilities are major
resources that are used by nations, communities and business.

Figure 3: Warning measures has been put.

• Management and organization

Management and organization is a key aspect of sustainable highway

construction. It is therefore a very complex and difficult subject to
address due to the breadth of their inter relationship and to a major
characteristic of the construction sector which is the strikingly large
number of actors involved in the process of activities. It starts from the
development phase up to the demolition phase through the operation
phase of each component of the built environment.

The barriers to progress are high and the challenges to be tackled

deal with several different aspects such as the design process, the
environmental quality construction, the re-engineering of the highway
process, the development of new highway concepts, the human
resources, the decision-making process, the project owners and
clients demand, education, public awareness, standards and
regulation or research.

Highway issues are concerned on how to optimize the characteristics of

highway construction project in order to improve the sustainability
performance taking into account such as background factors as climate,
culture, highway traditions and stage of construction development. Building
performance environmental methods by increasing the number of parameters
and exploiting suitable indicators will lead to a better assessment of the final
highway construction work.

Figure 4: Water quality sampling by Environmental Engineers.

1.6 Greener Highway Construction

In ensuring environmental is in a good condition during the
construction process, the civil engineers play a leadership role because they
have the rights and obligation in environmental related decision making. In
making a decision on environmental related there are two approaches of civil
engineers which are the ability to think in terms of alternatives and outcomes.

Green highway construction is a way for the project to move towards

achieving sustainable development which takes into account environmental,
socio-economic and cultural issues. Specifically, it involves issue such as
design and management of materials, road performance, energy and
resource consumption within the larger orbit of urban management and

2.1 Assessing Significant Impacts
A scoring method which is self-assessment method was then chosen
to assess significant aspects and impact of the highway construction
activities. The highway construction environmental objectives and targets
were identified and significant were further refined and developed into a
formal set of environmental objectives. Determining where control or
improvement is needed, identifying significant environmental aspects and

associated impacts were necessary. Ascertaining the legal needs and
requirements to which the organization subscribes and setting up the
objectives and targets to achieve them.

2.2 Impact Assessment

The method of numerical scoring system classifies the significance of
the impacts into three ranges: “very significant”, “significant” and “non-
significant”. Significance evaluation involves applying technical analysis and
the use of criteria, which should help an organization to establish which
environmental aspects and associated impacts it considers significant.

Establishing criteria for significance, the following should b considered:

a)Environmental criteria (such as type, size and
frequency of an environmental aspect);

0)Applicable legal requirements;

a)The concerns of internal and external interested
parties (such as noise, air, water and solidwaste).

An organization should maintain appropriate information on the

environmental aspects identified and those considered significant to facilitate
planning. The organization should use this information to understand the
need for and to determine operational controls.

Table 1: Keys criteria used for method are defined below:

Criteria Description

Legislation The Environmental Quality Act (EQA), 1974 (Act 127) and
(Act 672), was set up as the governing environmental
legislation in Malaysia
Local impact The severity of the environmental impact on a local scale
Impact on external The severity of the environmental impact to the community
community within a boundary of 115km from the facility
Regional/ Global Impact The severity of the environmental impact on a regional/
global scale

Policies/ Directives Is the impact governed under any Malaysian or international
policy or directive?
Management Cost The cost managing the impact
Public Satisfaction/ If an impact arises, how does the public respond?
Demand and Competitive
Frequency of occurrence What is the frequency of the occurrence?
Severity if Impact How serious is the impact?

The main score is determined for all keys criteria after which the score is
multiplied together with the score for frequency and seriousness. The scores
are then added up together and the significance of the impact is determined
from the following ranges:
“Very significant” : V is equal or above 65 (V > 65)
“Significant” : V is equal or above 30 but not more than 65 ( 30< V <
“Non-significant” : V is below 30
The calculation for the impact assessment is then carried out using the
frequency and seriousness of the criteria as shown below:

Frequency (F)

Seriousness S)

Frequency x Seriousness

Legal Requirement (LR)

Local Impact (LI)

Regional/ Global Impact (GI)

External Community (EC)

Policies/ Directives (P/D)

Management Costs (MC)

Public Satisfaction/ (CS)

Total T)

Significance level

The formula will be as below:
The total value resulting from the collection of elements T determine the level
of significance and can therefore indicate that the impact was very significant,
significant or non-significant.

Table 2: Significant Impact at Demolition Work
Activities/ Sub- Environmental Environmental Legislation
activities Aspects Impact Requirements

1)Hacking Work Emission to air of Air Pollution Environmental

Particulate Quality (Control
-This work involve matters, Sox, of Emission from
the use of Nox, CO & CO2 Air Pollution Diesel Engines)
machineries- from machineries Regulations,
mechanical breakers exhaust Disturbance 1996
to hack the existing Emission to air of to the existing Environmental
structure dust and structure/ Quality (Clean
particulate Nuisance Air) Regulations,
matters Noise 1987
Pollution Environmental
2)Transferring of Physical Quality (Motor
highway construction attributes- Vehicle Noise)
waste vibration Land use Regulations,
-This work involve Physical
the use of attributes-noise Environmental
excavators & lorry Quality (Motor
tippers to transfer Vehicle Noise)
the highway Generation of Regulations,
construction waste to construction 1987.
dumping ground/ waste Environmental
landfill Quality Act,

Table 3: Significant impacts for Earthwork
Activities/ Sub- Environmental Environmental Legislation
activities Aspects Impact Requirements

1)Excavation Works Emission to air of Air Pollution Environmental

Particulate Quality (Control
-This work involves the matters, Sox, Nox, of Emission from
use of machineries- CO & CO2 from Resource Diesel Engines)
hydraulic excavator, machineries Depletion Regulations,
compactor, back exhaust 1996
pusher, lorry tippers, Resource use- Land Environmental
shovel, motor grader diesel Contamination Quality Act, 1974
and scraper. Environmental
Leakage of Oil Water Quality Act, 1974
2)Movement of heavy onto Land Pollution
vehicles Class II or III of
Leakage of Oil into the Interim
- This work involve the Water National Water
use of excavators & Water Quality
lorry tippers to transfer Pollution Standards
the construction waste (INWQS), DOE
to dumping ground/ Discharge of Water Quality
landfill Surface Runoff Standard, Local
Class II or III of
the Interim
National Water
Water Quality

3.1 Impact Assessment

3.1.1 Environmental Aspect: Dust Emission
Environmental Impact: Air Pollution

Source of impact: In terms of dust during the highway construction


The baseline air quality monitoring was conducted at one monitoring station
established at Kampung Landai. The monitoring of ambient air quality should
be carried out close to sensitive receptors such as rural settlements
especially those with schools or clinics adjacent to the highway alignment. A
total of one air monitoring stations was proposed as the road construction at

these areas was expected to generate dust and affect nearby sensitive
receptors. The air monitoring will only be carried out when construction
begins at these locations. The locations of the proposed air monitoring
station were

PAI : within the road construction

PA2: within the nearby residential area

The air quality parameter to be monitored is Total Suspended Particulate

(TSP) as recommended. The air monitoring shall be carried out on a
quarterly basis with a sampling duration of 24 hours. The monitoring stations
should be relocated and scheduled accordingly as work progresses. The
result of air quality monitoring is as shown in the Table 4.

Table 4 : Baseline Air Quality Sampling Results

Sampling Total Suspended Particulate Malaysian Recommended Air
Location Quality Guidelines
PA1 34 260 (24 hours averaging time)
PA2 28 260 (24 hours averaging time)

Analysis of the dust emission which has been identified as one of the aspects
which effect the air pollution using the scoring method is calculated as below:

Legal Requirement (LR) =4

Local Impact (LI) =2
Regional/ Global Impact (GI) =2
External Community (EC) =2
Policies/ Directives (P/D) =3
Management Costs (MC) =3
Customer Satisfaction/ Demand (CS) =2
Frequency (F) =4
Seriousness (S) =1

T= 4*1 (4+2+2+2+3+3+2)
T= 72

72 > 65 that means it is very significant

3.1.2 Environmental Aspect: Wastewater Discharge
Environmental Impact: Water Pollution

Source of impact: The sewage, sullage and sanitation that comes from
waste, causes water pollution.

Based on the observation, it is proposed that only four streams need to be

observed at the Trans Eastern Kedah Interland Highway Project. The water
samples shall be taken at both the upstream and downstream points of the
crossing. A total of seven sampling stations are proposed during the
earthworks and construction stage where construction activities on site may
have potential impact upon the water quality. The description of the river
water monitoring stations was presented

PWQ1 : Ch. 71920, upstream

PWQ2 : Ch. 71920, downstream
PWQ3 : Ch. 74060, upstream
PWQ4 : Ch. 74060, downstream
PWQ5 : Ch. 74900, upstream
PWQ6 : Ch. 74900, downstream
PWQ7 : Ch. 80050, downstream

From monitoring results there are seven parameters which are dissolved
oxygen (DO), Biological Oxygen Demand (BOD5), Chemical Oxygen
Demand (COD), Ammonia cal Nitrogen (NH3-N), Total Suspended Solids
(TSS), pH, and E-coli. Water samples were obtained by the grab sampling
method and preserved for further analysis in the laboratory.

Table 5 : Baseline Water Quality Analysis Results

Station Temp pH BOD COD TSS Oil & NH3- DO E-coli Turbidity WQI &
Time Gas N Class
PWQ1 28.5 7.2 <2 6 <2 ND(<2) 0.6 5.1 100 <1 91.09
Class II
PWQ2 28.6 7.4 <2 7 4 ND(<2) 0.1 4.9 20 <1 89.03
Class II
PWQ3 28.6 7.1 <2 7 18 ND(<2) 0.4 4.9 500 <1 85.93
Class II
PWQ4 28.7 7.0 <2 8 4 ND(<2) 0.6 4.9 900 <1 89.16
Class II
PWQ5 28.7 7.7 <2 <2 <2 ND(<2) 0.1 5.0 ND <1 90.55
Class II
PWQ6 28.8 8.0 <2 <2 <2 ND(<2) 0.1 4.8 200 <1 89.47
Class II
PWQ7 27.4 7.9 <2 6 12 ND(<2) <0.1 4 2000 16 83.85
DOE - 5.5- <50 <100 <100 <7 - 5- <400 <200
Std. 9.0 7
Source: Environmental Science (M) Sdn. Bhd
•All result are expressed in mg/l except for pH value,
temperature (oC), E-coli (count/1 00ml) and Turbidity (NTU)

The samples were collected from monitoring stations. All results referred to
baseline of highway construction project as shown in table 5. The Water
Quality Index as shown in table was computed and it falls within Class II.
Class II represent water bodies of good quality. Most existing raw water
supply sources come under this category. In practice, no body contact
activity is allowed in the water for the prevention of probable human

Legal Requirement (LR) =4

Local Impact (LI) =3
Regional/ Global Impact (GI) =3
External Community (EC) =4
Policies/ Directives (P/D) =2
Management Costs (MC) =4
Customer Satisfaction/ Demand (CS) =2
Frequency (F) =3
Seriousness (S) =2

T= 3*2 (4+3+3+4+2+4+2)
T= 132

132 > 65 that means it is very significant
*result analysis is still in progression

Malaysia continuously supports initiative in implementing sustainable
highway construction development strategies including green growth.
Highway construction plans in Malaysia have long recognized the need for
green strategy by way of conservation, promoting green network and
sustainable highway transportation. Making a commitment to highway
construction designs will help to support low emission for the public involving
the widest possible range of contractors in green growth with green

Assessing an integrated efficient and user friendly transportation

system including environmental pathway, network for efficient connectivity
and to minimize the level of air pollution, noise pollution and water pollution.
The use of innovative technology in planning and services management is
aimed to promote the highway construction of green roadwork.

Therefore, a greater integration of social, economic and environmental

consideration is need as a sustainable development strategy in planning and
assessing environmental aspects and impacts at highway construction. Long
term public interest should be given more consideration than short-term
private interest as a sustainable development strategy at highway

A method to identify and evaluate the environmental aspects and
impacts at highway construction was developed that started with sustainable
development of the process followed by the identification and evaluation of
environmental aspects and impacts. To evaluate the significance of the
environmental impacts, a specific criterion was chosen which directly
corresponds to its environmental aspect. These criteria included elements
such as environmental perspective and business perspective.

Therefore, significant aspects are the ones that will be the focus of
management efforts to improve the environment. The associated aspect
needs to be managed if an impact is significant. The identification and
evaluation process is important because the aspects and impacts identified a
significant will be used in the development for other elements of the EMS.

Environment management was becoming increasing important to

highway construction as it was considered to be a major polluter to the
environment. A critical element of the EMS involves identifying and
prioritizing the environmental aspects and impacts associated with the
facility. Sustainable environment can be only achieved if development and
environmental issues and problems are given equal emphasis where basic
human needs must fulfill with living standards improved and ecosystems
sustained effectively.

CIB, (2002). Agenda 21 for Sustainable Construction in Developing
Countries. The International Council for Research and Innovation in
Building and Construction (CIB). UNEP-ITEC, Discussion Document.
South Africa.
Ahmad Husseini, (2001). Industrial environmental standards and their
implementation in the world. In : International seminar on environmental
managament 2001 (EM 2001). Kuala Lumpur Malaysia.
OECD, (2001). Encouraging Environmental Management in Industry. Report
on the OECD business and Industry Policy Form on Environmental
Management: Challenges for Industry, 27 September 2000.
Shan, L.Y. and Tam, W.Y. 2002. Implementation of environmental
management in Hong Kong Construction Industry. International Journal of
Project Management. Vol 20, pp.535-543.
Sheldon, C., & Yoxon, M., (1999). Installing environmental systems . UK:
Earthscan Publications Ltd
Sumiani Y., (2001). ‘An Overview of the Environmental Management
Framework for Industries and Life Cycle Assessment’ Proceedings on 4th,
SEA Regional Conference on Higher Engineering Education Network
(RECHEEN 2001) Kuala Lumpur, 15-17 April 2001.
Sumiani Yusoff (2004). “New Approaches for the Sustainability of Malaysian
Industries” in Issue 2/2003 IMPAK, Department of Environment, Ministry
of Science, Technology and the Environment, Malaysia. ISSN: 1394-
Sumiani Yusoff (2004). “Sustainability of ISO 14001 in Malaysian Industries”.
Paper presented on Environmental Management and Certification: Cross
Cultural Studies and Comparisons, Environmental Management Practices
Network Workshop, Roligheden, Copenhagen, 25-26 May, 2004.
Sumian i Y., (2006). “Greening of Malaysian Industries through Implementing
ISO 14001”, Journal of Industrial Technology, Vol. 15(2) (2006).
Starbizweek, The Star (23rd May, 2009).
Tibor, T., & Feldman, I., (1996). ISO 14000- A guide to the new
environmental management standards. USA: Irwin.
Welfrod, R (2000). Corporative environmental management 3: Towards
Sustainable Development . London: Earthscan Publications.





Md. Mahmud Hasan 1, Shamas Bajwa 2, Edmund Horan 3, Edward Chung 4

1, 2, 3 RMIT University, Melbourne, Australia

Queensland University of Technology, Brisbane, Australia

ABSTRACT: This study investigates the effect of rainfall and traffic performance on crash
risks. Road traffic performance can be described in terms of traffic mobility and roadway
safety. Traffic flow and speed are the two important indices used to quantify the traffic
performance while number of accidents is used as a performance index for roadway safety.
Severe weather condition such as rainfall affects the traffic performance and road safety.
During and immediately after the rainfall, the friction between tyres and road surface is
reduced, roads become submerged and visibility is deteriorated. Previous studies have
revealed that traffic condition changes significantly during rainfall and as a consequence,
traffic delay and accidents are increased which result in significant economic costs. For this
study, 93 km long section of The Pacific Motorway in Queensland, Australia is selected.
Hourly rainfall data from nearby weather stations, 5- minute interval traffic flow and speed
data along with crash information for the years 2009 and 2010 are used to analyse the effect
of rainfall and traffic on road accidents. Statistical analysis based modelling approach named
decision tree analysis is applied to develop the relationship among rainfall, traffic
performance and crash risks. The outcomes of the research will aid to improve the roadway
safety by considering traffic movements and severe weather condition.
Keywords: Rainfall, Traffic flow, Vehicle speed, Road accidents, Road safety.


Traffic performance and safety can deteriorate due to different natural and man-
made causes. Among the natural causes, adverse weather condition impedes
normal traffic movements and roadway safety. Adverse weather like rainfall is
considered as one of the hindrance against safe and normal service for the road
traffic. During adverse weather condition, road users may experience reduce
visibility, lesser friction between tyres and the road surface and also submersion of
the roadway. All these affect drivers’ behaviour and traffic movements on a roadway
and as a result, significantly impact on traffic mobility and road safety (Goodwin,
2002) and changes the normal pattern of travel demand and results in mode shift
(Faouzi et al., 2010).

This study investigates the relationship among rainfall, traffic performance and
crash risks. Road traffic performance can be described in terms of traffic mobility

and roadway safety. Traffic flow and speed are the two important indices used to
quantify the traffic performance while number of accidents is used as a performance
index for roadway safety. Statistical analysis method, decision tree, is used in this
study to evaluate how traffic performance parameters and rainfall affect road
accidents occurrence.

The rest of the paper is organised as follows: section 2 reviews the relevant
literature while section 3 describes the study area and data. The features of
accidents and traffic during rainy condition are explained in section 4 and then in
section 5, evaluation of the relationship among accidents, traffic and rainfall is
described thoroughly by detailing the model, analysis results and discussion.

Previous studies have found notable reductions in traffic volume and vehicle speed
during rainy condition in different locations. Smith et al. (2004) found 25%-30%
volume reduction during heavy rainfall using one year data while Agarwal et al.
(2005) found 5%-10% reduction during light rainfall using two years data of U.S.

In context of vehicle speed, Maze et al. and Agarwal et al. (2005) got quite similar
percentage (about 1% - 7%) of speed reduction on the freeways while Hranac et al.
(2007) revealed that light rain decreased free flow speed and speed at capacity at a
rate of around 2% - 4% and 8% - 10% respectively, and this reduction was higher
with the higher rain intensity.

Some previous studies have found that the reduction in volume and speed due
to rain calculated using HCM2000 method are not accurate (Agarwal et al.,2005 ;
Kyte et al.,2000). This may be because HCM 2000 method is based on research
carried using limited data from city of Toronto. So, more research in this area is
needed to increase the accuracy of rainfall impacts on traffic volumes and speeds.

It is obvious from different studies that road users are prone to accidents during
rainy conditions (Koetse and Rietveld, 2007; Eisenberg, 2003; Abdel-Aty and
Rajashekar, 2006). Chung et al. (2005) found that average frequency of accidents
on Tokyo Metropolitan Expressway during wet weather is 1.5 accidents per hour
which is significantly higher than the average frequency of 0.85 accidents per hour
during other times.

Keay and Simmonds (2004) found that rain increases accidents on freeways in
Melbourne (Australia) by up to nearly 5% depending on time of the day. The
database used for accident analysis in this study is restricted to accidents that
involved some sort of injury and required medical attention, not the damage-only
accidents. Changnon (1996) found that the impact of rain on road accidents also
varies spatially i.e. different impact in cities and rural areas. This may be because
of the differences of traffic volume and roadway conditions.


For this study, 93 km long section of the Pacific Motorway in Queensland (Australia)
connecting Brisbane and Gold Coast is selected as a case study. To analyse the
effect of rainfall on traffic parameters and accidents, both traffic and accident data
along with rainfall data is needed. So, hourly rainfall data from nearby weather
stations, 5-minute interval traffic flow and speed data collected by the detectors
along with crash information for the years 2009 and 2010 are used. In this study, all
types of accidents (damage-only, only injury or both) are considered for analysis.
Figure 1 shows the Pacific motorway (A) and three nearby three weather stations:

Figure 1. Pacific Motorway (Brisbane to Gold Coast) (ref. Google maps)

In order to find out the effect of rainfall, traffic parameters i.e. traffic flow and vehicle
speed on road accidents, the one year dataset is used for analysis. Statistical
software Minitab is used for the analysis. This analysis reveals the following results:

About 11% accident occurred during rainy conditions. Of these, 9 % occurred

when rainfall was in the range of >0 - 3 mm/hr and 2% occurred when rainfall was 3
- 5 mm/hr.

Figure 2. Road accident percentage at rain different rainfall amount

Speed- Flow diagram during accident and non-accident times shows that traffic
1.9% 8.8%
flow is in the free flow condition and vehicle travel on a free flow speed. But speed
during non-accident conditions has higher value during accident conditions, and in
Rainfall intensity
00 mm/hr
both cases, during rainfall time, speed are slightly less than non-rain time.
> 00 - 03 mm/hr
> 03 - 05 mm/hr
Figure 3. Speed – Flow 89.3%
diagram on different rainfall in accident time

Rain amount
0 mm >0 - 1mm
>1 - 2 mm
120 >2 - 3 mm
>3 - 4 mm






0 1000 2000 3000 4000 5000 6000 7000

Flow (veh/hr)

Rain amount
0 mm >0 - 1 mm
>1 - 2 mm
120 >2 - 3 mm
>3 - 4 mm






0 1000 2000 3000 4000 5000 6000 7000

Flow (veh/hr)

Figure 4. Speed – Flow diagram on different rainfall in non-accident time




In order to relate the rainfall, road accident and traffic parameters, the single
decision tree analysis method is adopted to show how the accidents occur. This
method is used here because it is easily understandable and can generate better
model using dataset having numerous missing data while other analysis method for
example regression analysis cannot work significantly in these cases. This type of
modelling acts by predicting the target variable based on the predictor variables. In
decision tree analysis, data comes in records of the form:

Where Y, is the target variable (dependent variable) and the vector x consists of the
predictor variables (independent variables) x1, x2, x3,..,xk etc.

For this study, accident number is considered as a target variable while rainfall
amount, traffic flow and vehicle speed are taken as predictors. Traffic flow and
speed at the time of accident and 5 minutes before the accident at the incident
location, at the nearest upstream location and at the nearest downstream location
are used as independent (predictor) variables for the modelling purpose.
Corresponding rainfall amount at the same time and location is also used as a
predictor variable.

DTREG statistical software package is used to analyse the dataset of about one
year (308 cases including accident occurring and non-accident occurring situation)
and to check proposed model which relate accident with rainfall and traffic

The output of the analysis results shows that the decision tree consists of total 71
nodes including 36 terminal nodes. Each node is built up by meeting the criteria of
specific value of flow or speed or rainfall and each node shows the probability of
accident for the specific criteria. Terminal nodes of each branch show whether the
accident occurs or not for the combination of predictor variables. By observing the
accident probability value of the terminal nodes, it is understandable that for which
combination e.g. flow, speed or rainfall, there will be an accident or no-accident. For
example, if one terminal node represents accident = 0, that means, there will be no
accident for this combination and if accident = 1, that means, there will be an
accident for the combination of flow and / or speed and / or rainfall. Accident value
between 0 and 1 represents the probability of accident occurrence for this
combination or branch.

The following figure shows the details of the decision tree model:

Figure 5. Decision tree model

The analysis output (Table 1) shows the importance percentage of each predictor
in the decision tree model. This describes priority levels of the variables to predict
the accident occurrence. The importance percentage of each variable is evaluated
by the most important variable whose importance percentage is taken as 100. It is
shown that flow at the location of incident 5 minute before the accident has got the
highest importance and rainfall has the least importance among the predictors.
Compared to speed, flow variable is more important for the prediction of accident

Table 1. Importance (in percentage) of each predictor in analysis

Variable Importance (in percent)

FLOW (location, 5 min before accident) 100.000
FLOW (upstream, at the time of accident) 84.717
SPEED (location, at the time of accident) 83.642
FLOW (location, at the time of accident) 82.630
FLOW (downstream, at the time of accident) 81.100
SPEED (location, 5 min before accident) 74.675
SPEED (upstream, 5 min before accident) 71.904
FLOW (upstream, 5 min before accident) 68.787
FLOW (downstream, 5 min before accident) 65.086
SPEED (downstream, at the time of accident) 64.023
SPEED (downstream, 5 min before accident) 51.226
SPEED (upstream, at the time of accident) 41.600


From the decision tree analysis, it is clearly shown that accident probability depends
more on the traffic flow and speed than rainfall. More specifically, accidents
occurrence depend mainly on traffic performance parameters i.e. flow and speed at
the accident location followed by traffic parameters upstream of the accident
location compared to the downstream condition.

In the decision tree analysis, the first and primary branch is built up by the criteria
of speed at the location of incident at the time of accident makes the other branches.
But it is observed that accident probability is more for the speed less than 102 km/hr
at the location of accident. The decision trees are added further with the criteria of
the flow at the location before 5 minutes of accident occurrence. Again it can be
seen from the decision tree that accident probability is higher for comparatively
lower value of traffic flow at the location of incident. For example, for

flow<=128veh/hr, incident probability is 0.833 whereas it is 0.50 for flow>128veh/hr.
The branches are further extended by the speed at the location before 5 minutes of
accident. It is noteworthy that node 10 shows that if the speed at the location before
5 minutes of accident is less than 76.5 km/hr by satisfying the upper criteria of
branches, the probability of accident is 1 i.e. there must be an accident. The
branches are further built up by the criteria of downstream flow at accident time
followed by upstream flow, upstream speed before 5 minutes of accident

Rainfall criteria are shown in the end nodes, node 38 and node 39. The criteria
used for rainfall effect is whether rainfall amount is less than or higher than 0.5 mm.
If the rainfall amount is more than 0.5 mm, after meeting previous criteria, accident
probability is 1 i.e. accident will happen. On the other side, there will be no accident
if the value is less than 0.5 mm.

There are few combinations of speed and flow for which, there will be an
accident on the roadway. For example:

1.Speed at the location > 102.5 km/hr + flow at the location 5 minutes before <=
258 veh/hr + speed at the upstream 5 minutes before > 97.5 km/hr.

2.Speed at the location <= 102.5 km/hr + flow at the location 5 minutes before >
129 veh/hr + speed at the upstream 5 minutes before <=76.5 km/hr.

3.Speed at the location <= 102.5 km/hr + flow at the location 5 minutes before <=
129 veh/hr + flow at the downstream <=4630 veh/hr.

This paper analyses how the traffic performance parameters and rainfall affect the
accident occurrence by using the decision tree analysis method has been found that
road accident probability on the studied freeway depends significantly on the
condition of traffic performance parameters: traffic flow and vehicle speed rather
than rainfall. Also, this study reveals the combination of flow, speed and rainfall for
which the road users on the roadway are prone to accidents. The analysis result
indicates that speed at the location, flow at the location 5 minutes before and speed
at the upstream 5 minutes before or flow at the downstream has the significant
contributing effect for occurrence of accident.

The findings of the study can be used to reduce road accidents by considering
the known combination of traffic flow, speed and rainfall and controlling traffic to
avoid these hazardous conditions resulting in accidents. Future research direction
will be conducted on how to improve the road traffic performance on the freeway by
adopting different traffic control measures in order to reduce crash risk probabilities.

Abdel-Aty, M. A. & Pemmanaboina, R. (2006) Calibrating a real-time traffic crash-

prediction model using archived weather and ITS traffic data. Intelligent
Transportation Systems, IEEE Transactions on, 7, 167-1 74.
Agarwal, M., Maze, T. & Souleyrette, R. (2005) Impacts of Weather on Urban
Freeway Traffic Flow Characteristics and Facility Capacity. in: Proceedings of the
2005 Mid-Continent Transportation Research Symposium.
Changnon, S. A. (1996) Effects of summer precipitation on urban transportation.
Climatic Change, 32, 481-494.
Chung, E., Ohtani, O., Warita, H., Kuwahara, M. & Morita, H. (2005) Effect of rain on
travel demand and traffic accidents. Intelligent Transportation Systems, 2005.
Proceedings. 2005 IEEE.

Faouzi, N.-E. E., Billot, R., Nurmi, P. & Nowotny, B. (2010) Effects of adverse
weather on traffic and safety: state-of-the-art and a European initiative. SIR WEC
2010: 15th International Road Weather Conference .

Goodwin, L. C. (2002) Weather Impacts on Arterial Traffic Flow. Prepared for the
FHWA Road Weather Management Program.

Hranac, R., Sterzin, E., Krechmer, D., Rakha, H. And Farzaneh, M. (2007) Empirical
Studies on Traffic Flow in Inclement Weather. Federal Highway Administration, U.S.
Department of Transportation, Washington, DC.

Keay, K. & Simmonds, I. (2005) The association of rainfall and other weather
variables with road traffic volume in Melbourne, Australia. Accident Analysis &
Prevention, 37, 109-124.

Koetsea, M. J. & Rietveldb, P. (2007) Climate Change, Adverse Weather

Conditions, and Transport: A Literature Survey. In Proceedings of the 9th NECTAR
Conference. CDROM. Network on European Communication and Transportation
Activities Research (NECTAR).

Kyte, M., Z. Khatib, P. Shannon, And F. Kitchener (2000) Effects of Environmental

factors on Free- Flow Speed. Transportation Research Circular presented at the
Fourth National Symposium on Highway Capacity, Maui, Hawaii, pp.108-119.

Smith, B. L., Byrne, K. G., Copperman, R. B., Hennessy, S. M. & Goodall, N. J.

(2004) An Investigation into the Impact of Rainfall of Freeway Traffic Flow. Paper
presented at the 83rd annual meeting of the Transportation Research Board


Mohd Farriz Basar1, Khalil Azha Annuar2, Firdaus Ab Halim3, Norhaslinda

Hasim4 and Muhamad Khairi Aripin5
1, 2, 3, 4, 5 Universiti Teknikal Malaysia Melaka (UTeM), Malaysia , , ,
norhaslinda@utem .edu .my,

ABSTRACT : This paper describes the design and development of Smart Motorcycle Safety
Vest (SMS-V) for motorcyclist in Malaysia. SMS-V is the reflex of light shirts and fitted with
red and bright yellow LEDs which become the third light for brake and signals. According to
statistics of road accidents in Malaysia, accidents involving motorcyclists has a high number
and it increases every year. As a result, many young people died and some have suffered
permanent disabilities and this is a great loss to society and country. Therefore, there should
be an alternative effort to ensure that this group is not a victim in the accident. Basically, the
accident on the motorcycle is happened because most of motorcyclist are not concerned
about safety where the mistakes that frequently happen is wearing dark colored clothing.
According to the research experimental works, a motorcycle can only be known by the
headlight and taillight of motorcycle at a distance of less than 180 meters. If a vehicle such
as car and lorry at a speed of 90km/h faced with a motorcyclist, the vehicle only has 6
seconds to be alert to the presence of motorcyclists. Instead, this product able to assist road
users about the presence of a motorcyclist at a distance of 500 meters, which is three times
better than normal conditions. Based on motorcycle registration statistics released by the
Malaysian Road Transport Department, as at January 2010, a total of 8,940,230 motorcycles
were registered in Malaysia. If SMS-V is used only 50% of the number stated above,
roughly, it will become a fashion and give a great impact to the motorcyclists in Malaysia as
the product itself offers an interesting look. Hence, it is believed that this research is able to
reduce the number of road accidents which often happen to motorcyclists in Malaysia.

Keywords : Safety Vest, Motorcyclist, Third Light, Light-Emitting Diode (LED), Malaysia

Road accidents have caused huge losses to the country. It involves an injury,
suffering and death but the histories of road accidents seem endless in
Malaysia. Thus, the desire to travel safely and smoothly remains as intention of
the government and the individual road users. Therefore, currently many
different efforts have been developed to avoid vehicle crash on the road
especially that involve with motorcyclists.In Malaysia, 96% of injuries and deaths
from accidents are caused by driver carelessness. Figure 1 shows the example
of motorcycle accident in Malaysia.

The invention basically is well suited for motorcycle rider, where the safety
vest (SMS-V) equipped with the third safety brake LED light, left and right signal
LED light and LED taillight that is more transparent to road users. The reason
why LED light is used because most of the road accident happens due to
visibility where the road user could not see the presence of the motorcyclist in
front or next to them.

Figure 1. The example of motorcycle accident in Malaysia

Obviously, the major factor to road accidents is the result of failure of the
road user himself when on the road. Table 1 shows the Statistics of Road
Accidents in Malaysia referred to the Kurikulum Pendidikan Pemanduan from
Road Transport Department of Malaysia (JPJ) where the number of road
accidents involving motorcyclists has a high number and it increases every year.
As a result, many young people died and some have suffered permanent
disabilities. This is a great loss to society and country as they are the heirs and
human capital for national development.

Table 1. Statistics of Road Accidents In Malaysia

Vehicle Involved 2005 2006 2007 2008 2009
Motorcycle 97,072 104,302 111,958 113381 113962
Car 376,061 411,976 428,475 441109 472307
Van 19,031 20,465 21,187 20501 19220
School Bus 1,106 1,235 1,274 1134 1160
Factory Bus 1,126 1,176 1185 1120 864
Express Bus 1,836 2,156 2,091 2056 2348
Small Lorry 12,372 13,915 17,464 18290 16459
Lorry 20,284 20, 406 19,138 18262 19188
Trailers and Tanker
9,406 10, 494 11,193 11101 11077
Four-wheel Drive 19,106 20,928 22,018 22656 23581
Taxi 7,043 7,754 8,816 8816 8669
Bicycle 2,751 2,857 2,693 2443 2486
Others 9,362 12,274 14,935 10183 9294
Total of Vehicle
581,082 635,082 668,173 676114 705623


Generally, the research works consists of literature review on previous work,

development of RF controller circuit, design and development smart vest
attached with LED and finally field experimental tests. The purpose of adopting
the research methodology is to obtain the desired experimental results during
the field experimental work. The main intention of this research is mainly to
develop a safety vest that can be used as the third light for brake and signal of
the motorcycle, where it is facilitated with wireless technology.

The following statement detailed the research methodology that was

practiced in this experimental project. Extensive literature reviews was done to
understand and identify the problem that going to be solved in this project.
Besides that, the studies have been done on the previous research works on the
safety matter.

In the Stage 1, the materials and components are assemble to build up the
transmitter and receiver controller circuit using the RF technology. The
transmitter controller circuit is located under the seat of motorcycle and the
receiver is attached at safety vest. Then, in Stage 2, the white, yellow and red
LED strip is attached at safety vest. In addition the LEDs are connected with the
receiver controller box. In Stage 3, the field testing is conducted in order to
investigate the performance of SMS-V and the field testing was conducted at
night. Besides that, the testing has been done to compare the effect of both woth
and without SMS-V. Analysis is carried out to verify the performance of SMS-V.
The analysis is aligning with the reliability and functionality of the RF controller
box in the SMS-V.

B.1 Literature Review

a. Nowadays, global economic crisis already becoming a gigantic anxiety

for some countries in the world. In general, almost medium earners have
to change their lifestyle including the vehicle they ride. Due to that, The
numbers of motorcycle riders has increased dramatically, which also the
demand of high quality light vest becomes imminent. Nearly half of
accident occurred correspond to the motorcycle riders are due to the
visibility of these motorcycle rider to the other road users. During night,
fog and even haze, the risk increase as the traffic visibility become less.

b. Refer to the [http ://www.themalaysian.
accidents-ranking-malaysia.html, August 2006.], most of these
transportation deaths and injuries occur on the nation’s highways. The
United Nations has ranked Malaysia 30th among countries with the
highest number of fatal road accidents, registering an average of 4.5
deaths per 10,000 registered vehicles.
c. According to Hafiz Muhammad Atiq et. all, nowadays vehicle detection
and shape recognition based on optical sensors has becoming an area of
active research among researcher and development departments of civil
defence, traffic police and universities. A successful vehicle detection
and shape recognition algorithm will pave the way for vehicle recognition,
vehicle tracking and safety improvements.

d. According to research done by Abdulhakam.AM.Assidiq et al., the effect

of variety of road condition contributes to road accidents. In addition, the
authors mention that, lane detection is not easy due to the varying road
condition while driving. For that reason, one of the main technology
involves is computer vision which become a powerful tool for sensing the
environment and has been widely used in many application by the
intelligent transportation systems (ITS). However, optical sensors and
computer vision is quite expensive and for that reason SMS-V project is
employed the RF technology.

e. In addition, many inventors have come out with many inventions of

jacket, vest and traffic signalling system. Refer to U.S Pat. No. 6,558,016
to Restauro (2003), a motorcyclist garment with illuminated traffic signals
comprising: a garment having a back side, a neck opening, and arm
openings; and a light-emitting assembly. The light emitting diode (LEDs)
are powered and triggered by the signal tapped directly from the signal at
the motorcycle tail, break and left and right signal fixture. Although the
garment conveniently getting the power source from the motorcycle tail,
the motorcyclist has inconveniently reconnect and disconnect when they
embark and disembark from the motorcycle especially when having to do
some work which require small time frame such as going to the toilet or
automated teller machine (ATM). On the other hand, the SMS-V employs
a wireless technology in order to provide comfort to the motorcyclist.
f. Then, U.S Pat. No. 7,109,857, a brake light warning system for safety
helmets includes a transmitter module adapted for mounting to a vehicle,

such as a motorcycle, and a receiver module adapted for mounting to a
safety helmet. The transmitter module is configured to continuously
transmit a transmission signal when a brake of the vehicle is disengaged
and discontinue transmission of the transmission signal when a brake of
the vehicle is engaged. The disadvantage of this invention is the limited
application, of only giving a breaking signal to the road users. By having
the breaking light warning system on the helmet could cause the system
to be very unreliable due to riders occasionally turns their head left or
right looking for direction or just distracted by the road objects.

g. U.S Pat. No. 2009/0134992 A traffic signaling device system for

motorcyclists comprising a safety brake and running light assembly
mounted to a motorcycle jacket. The safety brake light has an L.E.D.
housing supported by an outboard support member and a leather piece.
The disadvantage of this system is it requires the rider to mount the
LEDs on the jacket, connect some wiring to the motorcycle tail and
mount the control box to the gas tank or near the motorcycle controls.
This is very time consuming and inconvenient to the users.

B.2 Design and Development of Smart Motorcycle Safety Vest (SMS-V)

The smart motorcycle safety vest as shown in Figure 8, comprising third
safety brake LED light, left and right signal LED light and red LED taillight
embedded inside the safety vest. The strip LED is sewed between the
materials of the vest to have a rigid position.

The most significant concept in this project is the utilization of

motorcycle signal as the triggering signal for the LED on the SMS-V. Besides
that, the use of wireless controller (RF) for the purpose to transmit the
motorcycle signal to the receiver controller circuit located in the vest is also a
part of novelties. In addition, the wire connection and arrangement of
electronics component including the arrangement of LED strip; the
transmitter and receiver controller circuit in the SMS-V is original design.
Finally, the way of the transmitter controller box is tap with the existing
motorcycle wiring is also a new technique.

Stage 1
The intention of stage 1 is to build up a transmitter and receiver
controller for the SMS-V. The controller basically using the radio frequency
(RF) technologies with the frequency employed is 40MHz. Transmitter
controller box as shown in Figure 3 is powered by two AA 1 .5V batteries.
This controller need to be attached at the motorcycle and the best place is
under the seat of motorcycle.

Figure 3. Transmitter Controller Box of SMS-V

Figure 4 shows the situation before and after the transmitter controller
box connected with the motorcycle wiring system. In addition, this controller
box connected is placed under the motorcycle seat suit for the reason to
provide a convenient and easiness to the motorcyclist. Besides that, if the
transmitter controller box is kept under the motorcycle seat, it will prevent the
controller from any damage.

Figure 4. Transmitter Controller Box attached in the motorcycle

Basically, the user has to take only six wires from the existing wiring in
motorcycle and then connect the six wires to the transmitter controller box as
shown in Figure 5. It is good to know that, most of the motorcycles in
Malaysia have the wiring under the seat, thus it is easy to attach the
transmitter controller box under the motorcycle seat. The installation of the
transmitter controller box can be completed approximately less than five
minutes. In addition, the controller box is only need to be installed during the
first time used and it is permanently connected.

Figure 5. Transmitter Controller Box Connect to the motorcycle wiring.

Figure 6 depicted the receiver controller box of SMS-V where it is

powered by three AA 1 .5V batteries and this controller is attached inside the
vest. Essentially, this controller will receive a signal which is trigger and send
out by the transmitter control circuit. Simultaneously, the receiver will activate
the circuit for switch on the brake light, left and right signal attached at the
SMS-V dependable to the signal send by transmitter.

Figure 6. Receiver Controller Box of SMS-V


4.5V DC
12V DC Brake
(Red LED)

Figure 6 depicted the receiver controller box of SMS-V where it is

powered by three AA 1 .5V batteries and this controller is attached inside the
vest. Essentially, this controller will receive a signal which is trigger and send
out by the transmitter control circuit. Simultaneously, the receiver will activate
the circuit for switch on the brake light, left and right signal attached at the
SMS-V dependable to the signal send by transmitter.

Left Signal
Action Receiver Control
Right Signal
Action circuit
(6 wires from
circuit connect Left Signal Right Signal
Brake (Orange (Orange
Action with transmitter
controller LED) LED)

Figure 7. Wiring diagram of the SMS- V system

Figure 7 is the overall working system of the SMS-V. It can be

concluded that the system is not complex and due to that, this innovation is
not expensive and easy for troubleshooting if there is any problem in the

Stage 2
The invention of SMS-V has the strip LED which is sewed within the
vest. It has third safety brake LED light, left and right signal LED light and
LED tail light powered by a light weight rechargeable battery that offers high
intensity thus excellent visibility to the motorcycle riders. The invention
comprises of strip LED sewed within the vest, battery, receiving controller
circuit and transmission control box. The transmitter controller box is placed
under the rider seats, which connected directly to the motorcycle signal
which is brake light, left turn signal and right turn signal. Whenever the riders
operates the brake, left or right signal, the transmitter controller box is trigger
and transmit a corresponding signal to the receiver controller vest which is
located inside the SMS-V.
Referring to Figure 8, it can be seen that SMS-V comprises a flexible
strip LEDs yellow, red and white attached in front of the SMSV. Number 1
and 6 represent the LED taillight, number 7 and 8 represent right signal LED
light, number 9 and 10 represent the left signal LED light and number 11 and
12 represent the brake LED light. In addition, number 2, 3, 4 and 5 is an
additional for front light. Besides that, for the rear side of the SMS-V, number
Front 13, 14, 15 and 16 represent the LED taillight, number 17, 18 and 19, 20
represent the left and right signal respectively and number 21 to 24
represent the brake light.

Figure 8. The front and rear of the Smart Motorcycle Safety Vest (SMS-V)

The entire LED positive terminal is soldered together according to its

function while the negative terminal connected to the ground. All wires were
connected directly to the receiver controller box which is powered by dry cell
battery AA 3V. Besides that, the dry cell battery 12V is used to power the
LED strip of the SMS-V and can be attached to the rider’s pants.

Basically while the rider use brake, left or right signal, the transmitter
control box will trigger and transmit the respective signal to receiver control
box. Then, the receiver control box will respond accordingly. While the rider
wearing SMS-V and turn on the switch at the receiver control box, the tail
light (no 1, 6, 13, 14, 15 and 16) and front light (no 2 -5) will turn on. During
riding, while the rider pulls the front brake, the SMS-V brake light (no 11, 12,
21 - 24) will turn on. While the rider push the right or left signal, the SMS-V
right (no 7, 8, 19 and 20) or left (9, 10, 17 and 18) signal will blink. In
addition, the LED attached to the vest will blink in the same frequencies with
the motorcycle signal.

B.3 Field Testing Activities

The field testing activities have been conducted at Universiti Teknikal

Malaysia Melaka (UTeM) at night with a real motorcycle on the road to verify
the performance of the SMS-V.

Table 2: Time to Alert for Other Vehicles with Varying the Speed of Other Vehicle during
With and Without Using SMS-V
Without SMS-V With SMS-V
Distance : 150 meter Distance : 500 meter
Speed of the Speed of the
Time to alert Time to alert
Other Vehicle Other Vehicle
50 km/h 10.8 second 50 km/h 36.0 second
60 km/h 9.0 second 60 km/h 30.0 second
70 km/h 7.7 second 70 km/h 25.7 second
80 km/h 6.8 second 80 km/h 22.5 second
90 km/h 6.0 second 90 km/h 20.0 second
100 km/h 5.4 second 100 km/h 18.0 second
110 km/h 4.9 second 110 km/h 16.4 second
120 km/h 4.5 second 120 km/h 15.0 second

Table 2 shows the distance with varying speed of vehicle for two different
conditions which are with and without SMS-V. It shows that by using SMS-V,
the time taken for a lorry or a car to alert to the motorcyclist is greater than
without using the SMS-V.

According to the research experimental works, a motorcycle can only

be known by the headlight and taillight of motorcycle at a distance of less
than 160 meters as shown in Figure 9. If a vehicle such as car and lorry at a
speed of 90km/h faced with a motorcyclist, the vehicle only has 6 seconds to
be alert to the presence of motorcyclists. Instead, this product able to assist
road users about the presence of a motorcyclist at a distance of 500 meters,
which is three times better than normal conditions.

Figure 9: Difference in Distance between With and Without SMS-V

Figure 10 shows the graph where time to alert in seconds versus

speed of other vehicle in kilometre per hour. The graph indicates that the
user who wearing the SMS-V is extremely safer than motorcyclist where do
not use the SMS-V.

40 10: Different in Distance between With and Without SMS-V
20 there are several advantages of the invention which are very
In conclusion,
important in ensuring
the developed Smart Motorcycle Safety Vest (SMS-V) Se...
function as an5alternative way to reduce the number of motorcycle accidents in
50 60 70 80 90 100 110 120 130

Speed of the Other Vehicle (km/h )

Malaysia. The advantages are to provide the motorcycle riders to be more
conspicuous to other road user; to bring forward wireless triggering technology
into the LED safety vest, which is very convenient compared to having many
wires which need to connect manually to the motorcycle wiring. Furthermore,
according to the field testing result, the SMS-V proves that, it is three times
better than normal condition or without using SMS-V. All these features are
useful to assist other road user about the presence of motorcyclists. Hence, the
SMS-V will give a great impact to the motorcyclist in Malaysia as the product
itself offers an interesting look. Further research in terms of renewable energy
such as SMS-V powered by solar system will be conducted to enhance the
features of the SMS-V.


Hafiz Muhammad Atiq, Umar Farooq, Rabbia Ibrahim, et al. (2010). Vehicle
Detection and Shape Recognition using Optical Sensors: A Review.
Proceedings of the Bizarre Fruit Postgraduate Conference. pp 107 – 123.
University of Salford.UK

Abdulhakam.AM.Assidiq, Othman O. Khalifa, Md. Rafiqul Islam, et al. (2008). Real

Time Lane Detection for Autonomous Vehicles. Proceedings of the International
Conference on Computer and Communication Engineering 2008, May 13-
15,2008 Kuala Lumpur, Malaysia.

Yun Luo, Jeffrey Remillard, and Dieter Hoetzer (2010). Pedestrian Detection in
Near-Infrared Night Vision System. 2010 IEEE Intelligent Vehicles Symposium
University of California, San Diego,CA, USA. June 21-24, 2010.

Junji Hirasawa & Masayoshi Kakikura (2004). Motion Analysis of Motorcycles: A

Study on DCT Mechanism. Proceedings of the Bizarre Fruit Postgraduate
Conference. pp 107 – 123. University of Salford.UK

Evangelos D. Bekiaris, Andrea Spadoni, & Stella I. Nikolaou (2009). SAFERIDER

Project: New Safety and Comfort in Powered Two Wheelers. Humans System
Interaction 2009- University of Catania, Catania, Italy, May 2 1-23, 2009.

Marco Pieve, Francesco Tesauri, & Andrea Spadoni (2009). Mitigation Accident
Risk in Powered Two Wheelers domain: Improving effectiveness of Human
Machine Interface Collision Avoidance System in Two Wheelers. Humans
System Interaction 2009- University of Catania, Catania, Italy, May 2 1-23, 2009.

Marco Pieve, Francesco Tesauri, & Andrea Spadoni (2009). Mitigation Accident
Risk in Powered Two Wheelers domain: Improving effectiveness of Human
Machine Interface Collision Avoidance System in Two Wheelers. Humans
System Interaction 2009- University of Catania, Catania, Italy, May 2 1-23, 2009.

Road Transport Department of Malaysia (JPJ) (2010). Kurikulum Pendidikan

Pemanduan, Malaysia, pp 77 – 80, 2010.

RASHIDAH Ab. Rashid 1, ISMAIL Yusof 2 and Mohd ERWAN Sanik 3

1, 2, 3 Universiti Tun Hussein Onn Malaysia, Johor, Malaysia

ABSTRACT: Malaysia possesses the most complete and modern roads and highway
infrastructure that links the whole country. In the interest of safety, the government remains
constantly vigilant to possible hazards along the roads and highways. In Malaysia, there are
three typical types of road safety barrier which are widely used, namely wire rope, w-beam
guardrail and concrete barriers. Wire rope safety barrier is seen as the best solution to keep
vehicle damage to a minimum as well as to reduce the risk of serious injury. Wire rope
barrier has recently used in Malaysia due to the low cost of maintenance, repair time, and
ease of construction. However, to date, there is lack of study on traffic barriers especially
wire rope. Therefore, this study is carried out to conduct hypothesis testing using statistical
approach to verify either wire rope is safe or not. Another purpose of this study is to analyze
the accident characteristic based on the police reports which involve with traffic barriers.
Data collection consists of accident data and field work measurement for both federal and
expressway selected roads. The SPSS statistical software is used to analyze the accident
data. As expected, the result shows with sufficient evidence that wire rope is safer than
guardrail in term of accident casualties. Based on the inspection of the guardrail and wire
rope installation, most of the stretches generally complied with the JKR and the REAM
specifications. Finally, this pioneer study may be applicable for future research on traffic
barriers as well as in any justification of accident countermeasure along Malaysian roads.

Keywords: Guardrail, Road safety barriers, Statistical approach, Wire rope

Malaysia is facing a serious road accident problem. Every year for the last 5
years, more than 6000 people were killed on its road (MOT, 2006). There is a
need to provide highway design engineers with a choice of safe and effective
guardrail and medium barrier systems. The application of traffic barriers should
result in safer highways. However, traffic barriers have sometimes been
misused, perhaps due to a misconception of their function. They are not installed
to protect roadside objects or prevent accident occurrence, but to protect vehicle
occupants from possible serious injury (REAM, 2006). Roadside barriers are
designed to enhance the safety of the road infrastructure by containing errant
vehicles and reducing the severity of off-path collisions (REAM, 2006). However,
in Malaysia, there are lack of study on traffic barriers especially wire rope.
Usually in Malaysian road, there are three types of road safety barrier which are
widely used, namely wire rope, w-beam guard rail and concrete median but
which one have perform better in term of safety and how much it’s installation
comply with the specification?


B.4 Data Collection Technique

Based on the related works, this study is then started by identifying the study
locations or routes to be inspected in order to achieve the objectives. This
step led to the second stage which was the most challenging part in this
study. It is the data collection. Data were categorized into two types;
primary and secondary data. Data collection was carried out to obtain data
such as road geometry and layout, road history, and accident data statistic.
Primary data are collected using site inspection technique while secondary
data are gathered from the Jabatan Kerja Raya offices and the Police
Stations. The inspection is done by randomly checking for the barrier
specification at 10 points for each location. Table 1 shows the name of
location with respect to the Road ID. Type One (1) means wire rope and type
two (2) is w-beam guardrail.

Table 1. Road ID
Location Road ID Safety Barrier Type
Kuala Perlis – Changloon 1 2
Kuala Perlis – Changloon 2 1
Muar by-pass 3 1
Jasin 4 2
Alor Gajah 5 2

B.5 Data Analysis Approach

h.Graphical Analysis

Using graphical analysis, the pattern of accident with regards to safety

barriers type can be depicted and the different can be analysed in term of
location and traffic composition.

i.Statistical Analysis

Using statistical analysis, the different in term of safety risk can be

proven using t-test and F-test or ANOVA.

3.1 Graphical Analysis

Graphical analysis is carried out to depict the accident characteristic when using
wire rope and w-beam guardrail as the road safety barrier. The accident data
from 2007 to 2009 is taken and analyzed.

a) Total Road Accident

In this study, the types of road accident that taken into consideration are
“fatal”, “serious injury” and “slight injury”. These three types of accident can
be defined as Personal Injury Accident (PIA). These data are more reliable
and have higher weightage compared to damage-only crashes (Harnen et
al., 2003). Figure 1 shows the percentage of accident which have been
categorised by type of injury with different type of barrier. Referring to Figure
1, all type of accidents is higher at the location using guard rail as road
safety barrier compared to wire rope. Of these, fatal accident involving
guardrail location is 66.67% (16 cases), while 33.33% (8 cases) at wire rope
location, respectively. Then, for serious injury involving guardrail location is
63.64% (11 cases), while 36.36% (4 cases) at wire rope location and slight
injury involving guardrail location is 74.03% (23 cases) while 25.97%(13
cases) at wire rope location respectively. Therefore, it shows that locations
with wire rope installation are safer than guardrail with respect to the number
of accident.
Figure 1. Percentage of Accident at Location with W-Beam Guardrail and Wire Rope
40 73.33
30 66.67 63.89
20 36.11
33.33 26.67
fatal serious injury slight injury

wire rope guardrail

b) Accident Data by Traffic Composition
Figure 2 shows the number of accident by traffic composition at location w-
beam guardrail from 2007 to 2009. Referring to the figure, the number of
accident are fluctuated, however the percentage of fatal accident involving
car is the highest which is 50% (8 cases) than, the second highest is
motorcycle 31.25% (5 cases). While for serious injury, the highest involving
car is 54.55% (12 cases).

Figure 2. Number of Accident by Traffic Composition at Location W-Beam Guardrail

Figure 3 shows the number of accident by traffic composition at
50 54.55
location wire50rope from 2007 to 2009. Referring to the figure, the percentage
of fatal accident and serious injury involving motorcycle is the highest which
is 62.5% 31.25 (4 cases) respectively. From the analysis, it
(5 cases) and 100%
22.73 22.5 22.5
10that, wire rope is dangerous for motorcycle.
12.5 12.59.09 6.25 13.64
Figure 3. Number of Accident by Traffic Composition at Location Wire Rope
car motorcycle lorry bus

100 serious injury slight injury

62.5 61.54
25 0 23.08 12.5 0 15.38

car motorcycle lorry

3.2 Ranking Accident Point Weightage

Point weightage formula used to determine the study locations based on

accident point weightage. The factor or weightage values published by the
Highway Planning Unit (HPU) in the Interim Guide on Identifying, Prioritising and
Treating Hazardous Locations on Roads in Malaysia. Equation 1 is used to
calculate Accident Point Weightage at a section (Nor Esnizah, 2008).

Weightage Formula = X 1 (6.0) + X 2 (3.0) + X 3 (0.8) + X 4 (0.2) (1)

Where, X1, X2, X3 and X4 are the number of fatal, serious injury, slight injury,
and damage only, respectively.

3.3 Ranking of the Top Fourty Accident Section (Kilometre)

Table A and B in Appendix show the top-fourty accident ranking by weightage

for w-beam guardrail and wire rope, respectively. The analysis is based on
compile data in three years from 2007 to 2009.

3.4 Statistical Analysis

Two types of hypothesis testing are carried out namely T-Test or mean analysis
and also F-Test or analysis variance. Of these, variance analysis is more
specific to the location of barrier.

a) Mean Analysis (t-test)

Table 2 shows the statistical summary for both types of safety barriers. Two
tailed t-test is used to compare means between wire rope and guardrail. The
null hypothesis for t-test is that means for the two types of barriers are the
same for fatal, serious injury and slight injury accidents.


Where µi is the mean of group i, in this case barrier type i and the alternate
hypothesis is that means for two barrier types are significantly different.

H1 :µWR≠µ GR

At 95% confident, if the observed significant level or probability is less

than 0.05, null hypothesis is rejected and can be concluded that the test is
statistically significant. Referring to Table 11, the probability for serious injury
is higher than 0.05, so the null hypothesis not rejected and the difference is
significant. However, for fatal and slight injury, the probability is less than
0.05 and the null hypothesis is rejected and can be conclude that the
difference is statistically significant.

Table 2. T-Test Result for each Type of Accident

Table 3 shows the descriptive statistic for type of accident. By

referring table below, for fatal accident, the value of mean for type 1 is equal
to 0.23 and for type 2 is 1.18. From the result, it show that, the mean of fatal
accident at location using with wire rope is less than guardrail, so wire rope
is assumed safer than guardrail. Then, there have no different of serious
injury at both location either using wire rope or guardrail. Next, by referring to
slight injury accident, the value of mean for type 1 is equal to 0.50 and for
type 2 is 1.43. From the result, it show that, the rate of slight injury by using
wire rope as road safety barrier is less than guardrail, so wire rope is
assumed safer than guardrail.

Table 3. The Descriptive Statistic for Type of Accident

b) Variance Analysis (F-Test)

Table 4 shows the statistical summary for location study based on Road ID.
Analysis variance F-Test is used to compare the value of means between
Road ID. The null hypothesis for f-test is for all Road ID are the same for
fatal, serious injury and slight injury accidents.

H0:µ1=µ2 =µ3=µ4=µ5

Where µi is the mean of group i, in this case Road ID is i and the

alternate hypothesis is that means for all Road ID are significantly different.

H1:µ1=µ2 =µ3=µ4=µ5

At 95% confident, if the observed significant level or probability is less

than 0.05, null hypothesis is rejected and can be concluded that the test is
statistically significant. Referring to Table 4, the probability for all types of
accident include fatal, serious injury and slight injury, is less than 0.05, so the
null hypothesis is rejected and can be conclude that the difference is
statistically significant.

Table 4. F-Test Result for each Type of Accident

Table 5 show the descriptive statistic for fatal accident based on

Road ID. The mean value for Road ID 3 and 5 is subset, so there have no
different of fatal but for Road ID 1, 2 and 4, there have different mean of
fatal. Road ID 4 is the highest value fatal accident and the second highest is
Road ID 1. From the result, it show that, both of the location which have the
highest fatal accident use w-beam guardrail as road safety barrier. So, it can
be assumed that wire rope is safer than guardrail.
Table 5. The Descriptive Statistic for Fatal Accident based on Road ID

Table 6 shows the descriptive statistic for serious injury accident

based on Road ID. The mean value for Road ID 4 is subset, so there have
no different of serious injury but for Road ID 1, 2, 3 and 5, there have
different mean of serious injury. Road ID 1 is the highest value of serious
injury and the second highest is Road ID 3. From the result, it show that, the
location which have the highest serious injury use w-beam guardrail as road
safety barrier and the second highest is use wire rope as road safety barrier
but if study make in depth about the road condition for each location, at Road
ID 3, there have problem with the road settlement, so it can affects the
number of accident.

Table 6. The Descriptive Statistic for Serious Injury Accident based on Road ID

Table 7 shows the descriptive statistic for slight injury accident based
on Road ID. The mean value for Road ID 3 and5 is subset, so there have no
different of slight injury accident but for Road ID 1, 2, and 4, there have
different mean of slight injury accident. Road ID 4 is the highest value of
slight injury accident and the second highest is Road ID 1. From the result, it
show that, the location which have the highest and the second highest slight
injury accident use w-beam guardrail as road safety barrier. So, it can be
assumed that wire rope is safer than guardrail.
Table 7. The Descriptive Statistic for Slight Injury Accident based on Road ID

From the hypothesis testing, it’s clearly show that the fatal accident,
serious injury and slight injury accident is less at location using wire rope.
The fatal accident is the highest at Road ID 4 which is Jasin (PLUS
Expressway) and the second highest is Road ID 1 which is Kuala Perlis-
Changloon Highway. Both of road use w-beam guardrail as road safety

In conclusion, based on the graphical analysis, it shows that locations with wire
rope installation are safer than guardrail with respect to the number of accident.
Moreover, statistical analyses show that guardrail gives higher risk to the road
users’ safety compared to wire rope. Several recommendations for further
investigation are as follow:
i) The deflection of vehicles crashing into WRSBs needs to be better
understood to reduce the likelihood of serious secondary collisions.
ii) Detailed barrier location (the actual side of the road on which it is located);
iii) Detailed target crash analysis (which target crashes are relevant to the
barrier length);

iv) And other available detail such as, whether there appears to be significant
differences in effect between installation lengths, and if there exists a
minimum threshold length for effectiveness and impact of barrier distance
from travel lane.


BCO Ventures Sdn. Bhd. (2010). Brifen Wire Rope Safety Fence System. Kuala
Lumpur, Malaysia.: Trade Brochure

Che Ali, C.H., (2010). Improving the Design and Installation of Traffic Barriers along
Major Roads in Malaysia. Proc. of conference on traffic barriers. REAM.

Harnen,S., Radin Umar.R.S.,Wong,S,V., and Wan Hashim,W.I. (2003), Motorcycle

Crash Prediction Model for Non- signalized Intersection. IAISS Research,
Vol.27, No.2.

Inov Fence Sdn. Bhd. (2010). Wire Rope Guardrail System. Selangor, Malaysia:
Trade Brochure.

Nor Esnizah, I. (2008). Accident Prediction Model Federal Route 50 (Batu Pahat –
Ayer Hitam). University Tun Hussein Onn Malaysia. Thesis B.s.c.

NCHRP Report 350-Recommended Procedures for the Safety Performance

Evaluation of Highway Features, National Co-operative Highway Research
Program, U.S. Department of Transportation, Federal Highway Administration.

REAM (2006), Guidelines on Design of Result and Selection of Longitudinal Traffic

Safety Barrier, GL.9/2006

Statistical Report Road Accident, Royal Malaysian Police. (2007 to 2009). Get on 23
August 2010 from Kangar Traffic Police Station.

Statistical Report Road Accident, Royal Malaysian Police. (2007 to 2009). Get on 06
September 2010 from Muar Traffic Police Station.

Statistical Report Road Accident, Royal Malaysian Police. (2007 to 2009). Get on 20
September 2010 from Melaka Contingent Police Headquarter.
Table A. Top-Fourty Accident Ranking by Weightage (Guardrail)
Serious Slight Damage
Injury Injury Only
205 4 0 2 119 125 49.4 3.89 1 5
195 2 0 2 145 149 42.6 3.75 2 5
196 3 1 3 56 63 34.6 3.54 3 5
185 3 0 3 30 36 26.4 3.27 4 4
184 2 3 0 24 29 25.8 3.25 5 4
197 2 0 4 47 53 24.6 3.20 6 5
202 1 2 3 40 46 22.4 3.10 7 5
10 1 5 0 4 10 21.8 3.08 8 1
199 2 0 4 31 37 21.4 3.06 9 5
217 0 0 0 97 97 19.4 2.96 10 5
209 2 0 0 37 39 19.4 2.96 11 5
207 2 0 1 28 31 18.4 2.91 12 5
179 2 0 1 27 30 18.2 2.90 13 4
210 0 0 4 71 75 17.4 2.85 14 5
198 2 0 0 27 29 17.4 2.85 15 5
178 2 0 0 25 27 17.0 2.83 16 4
3 1 2 3 12 18 16.8 2.82 17 1
181 1 2 1 17 21 16.2 2.78 18 4
221 0 0 0 77 77 15.4 2.73 19 5
11 2 0 1 13 16 15.4 2.73 20 1
189 1 1 2 20 24 14.6 2.68 21 4
191 1 1 4 11 17 14.4 2.66 22 4
4 0 3 4 11 18 14.4 2.66 23 1
193 1 0 1 36 38 14.0 2.63 24 5
204 1 0 2 30 33 13.6 2.61 25 5
201 1 0 1 31 33 13.0 2.56 26 5
211 1 0 0 33 34 12.6 2.53 27 5
220 0 0 1 59 60 12.6 2.53 28 5
182 1 1 0 16 18 12.2 2.50 29 4
8 2 0 0 1 3 12.2 2.50 30 1
194 1 0 1 26 28 12.0 2.48 31 5
187 1 0 1 23 25 11.4 2.43 32 4
218 0 0 0 56 56 11.2 2.41 33 5
216 0 0 0 53 53 10.6 2.36 34 5
219 0 0 0 52 52 10.4 2.34 35 5
212 0 0 0 51 51 10.2 2.32 36 5
5 1 0 1 17 19 10.2 2.32 37 1
203 0 0 2 41 43 9.8 2.28 38 5
200 1 0 1 13 15 9.4 2.24 39 5
174 0 0 4 23 27 7.8 2.05 40 4

Table B. Top Fourty Accident Ranking by Weightage (Wire Rope)
Serious Slight Damage
Injury Injury Only
0 6 2 8 70 86 62.4 4.13 1 3
16 0 3 0 21 24 13.2 2.58 2 2
8 1 1 0 11 13 11.2 2.41 3 3
7 1 0 1 9 11 8.6 2.15 4 3
2 0 0 0 39 39 7.8 2.05 5 3
5 0 1 2 15 18 7.6 2.02 6 3
35 1 0 0 1 2 6.2 1.82 7 2
21 0 1 1 10 12 5.8 1.75 8 2
9 0 0 1 17 18 4.2 1.43 9 3
3 0 0 1 16 17 4.0 1.38 10 3
18 0 0 2 12 14 4.0 1.38 11 2
22 0 1 0 5 6 4.0 1.38 12 2
14 0 1 0 2 3 3.4 1.22 13 2
17 0 0 1 12 13 3.2 1.16 14 2
37 0 1 0 1 2 3.2 1.16 15 2
27 0 1 0 0 1 3.0 1.09 16 2
1 0 0 0 12 12 2.4 0.87 17 3
6 0 0 0 11 11 2.2 0.78 18 3
31 0 0 0 9 9 1.8 0.58 19 2
11 0 0 0 7 7 1.4 0.33 20 3
43 0 0 1 2 3 1.2 0.18 21 2
29 0 0 0 5 5 1.0 0.00 22 2
36 0 0 1 1 2 1.0 0.00 23 2
23 0 0 1 0 1 0.8 -0.22 24 2
25 0 0 0 4 4 0.8 -0.22 25 2
33 0 0 0 4 4 0.8 -0.22 26 2
4 0 0 0 3 3 0.6 -0.51 27 3
30 0 0 0 3 3 0.6 -0.51 28 2
40 0 0 0 3 3 0.6 -0.51 29 2
20 0 0 0 2 2 0.4 -0.91 30 2
28 0 0 0 2 2 0.4 -0.91 31 2
10 0 0 0 1 1 0.2 -1.60 32 3
12 0 0 0 1 1 0.2 -1.60 33 2
13 0 0 0 1 1 0.2 -1.60 34 2
15 0 0 0 1 1 0.2 -1.60 35 2
19 0 0 0 1 1 0.2 -1.60 36 2
24 0 0 0 1 1 0.2 -1.60 37 2
34 0 0 0 1 1 0.2 -1.60 38 2
38 0 0 0 1 1 0.2 -1.60 39 2
39 0 0 0 1 1 0.2 -1.60 40 2


Cik Wan NORBALKISH Jusof 1, ISMAIL Yusof 2 and Mohd ERWAN Sanik 3

1, 2, 3 Universiti Tun Hussein Onn Malaysia, Johor, Malaysia

ABSTRACT: Road accident is an unfortunate event which is a matter of serious concern to

the authority. An accident reduction measure taken in reducing the rate of accidents is by
identifying hazardous locations for treatment. The purposes of this study are to determine
the blackspot location along Federal Route 1, Federal Route 5 and Federal Route 24 in the
Batu Pahat area and to develop an Accident Prediction Model by using multiple regression
method as well as to carry out sensitivity analysis on the developed model. This study is
limited to the two-lane federal road type and therefore multilane highway is not taken into
account. Data for explanatory variables such as volume, speed, and road geometry are
obtained using the Metrocount equipment and field measurements. Accident point weightage
calculation is carried out in order to determine road accident trends and blackspot ranking
using accident data provided by the Royal Malaysian Police. Using these data, an accident
prediction model is successfully developed. The results from sensitivity analysis revealed
that the increment of accident point weightage can be explained by either the rise in speed,
number of major access point, and traffic volume. Meanwhile, an increment of lane width
and shoulder width will reduce the weighting point rates.
Keywords: Accident Point Weightage, Blackspot ranking, Sensitivity analysis


One of comprehensive definitions of road traffic accident is a rare, random,

multi-factor event always preceded by a situation in which one or more road
users have failed to cope with their environment, resulting in a collision on the
public highway which should be recorded by the police (Mohamed Shafii et al.,
1995). The first death involving a motor vehicle is said to have taken place in
London in 1896 (PIARC, 2003). According to the World Health Organization,
every year, nearly one million people killed, three million are severely disabled
for life and thirty million are injured in road accidents. Road accidents lay in the
ninth place out of a total of over a hundred separately identifiable causes of
death or disability (Murray et al., 1996). And, by the year 2020, it has been
predicted that road accidents will be the third leading cause of death worldwide
(TRL, 1998).

It is possible to identify hazardous sections of the road network so that

appropriate remedial measures can be undertaken to reduce the likelihood and
severity of accidents at those locations. The term ‘blackspot’ is said to derive
from the method that was originally used to identify hazardous sites. The terms

‘hazardous location’ and ‘high accident location’ often used as a synonym to
blackspot. A blackspot was originally a road location of limited area with a high
concentration of accidents. This definition has progressively evolved as several
researches now recommend including the concept of potential for improvement
(McGuigan, 1981; Elvik, 1988; Hauer, 1996).

An Accident Prediction Model (APM) is a mathematical equation that

expresses the average accident frequency of an entity as a function of traffic
flow and other road characteristics. It is important to examine the nature of
relationships between roadway, environmental and operational factors and,
accidents to understand the causal mechanisms involved in accidents on the
one hand and to better predict their occurrence on the other hand. APMs are
one path of inquiry often used to gain these insights (Reurings et al., 2005). The
Multiple Linear Regression is a statistical methodology describing relationships
between a continuous outcome and a set of explanatory variables (Kutner et al.,
2005). This technique is appropriate to describe relationships between
continuous accident outcomes and explanatory variables. A comprehensive
study of road safety (Treat et al., 1977) found that human error was the sole
cause in 57% of all accidents and was a contributing factor in over 90%. In
contrast, only 2.4% were due solely to vehicle and only 4.7% were caused only
by road and environment factor. Based on categorical representation of lane
width, for two-lane rural, two-lane urban, four lane urban undivided, and urban
freeways, widening lane width up to 4.0m, 3.7m, 4.0m and 4.0m respectively
could be expected to decrease crash rates (Hadi et al., 1995). In Australia,
excessive speed is an important factor in approximately 20 percent of fatal
crashes (Haworth et al., 1993) and speed is a probable or possible cause in 25
percent of rural crashes (Armour et al., 1990). Meanwhile, roads with higher
ADT and pedestrian traffic are associated with higher accident frequencies for all
highway types (Berhanu, 2004).

Four main stages in the methodology are data collection, analysis of accident
data and identification of blackspot location, field work, and the development of
an accident prediction model.

B.6 Data Collection

In this study, data collection involves two types of data namely primary data
and secondary data. Secondary data are accident data obtained from the
Royal Malaysian Police (RMP). Accident data was collected from the Batu
Pahat Traffic Police Station, Batu Pahat Public Work Department, Bukit
Aman RMP Headquater and Jabatan Keselamatan dan Jalan Raya (JKJR).
The accident database is extracted from the POL27 and the accident record
collected from the year 2007 until May 2010. Meanwhile, primary data
includes geometrical data and field work data, specifically from field
observation and Metrocount data.

B.7 Analysis of Accident Data

The purpose of preliminary analysis of the traffic accident data is to
determine the broad nature of the accident problem. At this stage, the
general patterns and trend of accident are established. Types of trend and
statistics that take into account for the Federal Route 1 Johor Bahru – Labis
(FT001), Federal Route 5 Johor Bahru – Melaka (FT005) and Federal Route
24 Yong Peng – Muar (FT024) are as follow:

j. Road accident statistics for year 2007 to May 2010

k. Accident and casualties for year 2007 to May 2010
l. Accident by hours of the day for year 2007 to May 2010 (every two

B.8 Identification of Blackspot Location

Identification and prioritization of accident blackspot location is carried out
through ranking by accident point weightage.

B.9 Field Work

The purpose of field work is to collect primary data at selected 8 locations
from all route. Furthermore, the field works also important to document the
site findings and photographs taking as evidence. The road sections
selected for data collection are Km88, Km106 and Km112 at FT001, Km93
and Km 122 at FT005 and Km5, Km1 1 and Km27 at FT024. These locations
are selected based on the ranking of the Accident Point Weightage (APW).
The Metrocount equipments are installed for 12 hours at each location in
order to obtain data such as traffic volume and 85th percentile speed. Figure

1 depicts the Metrocount equipment used in the field work. The
measurements of lane width, shoulder with and number of access point are
carried out manually at the study locations. Figure 2 shows the odometer
used to measure the road geometry.

Figure 1. Metrocount

Figure 2. Odometer

B.10 Accident Prediction Model and Sensitivity Analysis

In this study, Microsoft Excel and Minitab statistical software are used in
analyzing all data and developing an APM, furthermore to validate the
coefficients and the model. The dependent variable for this model is the
APW and the explanatory variables are the 85th percentile speed, hourly
traffic volume, lane width, shoulder width, total number of access point, and
type of road geometric (straight, hilly, curvy). Later, the sensitivity analysis is
carried out on the model to determine the effect of each independent
variable on the dependent variable; for instance, the effect that changes in
speed will have on the APW.


The blackspot locations are determined with the ranking of the APW. The data
obtained from 9 selected locations which have been installed with the
Metrocount equipment are analyzed using the Microsoft Excel before preceded
with the development of the APM by using multiple regression method.
Furthermore, sensitivity analyses based on the developed model is carried out to
check the differences if changes in value made on the independent variables.

3.1 Road Accident Statistics at FT001, FT005 and FT024

Figure 3 shows the accident statistics at the FT001 (Km85-130), FT005

(Km88-151) and FT024 (Km 1-36) from year 2007 until May 2010. Referring
to the figure, the highest number of total accidents is recorded at the FT005
with 2774 cases followed by FT001 with 1033 cases and FT024 with 925
cases. Of these, 179 fatalities are recorded at all 3 routes in this period of
time. The most frequent accident is the damage only cases.

Figure 3. Accident Statistic for 2007-May 2010

3.2 Accident and Casualty

a) Accident and Casualty at FT001

In this study, records of FT001 involve Km85 to Km130 which is about 45

kilometers stretches. Figure 4 shows that in 2007 and 2008, the total
number of accidents recorded along FT001 is the highest with 317 cases in
each year. However, the total number of cases reduces in 2009 until May
2010. The total number of accidents recorded from 2007 until May 2010 at
the route is 1033 cases. Of these, 45 cases are fatal, 3 cases are serious
injuries, 63 cases are slight injuries and 922 are damage only cases.

Figure 4. Accident and Casualty at FT001 (2007-May 2010)

b)Accident and Casualty at FT005

Records of FT005 involve Km88 to Km151 which is about 63 kilometers

stretches. Figure 5 shows that in 2009 the total number of accidents
recorded along FT005 are the highest with 844 cases. The total number of
cases increased every year. The total number of accidents recorded from
2007 until May 2010 at this route is 2774 cases. Of these, 91 cases are
fatal, 31 cases are serious injuries, 357 cases are slight injuries and 2295
cases are damage only.

Figure 5. Accident and Casualty at FT005 (2007-May 2010)

c)Accident and Casualty at FT024

In this study, records of FT024 involve Km1 to Km36 which is about 36

kilometers stretches. Figure 6 shows that in 2009 the total number of
accidents recorded along FT024 are the highest with 295 cases. The total
number of accidents recorded from 2007 until May 2010 at the\is route is 925
cases. Of these, 43 cases are fatal, 8 cases are serious injuries, 77 cases
are slight injuries and 797 cases are damage only.

Figure 6. Accident and Casualty at FT024 (2007-May 2010)

3.3 Accident Data by Hours of the Day

Figure 7 through 9 show the total number of accident for every 2 hours
beginning midday to midnight throughout the year from 2007 to May 2010.
The analyses are carried out for every route and the largest number of
accidents recorded at 14:01 to 18:00 or 2:01 to 6:00 p.m. This pattern might
happened because the period especially 4:01 to 6:00 p.m. is the peak hours
where the density of traffic is high on roads and since the federal roads are
at the rural area, speeding may also being a factor contributes to accident

a)Accidents by Hours of the Day at FT001

Figure 7 shows that the highest number of accidents occurred at FT001

recorded from 16:01 to 18:00 with 150 cases followed by 139 cases at
14:01-16:00. The least number of accidents happened is at 04:01 to 06:00
with 29 cases are recorded.

Figure 7. Accident by Hours of the day at FT001

b)Accidents by Hours of the Day at FT005

Figure 8 shows that the highest number of accidents occurred at FT005

recorded from 14:01 to 16:00 with 339 cases followed by 328 cases at
12:01-14:00. The least number of accidents recorded is at 04:01 to 06:00
with 54 cases are recorded.

Figure 8. Accident by Hours of the day at FT005

c) Accidents by Hours of the Day at FT024

Figure 9 shows the highest number of accidents occurred at FT024 recorded

from 16:01 to 18:00 with 130 cases followed by 123 cases at 14:01-16:00.
The least number of accidents recorded is at 04:01 to 06:00 with 26 cases
are recorded.

Figure 9. Accident by Hours of the day at FT024

3.4 Accident Point Weightage Ranking

The calculation for APW is carried out using Equation (1) and ranked using
the Microsoft Excel.

APW = X 1(6.0) + X 2(3.0) + X 3(0.8) + X 4(0.2) (1)

Where, X 1, X2, X3 and X 4 are the number of fatal, serious Injury, slight injury
and damage only, respectively

3.5 Selected Study Location

Using Equation (1), APW for every kilometer in a route is calculated and the
sections are accordingly ranked. From rank 1 to 5 of every route, 8 sections
have been selected for further investigation. Table 1 shows the list of
selected sections in every route with regards to type of accident and APW

Table 1. 8 selected sections for FT001, FT005 and FT024

Section Serious Slight Damage
Route Rank Fatal number of APW
No. Injury Injury only
FT001 1 112 3 0 8 145 156 53.4
4 106 3 0 1 14 18 21.6
5 88 3 0 1 7 11 20.2
FT005 2 93 2 2 20 87 111 51.4
4 122 3 2 13 56 74 45.6
FT024 1 27 4 0 13 81 98 50.6
2 5 4 0 1 37 42 32.2
5 11 4 0 2 16 22 28.8

3.6 Number of Accidents at Study Locations

Figure 10 depicts the accident data at every selected section. Referring to

the figure, for the fatal accident, Km 5, Km 11 and Km27 from FT024
recorded the highest number with 4 cases at each location while the lowest
is recorded with 2 cases at Km93 (FT005). Furthermore, for the serious
injury accident, Km93 (FT005) and Km122 (FT005) record 2 cases at each
location while the other locations record none. Km93 (FT005) records the
highest number of slight injury accident with 20 cases while Km106 (FT001),
Km88 (FT001) and Km5 (FT024) record the lowest number with 1 case at
each location. The highest number of damage only accident is at Km1 12
(FT001) with 145 cases and the lowest is at Km88 (FT001) with 7 cases.
For overall, the location with highest total number of accidents is Km112
(FT001) with 156 cases while the lowest is Km88 (FT001) with 11 cases.

Figure 10. Accident statistic at study locations

3.7 Model Development

In this study, multiple regression analysis is conducted using the Minitab.

After considering the best fitted model and with transformation of level-level
linear regression to log-level linear regression, the APM that has been
developed is as follows:

ln (APW) = 0.000098(HTV) + 0.00469(85 th

PS) + 0.0476(AP) + 0.137(ST) –
0.4386(HI) – 0.170(LW) – 0.0602(SW) + 3.49

APW = accident point weightage
HTV = hourly traffic volume
85th PS = 85th percentile speed
AP = number of access points per kilometer
ST = straight road geometry

HI = hilly road geometry

LW = lane width
SW = shoulder width

a) From the analysis, the developed model has an R-squared of 0.994 and
adjusted R-squared of 0.993, which means 99.4% and 99.3% of the variation
in the number of accidents has been explained the regression line. Table 2
shows the range of explanatory variables values can be used for this model.

Table 2. Range of values for independent variables in Accident Prediction Model developed
Factor Range of Values
Hourly Traffic Volume (HTV) 494 - 2415veh/hr
85 th Percentile Speed (85 th PS) 60.8 – 98.3 km/hr
Number of access points (AP) 0 – 14
Lane width (LW) 3.4 – 4.2 m
Shoulder width (SW) 0.35 – 2.8 m

This study used the APW from the HPU to determine the black spot locations
before ranking them in one kilometer at each location. Further analysis of this
study is to develop predictive model relating accident weighting point with the
selected independent variables has been achieved. Multiple regression
techniques are used to build the model. The model is as shown in Section 3.8.
The result of the sensitivity analysis provide sufficient evidence to support the
hypothesis that the independent variables include Hourly Traffic Volume, 85th
Percentile Speed, total number of access points, lane width and shoulder width
are related to the accident weighting point. It can be conclude that for instance:

Doubling the hourly traffic volume is expected to cause a 26.7% increase

in percentage of the predicted APW.

A 10% increment of 85th percentile speed is expected to cause a 4.7%

increase in percentage of the predicted APW.

Widening to lane by 10% increment from the existing width is expected to

reduce the percentage of the predicted APW for about 6.9%.

Widening to shoulder by 10% increment from the existing width is

expected to reduce the percentage of the predicted APW for about 1.7%.

An increase of 5 point in approach number of access point is expected to

cause a 26.9% increase in the percentage of the predicted APW.

Finally, the developed Accident Prediction Model in this study is expected to

be applicable in road safety improvement especially to reduce the number of
accidents at two-lane federal roads by emphasizing into consideration the
factors that predict the model at the design and planning stages.


We would like to thanks the Batu Pahat Police Station and the Jabatan Kerja
Raya, Batu Pahat Branch for their cooperation in getting the data.

Armour, M. & Cinquegrana, C. (1990). Victorian study of single vehicle rural
accidents. Proceedings of the 15 th Australian Road Research Board Conference,
Part 7. pp 79-91. Australian Road Research Board.Melbourne

Berhanu, G. (2004). Model relating traffic safety with road environment and traffic
flow on arterial roads in Addis Ababa. pp 697-704. Adis Ababa University

Elvik, R. (1998). Some Difficulties in Defining Populations of ‘entities’ for Estimating

the Expected Number of Accidents. Accident Analysis and Prevention. V20,
N4, pp 261-275

Hadi Mohammed A., Jacob Aruldhas, Lee-Fang Chow, and Joseph A. Wattleworth
(1995). Estimating Safety of Cross-Section Design for Various Highway Types
Using Negative Binomial Regression. Transportation Research Record, 1500 ,

Hauer, E. (1996). Identification of Sites with Promise. Transportation Research

Record 1542 . pp 54-60

Kutner M.H., Nachtsheim C.J., Neter J. & Li W. (2005). Applied Linear Statistical
Models . 5th edition

McGuigan D.R.D. (1981). The Use of Relationship between Road Accidents and
Traffic Flow in ‘blackspot’ Identification. Traffic Engineering. V12, N1, pp 448-

Mohd Shafii, Subiah Sulaiman, Norliah Saidin, PM Radin Umar, R.S, P/PPP Ruslan
Khalid, PPP Ooi I.B, Ir Sabudin Mohd. Salleh, Sanusi Ismail (1995). Interim
Guide on Identifying, Prioritizing and Treating Hazardous Locations on Roads in
Malaysia . Kuala Lumpur: Percetakan Ibu Pejabat JKR Malaysia

Murray, C. & Lopez, A.D. (1996). The Global Burden of Disease . World Health
Organization, Harvard School of Public Health: World Bank, Boston.

Reurings M., Janssen T., Eenink R., (SWOV), Elvik R., (TOI), Cardoso J., (LNEC),
and Christian S., (KfV), (2005). Accident Prediction Models and Road safety
Impact Assessment . RI-SWOVWP23- R1 -V2

Ross Silcock Partnership (1991). Towards Safer Road in Developing Countries, a

Guide for Planner and Engineers . TRRL: First Edition

Treat, J.R., Tumbas, N.S., McDonald, S.T., Shinar, D., Hume, R.D., Mayer, R.E.,
Stanisfer, R.L. & Castellan, N.J. (1997). Tri-level Study of the Causes of Traffic
Accidents. Report No. DOT-HS-034-3-535-77 (TA C)

World Road Association- PIARC (2003). Road Safety Manual . Quebec.

Mohd Hanifi Othman 1, Zaiton Haron 2, Khairulzan Yahya 3, Haryati Yaacob 4 ,
Shamila Azman 5

Universiti Teknologi Malaysia, 81300, Skudai, Johor, Malaysia

ABSTRACT: Centerline Rumble Strips (CRS) is one form of traffic safety devices that work
to prevent the occurrence of head-on-crashes and opposite direction sideswiped crashes
caused by inattentive driver as a result from fatigue, sleepy or drowsiness factor. CRS are
installed extensively in developed countries such as USA, Japan and Canada in order to
reduce the accident rates. This paper discusses the installation of CRS practices in these
countries including type, design, and installation guidelines. This paper also reviews on
effectiveness of CRS in reducing head-on-crashes caused by inattentive drivers and the
application of the CRS. It was found that CRS could be introduced in Malaysia as it is not
only economical, but also very effective in reducing the head-on-crashes caused by the
inattentive driver. This is possibly a method that can be considered by the relevant
authorities in order to reduce the high accident rates in Malaysia.

Keywords: Centerline Rumble Strips, head-on-crashes, inattentive driver


Rumble strips has been used as part of traffic calming devices, as an early warning
to drivers to improve their driving and hence to reduce accident at hazardous
location. The most common device used in Malaysia is transverse rumble strips
(TRS) which is generally consist of yellow thermoplastic lines (3mm to 7mm thick)
laid across the carriageway. TRS designs are uniform for the entire country
according to specifications set in REAM -GL8/2004 (Guidelines on Traffic Control
Devices and Management), Part-4, pavement marking and Delineation. One form of
rumble strips that work to prevent the occurrence of head-on-crashes and opposite
direction sideswiped crashes caused is Centerline Rumble Strips (CRS). This an
unfamiliar feature here but they are installed extensively in developed countries
such as USA, Japan and Canada. The primary purpose of CRS (Figure 1.1) is to
warn inattentive drivers whose vehicles are crossing centerlines of two-lane, two-

way roadways to avoid potential crashes with opposing traffic. CRS give warning to
the drivers in the form of vibration and sound created when vehicles passess
through it. There are two types of crashes are considered correctable by CRS i.e
head-on collision (HOC) and opposite-direction sideswipes (ODS) which often
referred as cross-over or cross-centerline crashes.

Figure 1.1: Typical CRS (Hirasawa et al., 2005)

In Malaysia, the authorities sometimes use pavement marker for mitigating HOC
and ODSC accidents. Function of pavement marker is similar to CRS and its able to
deflect light for safety at night. However, according to Hirasawa et al (2005) the
costs for pavement marker are twice the cost of CRS. A typical pavement marker is
shown in Figure 1.2.

Figure 1.2 : Pavement Marker (Hirasawa et al., 2005)

This paper reviews the effectiveness of CRS in reducing accidents caused by

inattentive drivers and problems that arise by the application of the CRS. The
installation of CRS practices in these countries including type, design, and
installation guidelines are also highlighted.


HOC typically occurs when a vehicle crosses the median and crashes with a vehicle
traveling in the opposite direction (typically called a cross-median crash or median-
crossover crash) as shown in Figure 2.1. While ODSC is also involves collisions
from the opposite direction but with impact is on the surface of the vehicle side. A
HOC can occur when a vehicle inadvertently travels the wrong way in the opposing
traffic lanes. Neuman et al. (2008) also reported that HOC is typically the result of
inadvertent actions by a driver potentially in combination with other adverse
circumstances such as weather conditions or motorist fatigue.

Figure 2.1: Oncoming vehicle in wrong lane that can resulted HOC

According to a study in US back in 1990, head-on collisions were the cause of

approximately 40% of all fatal, multivehicle crashes. The study also showed that
more than 86% of fatal head-on collisions on two-lane highways were caused not by
a driver attempting to pass another vehicle but rather typically either by entering a
curve at too high speed or by drifting across the road after falling asleep or being
inattentive (Russell and Rys, 2005). Furthermore, Neuman et al. (2008) reported
that most head-on collisons are likely to result when a driver falls asleep, is
distracted, or travels too fast in a curve and majority of head-on fatalities on two-
lane rural roads are on tangent sections (63% vs. 37%).

Torbic et al.(2009) stated there are seven characteristic of typical crashes due to
drowsiness/fatigue; (i) Crashes occur during late-night hours—Most crashes
involving drowsiness/fatigue occur from midnight to the pre dawn hours with a small

peak in the middle of the afternoon. This is consistent with human sleeping patterns;
(ii) Crashes happen at high speed—Because more long trips occur on higher speed
roadways, there is likely a higher proportion of drowsiness/fatigue crashes on
roadways with speed limits of 55 to 65 mph (88 to 105 km/h); (iii) Crashes are likely
to be serious—Injury and fatality rates are higher for drowsiness/fatigue crashes
than with other types of crashes. The higher rates could be a factor of the crashes
happening at higher speeds; (iv) Single vehicle leaves the roadway—A majority of
drowsiness/fatigue crashes involve single vehicles leaving the roadway. Rear-end
crashes and head-on crashes may also be increased due to drowsiness and
fatigue.; (v) No attempt to avoid crashes—Evidence of avoiding actions such as skid
marks or brake lights are less likely in drowsiness/ fatigue crashes than in other
types of crashes; (vi) Driver is alone in vehicle—Drowsiness/fatigue crashes often
involve single-occupant vehicles.

According to Torbic et al. (2009) there are three populations that have higher risk of
being involved in a drowsiness/fatigue-related crash: (i) Young people, especially
young men—Drivers under age 30 are four times more likely than other drivers to be
involved in a drowsy-driving crash. Men are five times more likely than women to be
involved in a drowsy-driving crash; (ii) Shift workers—The information concerning
shift workers and drowsiness/fatigue-related crashes has come from self reporting
and interviews rather than crash reports; but due to changing sleeping patterns, loss
of sleep, and more driving done in the early morning hours, it is assumed that there
is a greater risk of drowsiness/fatigue crashes among shift workers; (iii) People with
untreated sleep apnea syndrome and narcolepsy—The total number of
drowsiness/fatigue crashes involving drivers with sleep disorders is low, but the risk
is higher among drivers with untreated sleep disorders than among other drivers.


The most significant effectiveness of CRS had been recorded by the Delaware
Department of Transportation, US. From their investigation on HOC three years of
before and six years of after installation of CRS on Route US 301, showed that CRS
had decreased crashes by 90 percent. Crashes caused by drivers crossing over the
centerline rumble strips decreased by 60 percent. The benefit/cost ratio for CRS
installations was estimated to be 110. Investigation in California on the effects of

CRS in no-passing zones showed that crash data from three years of before and
after CRS installation had reduced by 11 percent and fatalities reduced by 77
percent. However, the situation is different in Minnesota where they installed CRS at
two sites on rural roads with 55 mph speed limits. A review of three years of before
and after crash data found no reduction in HOC (Russell and Rys, 2005).

Table 3.1 shows the numbers of head-on collisions in 2002 and 2003, before and
after installation of center rumble strips at 24 locations in Hokkaido, Japan. The
number of accidents for “before” is that for the two years before installation and the
number of accidents for “after” is from the day of installation to December 31, 2004.
HOC before installation was numbered 42 at 24 locations, and after installation was
numbered 15 respectively. To compare the accident rate of before and after, the
figure for after was extrapolated, which yielded a figure of 18.8 collisions. The
accident reduction rate was calculated to be 55.2% (Hirasawa et al., 2005). On the
other hand, Russell and Rys (2005) thought that crashes in reality are a rare
occurrence and, for a given roadway segment, several years of data or several
combined databases are generally required to apply sophisticated statistical
techniques that produce statistically significant results indicating that a treatment
resulted in an effect caused by the treatment and not by chance.

Table 3.1: Number of accidents before and after installation of CRS in Hokkaido, Japan
(Hirasawa et al., 2005).
Route Length Date of Number of HOC
installed construction
(m) Before From Extrapolated
installation installation after
(2 years) to Dec. 1, installation
2004 (2 years)

1 5 727 2002/2/22 1 0.0

2 274 2708 2002/11/6 3 3 3.0
3 5 457 2002/12/10 1 0.0
4 37 6197 2003/5/13 1 1 1.2
5 5 1500 2003/5/26 3 0.0
6 40 1178 2003/6/2 2 0.0
7 274 2860 2003/6/9 3 0/0
8 274 5050 2003/6/16 7 6 7.8
9 274 3815 2003/6/23 3 3 3.9
10 5 1507 2003/7/1 3 1 1.3
11 275 730 2003/7/8 1 0.0
12 39 1100 2003/7/22 1 0.0
13 44 400 2003/7/25 0 0.0
14 230 1057 2003/7/30 1 0.0
15 5 300 2003/8/5 1 0.0
16 230 1057 2003/8/26 2 0.0
17 5 522 2003/9/22 1 1 1.6
18 276 3448 2003/9/4 2 0.0
19 5 600 2003/9/3 1 0.0
20 303 440 203/9/16 0 0.0
21 5 442 2003/10/2 0 0.0
22 40 382 2003/10/22 1 0.0
23 236 200 2003/10/27 1 0.0
24 38 721 2003/11/1 3 0.0
Total 39284 42 15 18.8
Reduction rate for HOC: (42-18.8)/42*100=55.2%

Hirasawa et al. (2005) concluded in their study that the CRS has the following
advantages:- (i) A high degree of warning is given to drivers who deviate to the edge
of the road. (ii) Two-wheel vehicles can travel more safely on sections with CRS
than on those with center poles or chatter bars. (iii) Rumble strips do not hinder
snow removal. (iv) The costs are low (half of that for center poles, and one-third of
that for chatter bars). (v) Because the rumble strips are not installed where the
wheels of vehicles pass, they cause very little tire abrasion and they do not affect
the traveling speed and (vi) They can be highly effective against head-on collisions
in cold, snowy regions like Hokkaido. Russell and Rys (2005) also reported a study
in Pennsylvania found that there was a movement of vehicles away from the CRS

on the order of several inches and thus would increase the separation of opposing
vehicles and potentially increase safety.

Anunda et al. (2008) however showed that CRS cannot be really effective and there
are no effects observed due to type of strip. This is because the effects of
drowsiness or sleepy which are missing as a result of vibration and noise can be
reappear during 5 min after hitting CRS.


CRS are installed in the form of milled, rolled, and raised (Finley and Miles, 2007).
Milled rumble strips are installed on both new and existing concrete and asphalt
roadways by a mechanical milling device while rolled rumble strips are depressions
on asphalt roadways formed by steel pipes welded onto a roller at a uniform spacing
applied only on new or reconstructed asphalt surfaces. Raised rumble strips are
rounded or rectangular markers or strips adhered to the roadway including traffic
buttons, profile markings, preformed thermoplastic, or raised sections of asphalt
pavement. Raised rumble strips can be applied to any roadway; however, typically
they are restricted to warmer climates because cooler climate regions may require
snowplowing that may damage the rumble strips and/or the snowplowing equipment
(Makarla, 2009).

In order to tackling the problem of HOC and ODSC accidents, CRS is located in the
middle of the road that separates the opposite direction as shown in Figure 4.1.
However, CRS dimension size is different because in general, the transportation
departments use the size of its own design. Figure 4.2 (a) and (b) shows an
example of typical dimensions used CRS in the USA.

Figure 4.1: Typical CRS installation (Torbic et al., 2009).


Figure 4.2: Typical design of CRS in USA (Torbic et al., 2009)

Apart from rectangular shaped, CRS can also be found in football shaped as
shown in Figure 4.3 (Gardner et al., 2007).

Figure 4.3: Rectangular and football CRS (Gardner et al., 2007)


In general, each place has its own installation guidelines as a result from
investigation by their transportation agency. However, commonalities in the
guidelines are as follows: 1)Roadway type—Rural two- or three-lane undivided;
2)Crash history—All documents address numbers of crashes or crash rates and
indicated that CRS should be used on sections where some number (unspecified) of
cross-over crashes have occurred (Russell and Rys, 2005).

Only California policy contains a specific, weighted average number based on a
value for various levels in five categories: 1) number of total crashes, number of
deaths, fatal accident rate, death rate, and total accidents per mile. The five
categories are summed at a value of 40 (plus a cross-over fatality during 1998)
which triggers an investigation of the site;1) Speed—50 mph or greater;2) ADT
threshold—1,500 to 3,500; 3) Lane width—No less than 10 ft, with 11 or 12 ft more
common;4) Pavement type—Primarily asphalt in good condition with minimum
depths of 2.5 in. to 2.75 in.;5) Noise—Consider noise; and 6) Coordination—
Coordinate with all other project tasks and install CRS last (Russell and Rys, 2005).


The cost per linear foot of roadway for CRS installation varied from $0.20 to $0.60,
excluding traffic maintenance. This has led Delaware Department of Transportation
(Del-DOT) to claim a benefit–cost effectiveness of 110 to 1 this means that for
every dollar Del- DOT spent, society has saved $110 in medical costs, insurance,
legal fees, long-term care, property damage, and other expenses related to serious
accidents and fatalities (Gardner et al., 2007).


There are problems regarding CRS arising from other road users such as bicyclists.
Bicyclists claim vehicles on roads with CRS hesitate moving left and create a
potentially dangerous situation for them. Bicyclists are particularly concerned on
winding roads and roads with no shoulder. This effect needs more study and
certainly should be considered in any decision to use CRS (Russell and Rys, 2005).

In addition, the sound produced by the CRS for alerting drivers can also cause
external noise toward roadsides and its effect on roadside residences should be
considered. Many responses noted it should be considered, but no definite numbers
were presented except in Transportation Association of Canada (TAC) report that
states CRS terminated 200 m before residential or urban areas produce tolerable
noise impacts on residents and that at 500 m the noise is negligible (Gardner et al.,
2007). In Sweden, there is a rule rumble strips should not be used within a distance

under 100 m to areas where a maximum of 70dB (A) was imposed (Breyer, 2010).
Other methods to control noise of CRS are through the dimension and type of CRS
itself. Kansas Transportation Department conducted a study on centerline rumble
strips focusing on how states were constructing and placing milled centerline rumble
strips, and the associated noise and vibration produced by different rumble strip
patterns. A study reported the relationship between the dimensions and produced
sound which are as follows: (i) A one unit increase in the rumble strip depth (in.) is
associated with a 4.494 dBA increase in the sound level difference. (ii) A one unit
increase in the rumble strip spacing (in.) is associated with a 0.394 dBA decrease in
the sound level difference. (iii) Milled rumble strips are associated with a higher
sound level difference when compared to rolled rumble strips. (iv) A concrete
pavement surface is associated with a lower sound level difference when compared
to an asphalt pavement surface. (v) A wet pavement surface is associated with a
lower sound level difference when compared to a dry pavement surface (Torbic et
al., 2009). With this, the dimensions of CRS can be adjusted to the distance from
CRS location and residential areas.


From this review, it was found that the advantages of CRS out weight its
disadvantages. CRS is simple and inexpensive to install, effective in reducing HOC
and ODSC accidents, and has a high ratio of benefit/cost. While its disadvantages,
such as creating danger to bicyclists, and noise problems to surrounding residents
can be overcome from time to time. Thus, CRS is now increasingly used in
developed countries like USA, Japan and Canada. Therefore, in order to tackle the
high accidents rates problems particularly in term of HOC and ODSC accident
caused by inattentive driver for this country, the relative authorities should start to
use SRS as one of the traffic safety devices.


Special thanks to Ministry of Science and Technology (MOSTI), UTM Research

Management Centre and also Universiti Teknologi Malaysia for funding this
research under grant Vote EScience No. 79370.


Anunda, A., Kecklunda, G., Vadebyb, A., Hjälmdahlb, M. & Åkerstedta, T. 2008. The
alerting effect of hitting a rumble strip—A simulator study with sleepy drivers.
Accident Analysis and Prevention, 40, 1970–1976.

Breyer, G. 2010. Shoulder & Median Rumble Strips . Conference of European

Directors of Road. Paris

Gardner, L. W., Rys, M. J. & Russell, E. 2007. Comparison of football shaped

rumble strips versus rectangular rumble strip . Kansas Department of
Transportation, Kansas State University, University of Kansas. Report No. K -
TRAN : KSU-00-4P2.

Hirasawa, M., Motoki, A. & Saito, K. 2005. Study on development and practical use
of rumble strips as a new measure for highway safety. Journal of Eastern
Asia Society for Transportation Studies, 6, 3697-3712.

Neuman, T. R., Nitzel, J. J., Antonucci, N., Nevill, S. & Stein, W. 2008. Guidance for
Implementation of the AASHTO Strategic Highway Safety Plan: Volume 20:
A Guide for Reducing Head-On Crashes on Freeways . Transportation
Redearch Board. Washington DC
Russell, E. R. & Rys, M. J. 2005. Centerline Rumble Strips . Transportation
Research Board. NCHRP Synthesis 339. Washington D.C

Torbic, D. J., Hutton, J. M., Bokenkroger, C. D., Bauer, K. M., Harwood, D. W.,
Gilmore, D. K., Dunn, J. M. & Rochento, J. J. 2009. Guidance for the Design
and Application of Shoulder and Centerline Rumble Strips . Transportation
Research Board. NCHRP Report 641. Washington D.C

Mohd Hanifi Othman 1, Zaiton Haron 2, Khairulzan Yahya 3, Haryati Yaacob 4 ,
Shamila Azman 5

Universiti Teknologi Malaysia, 81300, Skudai, Johor, Malaysia

ABSTRACT : Malaysia is among highest accident rates countries if measured from the
aspect ratio of accidents per capita of population. Among the main causes of accidents are
fatigue and drowsiness experienced by driver, especially during long term and monotonous
journey. One approach taken by other countries to solve the problem is Shoulder Rumble
Strips (SRS). Shoulder Rumble Strips is one form of traffic safety devices that work to
prevent the occurrence of single vehicle run of road (SVROR) accidents caused by
inattentive driver as a result from fatigue, sleepy or drowsiness. This paper discusses the
installation of SRS practices in USA. The reviews focus on type, design method, installation
guidelines and performance of SRS in reducing SVROR type of accident. Other related
issues that arise by the application of SRS are also discussed. It was found that SRS is an
economical and effective approach and it should be introduced in Malaysia in order to
reduce the number of SVROR accident.
Keywords : Shoulder rumble strips, single vehicle run of road (SVROR) accident, inattentive

Road traffic accidents which involves cars, motorcycles and public transport
is a major problem in Malaysia. According to the Malaysian Institute of Road
Safety (MIROS), the rate of road accidents in Malaysia is one of the highest
compared to other countries in the world. A ten-year road traffic statistic
showed that the total number of accidents had raised from 215,632 cases in
1997 to 363,314 cases in 2007. This is equivalent to 3.73 deaths for every 10
000 registered vehicles in the same year (Kee et al., 2010). Based on the
accident data collected by Malaysian traffic police, SVROR accidents
constituted a significant proportion of serious injury cases (25.7%) and fatal
cases (38.6%) of the total road traffic collisions in 2006 (Abidin et al., 2009).
SVROR occurs when vehicles are likely to leave the road and collide with
objects (poles, posts, trees, walls, bridge supports or fences) within the road

shoulder or could involve contact with an embankment, slope or ditch (Abidin
et al., 2009). According to research, SVROR occurred due to factors of
sleepiness, fatigue and drowsiness and the factors contributed were
extremely high proportion which is 91.2% (Liu and Subramanian, 2009).

In USA, at least 100 000 accidents were reported in 1995 due to drowsiness
and fatigue. According to Perillo (1998) sleepiness reason when driving
caused approximately 1500 deaths and 71000 injuries per year in USA.
Drowsiness impairs driver’s performance by changing reaction times and
even small changes in reaction time can cause serious consequence,
particularly at higher speeds (Torbic et al., 2009). SRS a one method which
have been increasingly used by some developed countries to prevent driver
drifting their vehicles from travel lane as a result from inattentive, distracted,
drowsy, or fatigue. SRS was firstly used in New Jersey, also dubbed as the
singing shoulder as it generated sound and vibration when it is crossed by
vehicles which can alert an inattentive driver to correct their trajectories
(Perillo, 1998). However, SRS should not be expected to function as a
solution from the following: vehicles mechanical problems; evasive
maneuvers to avoid objects in travel lane; or driver error due to medical
condition (e.g., heart attack or seizures) (Torbic et al., 2009). Currently,
Canada, Japan, New Zealand and several other countries have also begun
to install SRS extensively.


Two types of SRS have been identified depending on its pattern on

pavement; sank and raised. Sank SRS are produced using "milled-in” and
"rolled-in" methods. The first design (milled-in) is made by cutting (or
grinding) the pavement surface with carbide teeth affixed to a 24 inches (600
mm) diameter rotating drum as seen in Figure 2.1 and Figure 2.2. The latter
(rolled-in) are by compacting a steel wheel roller (half sections of metal pipe
or solid steel bars are welded) into hot asphalt surface (Carlson and Miles,
2003, Hirasawa et al., 2005). SRS milled and rolled are illustrated in Figure
2.3 and 2.4.

While raised type RS patterns include “formed” and “raised”. Installing
method for “formed” SRS is by adding the formed SRS into Portland Cement
Concrete (PCC). Perillo (1998) stated that “formed” SRS were installed in the
late 1980's in several locations by being formed in the surface of fresh PCC
but is no longer a popular choice now due to the commonly used asphalt
shoulder. Instead, "Raised” rumble strips is the most commonly found in
Malaysia but in form of Transverse Rumble Strips. The design can be made
from a wide variety of products and installed using several methods. Carlson
and Miles (2003) adds this type may consist elements such as pavement
markers, a marking tape affixed to the pavement surface or asphalt material
placed as raised bars on the shoulder surface. In Malaysia, the materials
typically used are thermoplastic. "Raised" SRS is only suitable for places with
no snow. If this type of RS installed in snowy places, it may be damaged by a
snowplow blade passing over the SRS (Carlson and Miles, 2003). Some
researchers speculate that freeze-thaw cycle as a result from water collects
in the SRS grooves might damage the pavement. Once again, Carlson and
Miles (2003) noted that: “Tests show that vibration and the action of wheels
passing over the rumble strips in fact knock debris, ice, and water out of the
grooves”. Figure 2.5 shows example of raised SRS used on asphalt

Figure 2.1: Milling process to install “milled-in” SRS (Hirasawa et al., 2005)

Figure 2.2: Machine to install “milled-in” SRS (Hirasawa et al.,2005)

Figure 2.3: Milled SRS(Hirasawa et al .,2005) Figure 2.5: Rolled SRS with the roller device
(Morena, 2002)

Figure 2.4: Raised SRS (The Southland Times,2010)


In the USA, there are no standard design guidelines for SRS in national level,
however every state has their own design guidelines based on their own
studies. Typical design of SRS can be seen in Figure 3.1 and Table 3.1.
According to Torbic et al. (2009), dimensions shown in Figure 3.1(b) are as
follows: (i) Offset (A) - Lateral distance from the edge of the travel way to the

inside edge of the SRS; (ii) Length (B) - Dimension of the SRS measured
lateral to the travel way. Sometimes also known as transverse width; (iii)
Width (C) - Dimension of the TRS measured parallel to the travel lane; (iv)
Depth (D) - the vertical distance measured from the top of pavement surface
to the bottom of a SRS which refers to the maximum depth of the cut or
groove. (v) Spacing (E) - distance measured between SRS patterns.
Typically this dimension is measured from the center of one rumble strip to
the center of the adjacent rumble strip, or it could be measured from the
beginning of one rumble strip to the beginning of the adjacent rumble strip;
(vi) Recovery Area (F) - Distance from the inside (i.e., left) edge of the rumble
strip to the outside edge of the shoulder. The recovery area can also extend
beyond the edge of the shoulder to the nearest roadside object; (vii) Gap (G)
- Distance, measured parallel to the roadway, between groups of rumble strip
patterns. Gaps are designed primarily to allow bicyclists to navigate to the
other side of the rumble strip pattern without having to encounter a rumble
strip; (viii) Height (H) - This dimension is not depicted in Figure 3.1 , but it
refers to the vertical distance measured from the pavement surface to the top
of a raised rumble strip. This dimension corresponds to the depth dimension
of milled, rolled, and formed rumble strips. The road pavement must be thick
enough to be installed SRS. New York Department of Transportation set the
minimum thickness is 60 mm thick (Perillo, 1998); (viv) Lateral Clearance (I) -
Distance from the outside (i.e., right) edge of the rumble strip to the outside
edge of the shoulder. This is the portion of the shoulder to the right of the
rumble strips available for bicyclists to ride along the shoulder without
encountering the rumble strips. The lateral clearance can also be measured
to the nearest roadside object rather than the outside edge of the shoulder;
(x) Departure Angle ( α) - Angle at which a motor vehicle departs from the
roadway. This angle is a function of the steering angle and the curvature of
the roadway. Design of SRS is not necessarily in rectangular shape, SRS
also can be found in football shapes as shown in Figure 3.2.


Figure 3.1: Typical design of SRS (Torbic et al., 2009)

Table 3.1: Typical measurement of SRS dimension (2000)

Dimension Name Measurement (mm)
A Offset 150-600
B Length/Transverse 400
C Longitudinal Width 180
D Depth 13

Figure 3.2: Football shaped SRS (Gardner et al., 2007)


There are a number of concerns regarding the installation of SRS. SRS
should be installed for new asphalt with a minimum thickness of 1 3/4 in. on

top of a solid base with posted speed limit of 50 mph or greater. SRS should
not be installed if any of the following conditions exist: (i) The asphalt is old
and appears vulnerable to raveling if the grooves are cut; (ii) Shoulder
thickness is less than 3 3/4 in. (Minessota Department of Transportation,


Since the past ten years many countries including USA are actively using the
SRS based on finding from several researches that highlighted the
effectiveness of the SRS. The researches came to conclusion that SRS is an
effective traffic safety device. There is a significant reduction in the total
number of SVROR accident from the period 1991-1993(before/during
installation) as compared to the period 1996-1997(after installation). The New
York State Department of Transportation reported that their 5071 shoulder-
kilometers of SRS have reduced SVROR accidents due to driver inattention,
fatigue, and/ or drowsiness by 65 percent. The effectiveness of SRS can be
explained from Table 5.1 and Figure 5.1.
Table 5.1 : Before-After Data for Installation of Shoulder Rumble Strips in
New York State (NYSTA) (Perillo, 1998)

Figure 5.1 : SVROR accident reduction from 1991 to 1997 in New York

State (NYSTA) (Perillo, 1998)

In Finland, SRS was found to improve the visual guidance of drivers.

Vehicles are more likely to stay away from the barrier line therefore reducing
the risk crossing over barrier line (Rasanen, 2005).


Cost of installation of milled SRS reported by the New York State Thruway
Authority is shown in Table 6.1. These costs are dramatically declining over
the years due to the increased of numbers of installation and technological
advances. The cost of installation is also depends on the size of a project.
The bigger the project could cost for a meter SRS decreases. Therefore, the
reduction of prices was due to the increasing of large scale of SRS

In terms of benefit-cost ratios, Virginia Department of Transportation reported

that the benefit/cost ratio for SRS was +45. Meaning that every dollar
invested will return benefit of $45. According to Federal Highway
Administration (FHWA), the benefit/cost ratio were between +60 to +128
from the use of SRS to reduce SVROR accidents (Neumann and Richard,
2002). Based on research in the state of New York, Perillo (1998) reported
that total benefit/cost ratio for the SRS is +182.

Table 6.1: Milled SRS installation cost (Perillo, 1998)


RS also have some risks causing accidents to bicyclists. Researchers has been
trying to find a solution to this problem by finding the size of RS based on principle
providing a minimum of vibration to the bicyclist, but at the same time able to alert
inattentive or drowsy drivers. The researchers recommended the adoption of two
new “bicycle-tolerable” rumble patterns, one for non-freeway facilities operating near
55 mph (89km/hour) and the other for those operating at 45 mph (72 km/hour)
(Carlson and Miles, 2003).

Alaska Department of Transportation and Public Facilities took an approach by

providing a route for the bicyclist when passing SRS (Figure 7.1). Figure 7.2 shows
the SRS with the same approach in Kansas. The design may not be applied to the
entire roadway, but only to the roadway that is often used by bicyclists (FHWA,

Figure 7.1: Design that takes into account the needs of bicyclist to cross the
SRS (FHWA, 2000)

Figure 7.2 : SRS in Highway 36 in western Kansas (a suggested cross-state bicycle
route)(Kansas Cyclist, 2010)


Sound from the SRS is considered necessary to alert drivers but it also
annoying to local residents. SRS installed have caused complaints from
localities. The noises are more significant at night due to lack of background
noise (Estrada, 2008, November 9). Finley and Miles (2007) suggested that
velocity of the vehicle, size of vehicle, and type of rumble strips (RS) are
factors that influence the magnitude of exterior noise level produced by RS.
They found RS generated additional exterior noise by 13 percent more than
the highest noise level measured on smooth roads. They also reported that
exterior noise is higher at high speed and lower at low speeds (70mph and
55mph speed has been used in the study). Regarding vehicle size, they
found that heavier vehicle (commercial van as a sample) generated less RS
exterior noise compared to lighter vehicle (sedan car as sample). The results
of the study also supported the claim made by Perillo (1998) and Makarla
(2009) that "milled" and "rolled" RS produces the most powerful sound.
Finley and Miles (2007) also stated the change in the exterior noise level was
also affected by the pavement type (chip seal versus hot-mix asphalt.

Bendtsen et al. (2004) however suggested that condition of stripes may also
have influenced sound generated. He found that the RS noise level can be
lowered by the followings: (i) the distance between the individual stripes is
increased; (ii) the width of the individual stripes is decreased; and (iii) the
thickness of the individual stripes is reduced. The Transportation Association
of Canada proposed that SRS must be located at least 200m from residential
areas. Miska (2009) come up with more strict suggestion by insisting
distance between rumble streets to nearby residences must be 500 m in rural
environment and 200 m in urban environment.


SRS is one form of traffic safety devices that work to prevent the occurrence
of SVROR accidents caused by inattentive driver due to fatigue, sleepy or
drowsiness. It is successfully applied in USA and other countries. This paper
discussed types, design method, installation guidelines, and other related
issues that arise by the application of SRS such as impact on bicyclist and
noise annoyance to the neighborhood. Although there are some negative
aspects to their installation, the positive aspect appear to far outweigh the
negative one. This can be seen from the effectiveness of SRS in reducing
SVROR accidents, good aspect of benefit/cost ratio and easy installation
process. Although some issues might arise such as problems to the bicyclist
and noise annoyance, but these minor issues of SRS already have their
solutions as discussed earlier in the paper. Thus, in order to tackle the high
accidents rates particularly in term of SVROR accident caused by inattentive
drivers, authorities are recommended to use SRS as one of traffic safety


Special thanks to Ministry of Science and Technology (MOSTI), UTM

Research Management Centre and also Universiti Teknologi Malaysia for
funding this research under grant Vote Escience No. 79370.


FHWA.2000. Synthesis of Shoulder Rumble Strip Practices and Policies

[Online]. Federal Highway Administration (FHWA). Available: dept/pavement/rumble stri ps/reso
urces/synthesis/index.cfm#toc [Accessed 1 September 2010].
Kansas Cyclist .2010. Kansas Rumble Strips Policy [Online].Avaiable:
strip policy/ . [Accessed 16 December 2010].
The Southland Times .(2010, 30 March 2010). Woman Caught By Rumble
Abidin, A. N. S. Z., Rahim, M. M. A., Voon, W. S. & Sohadi, R. U. R. 2009.
Single-Vehicle Accidents Involving Trees in Malaysia – a Preliminary
Study . Malaysia Institute of Road Safety Research. Kajang

Bendtsen, H., Haberl, J., Sandberg, U. & Watts, G. 2004. SIL VIA Deliverable
12: Traffic management and noise reducing pavements
Recommendations on additional noise reducing measures .

Carlson, P. J. & Miles, J. D. 2003. Effectiveness Rumble Strips On Texas

Highways: First Year Report. Texas Transportation Institute.
Estrada, H. M. 2008, November 9. Rumble strips are rising grumbles. Star [Online]. Available: [Accessed
Gardner, L. W., Rys, M. J. & Russell, E. 2007. Comparison of football shaped
rumble strips versus rectangular rumble strip . Kansas Department of

Transportation, Kansas State University, University of Kansas. Report
No. K-TRAN : KSU-00-4P2.
Hirasawa, M., Motoki, A. & Saito, K. 2005. Study on development and
practical use of rumble strips as a new measure for highway safety.
Journal of Eastern Asia Society for Transportation Studies, 6, 3697-

Kee, S. S., Tamrin, S. B. M. & Goh, Y. M. 2010. Driving Fatigue and

Performance among Occupational Drivers in Simulated Prolonged
Driving. Global Journal of Health Science, 2 (1).
Liu, C. & Subramanian, R. 2009. Factors Related to Fatal Single-Vehicle
Run-Off-Road Crashes . National Highway Traffic Safety
Administration. Washington D.C
Minessota Department of Transportation 2010. Construction Inspection
Guidelines for Sec 626

[Online] .Available: php?title=626.3 Constru

ction Inspection Guidelines for Sec 626 [Accessed 2 September

Miska, E. (2009). Transverse Rumble Strips (TRS) . Chief Traffic, Electrical,

Highway Safety and Geometric Engineer.
Morena, D. A. 2002. The Nature and Severity of Drift-Off Road Crashes on
Michigan Freeways, and the Effectiveness of Various Shoulder
Rumble Strip Designs . Federal Highway Administration – Michigan
Division. Lansing
Neumann & Richard. 2002. Safer on Reflection. Traffic Technology
International , 96-97.
Perillo, K. 1998. The Effectiveness and Use of Continuos Shoulder Rumble
Strips. In: ENGINEER, H. (ed.). Albany: Federal Highway

Rasanen, M. 2005. Effects of Rumble Strips Barrier Line on Lane keeping in

curve. Accident Analysis & Prevention, 37, 575-581.
Riley, G. 2010. The hidden persuaders. Contractor. Auckland: Contrafed
Publishing Co. Ltd.

Torbic, D. J., Hutton, J. M., Bokenkroger, C. D., Bauer, K. M., Harwood, D.
W., Gilmore, D. K., Dunn, J. M. & Rochento, J. J. 2009. Guidance for
the Design and Application of Shoulder and Centerline Rumble Strips .
Transportation Research Board. NCHRP Report 641. Washington D.C

Van Berkel, C. 2009. Use of Transverse Rumble Strips in Rural Areas

(PW09031) - (City Wide) Public Works Outstanding Business List .


Nurul Elma Kordi1, Intan Rohani Endut1,2, Bahardin Baharom1,2 and Md Yunus Ab
Malaysia Institute of Transport (MITRANS), Universiti Teknologi MARA, 40450

Shah Alam, Selangor, Malaysia

Faculty of Civil Engineering, Universiti Teknologi MARA, 40450 Shah Alam,

Selangor, Malaysia

ABSTRACT: The attention of damages on flexible pavements due to the increased of legal
load limit is quite important for the authorities obtain some guide and evidence to decide
whether to allow the increase of legal load limit or to maintain the legal load limit. Among the
heavy trucks' operators, they want to increase in axle load from the current legal limit of 12
tons per axle. The ideas are to avoid delivery delays of goods to customers, encounter the
higher chargers of storage and also make the win-win situation between transporters and
their customers as it will enable the industry players to become more competitive, especially
among the neighboring countries. This paper’s aims are to identify and discuss the
seriousness of the type of damages occurred due to the increased of legal load limit and the
effects of increased of legal load limit on Malaysian Federal Road. By applying the Statistical
Package for the Social Sciences (SPSS) computer program for statistical analysis, the
seriousness of failure and the major group of distress can be determined. The findings have
been identifying major of pavement damages is fatigue cracking, followed by longitudinal
cracking and patch deterioration. Mostly, the level of seriousness is moderate.

Keywords: Axle Load, Damages, Flexible Pavement, Legal Load


This paper reports part of an ongoing research study. The main research is
mapping all types of damages in Malaysian Federal Road Route One (FT01) and
identify the types of damages occurred due to the increased of legal load limit.
For each type of damages, the possible causes and potential rehabilitation
options will identify as well.
The aim of this present paper is to report a study on the analysis of the major
types of damages on FT01 at Selangor, the seriousness of the types of damages
on FT01 at Selangor and the effects of increased of legal load limit on Malaysian
Federal Road. In order to achieve the set aim, the study observed, identifies and
discusses the type of damages and the criteria that affect the failure of pavement

The study combines a literature review with observations about types of

damages along the FT01 at Selangor and the analysis to achieve the aims. Data
obtained from the sites was analyzed using Statistical Package for the Social
Sciences (SPSS) computer program to produce statistics. By applying the SPSS
computer program for statistical analysis, the major group of distress and the
seriousness of failure can be determined. Furthermore, reports by others were
review and analyzed, together with the information obtained in this study to
develop engineering recommendations.


Road has played an important role in the trade and transportation system
throughout the world, and it becomes a rapid increase in the pavement
infrastructure development in Malaysia. Among the operators of heavy trucks,
they want to increase the axle load from the current legal limit of 12 tons.
President of Pan Malaysia Lorry Owners Association (PMLOA), Er Sui See said
lorries running on heavier loads had been doing so for more two years ago
without causing any problem. These lorries did not pose a risk to other road users
or damage on roads (The Star, 2010). The ideas to increase the axle load limit
are to avoid delivery delays of goods to customers, encounter the higher charges
of storage and also make the win-win situation between transporters and their
customers as it will enable the industry players to become competitive, especially
among the neighboring countries.
Although there are many potential benefits to increasing the legal weight limit

for trucks, many argue that it will only be more damaging to the infrastructure. On
behalf Work Ministry, heavy vehicle is one major factor contributed to
deterioration of the road. There were 19.3 million registered vehicles on the
Malaysia’s road, and the government spent RM5 billion between 2001 and 2010
to keep sustaining all the Federal roads (New Strait Times, June 2010). The
investigated has been done, and the result shows the total of heavy vehicle
breakdown three times more than light vehicles and this situation will give big
impact on traffic congestion (Berita Harian, 2010). The major factors of damages
on the pavement because of what need to be investigate.

The permission from heavy trucks’ operators to increase the axle load still
argues the responsibilities’ authorities because they need some guides and
evidence to ensure no problems occur on the pavements in the future. From the
findings, it will give some guide and evidence to decide whether to allow increase
the legal load limit or maintain the legal load limit.

3.1 Axle Load Limits
Damage is primarily caused by the heavier axle loads associated with large
commercial vehicles (Paul & David, 1997). There were 19.3 million registered
vehicles on the Malaysia’s road, and the government spent RM5 billion between
2001 and 2010 to keep sustaining all the Federal roads (New Strait Times, June
2010). For this reason, every country legislated for a maximum axle load limit and
maximum gross vehicle weight. Table below summarizes the maximum single
axle load applying among the various countries. From the Table 1 below, it shows
the highest axle load limit is Malaysia, and the lowest is Thailand. Different load
limits among Thailand and neighboring countries might also cause problems for
transshipment of goods across national borders and complicate international
trade negotiations (Transport Sector, 2008). The international standard of axle
load limit is 13 tons.

Table 1: Comparison of Single Axle Load among Various Countries

Country Axle Load Limit (Ton)
Malaysia 12
Thailand 8.2
Singapore 10
Japan 11
Europe 10-13
The People’s Republic of China 10

ASEAN member countries 10
Hawaii 10.9
New York 10.2
Texas 9.1
Washington 9.1

(Source: Transport Sector, 2008 and Paul & David, 1997)

Increasing the permitted gross weight of commercial vehicles can result in a

saving up to five percent in haulage costs (Paul & David, 1997). The damages on
road pavements will not be increased by an increase in the number of axles and
number of tires per axles because the load will distribute evenly among the axles
Marshek et al. (1986).

Since 1974, the legal weight limit for trucks in the U.S. has been 356 kN.
Due to increases in truck traffic, there is a need to raise this limit to make the
trucking fleet more efficient. There has been a proposal in the U.S. to increase
the limit to 431 kN. In order to satisfy the federal bridge formula, this weight
needs to be carried by adding an axle to the rear of the trailer. To quantify the
effects of this change in axle configuration on pavement damage, Peter & Tim,
2008, was utilizing the M-E Pavement Design Guide (MEPDG). The results
showed no change in design life for a variety of pavement cross-sections if the
overall weight of the traffic stream was held constant.

A weight limit change will improve truck safety by reducing the number of
trucks needed for delivery and hence reducing total truck miles driven (Runyan,
2010). For example, a 2009 study by the Wisconsin DOT study looked at truck-
related accident data in the state during 2006. Runyan (2010) says that from that
data, a reasonable estimate is that that if a law like SETA had been in place in
2006, it would have prevented 90 truck-related accidents in the state that year.
That follows actual data from the United Kingdom, which raised its gross vehicle
weight limit to 97,000 pounds for six-axle vehicles in 2001. Since then, fatal truck-
related accident rates have declined by 35%, as actual truck vehicle miles
travelled declined substantially. Table 2 shows the list of advantages and
disadvantages of increased axle load.

Table 2: Effect of Increased Axle Load

Advantages Disadvantages
Improve productivity Pavement damages
Air quality Low speed
Reducing carbon emissions Pavement rough
Help high meet maintenance cost Decreasing life of pavement

3.2 Types of Distress
Probably, the most common reason pavement damages are a loss of base,
subbase or subgrade support from things like poor drainage or spring thaw
(WAPA, 2011). It also due to excessive deflection of the pavement due to an
unstable base or due to repeated traffic loads heavier than what the pavement
was built to withstand (Querycat, 2011). For example, patch is replaced or adds
the new materials to the portion of pavements surface that has been damages
because of fatigue cracks, potholes and others. However, sometimes the patch is
not done properly by the contractor. If the patch not done properly, it will expand
more damages called patch deterioration and block cracking as well.

Indicator of structural failure, cracks allow moisture infiltration and make the
pavement roughness. If this cracking is not addressed then it will expand more
damage such as potholes (Pavement Interactive, 2010). Table 3 below shows the
types of damages that have been grouped into cracking, surface deformation,
and surface defect and patching and potholes. There have 19 types of damages
on flexible pavement.

Table 3: Types of Damages on Flexible Pavement


Cracking Fatigue cracking
Block cracking
Crescent shaped cracks
Longitudinal cracking
Transverse (thermal) cracking
Reflection cracking at joints
Edge cracking
Edge breaks
Edge drop-offs
Surface deformation Shoving
D e p r e s s i o n
Surface defect Bleeding
Polished aggregate
Patching and Potholes Patch Deterioration
(Source: Distress Identification Manual, 2003, Austroads, 1987 and Rani, 2007)

Rutting is defined as longitudinal deformation in the wheel paths, which occur
after repeated application of axle loading. It may occur in one or both wheel paths
of a lane. It can be categorized as either traffic load associated deformation, wear
related or the combination of both. The causes include traffic load, age of
pavement and deformation of the entire pavement structure or instability in the
form of one or more pavements. Rutting will make the road surface rough, patchy
and bumpy, which can lead to safety problems. The unit use to measure rut
depth is the millimeter (mm). Plotting rutting and number of year will give us the
following graph (Figure 1) which will assist the pavement authorities to decide as
to when to start rehabilitation or maintenance work on the pavement. If the
authorities only accept rutting below than 12.0 mm; the authorities will need to
maintain the surface at year five (Mohd Isa et al., 2005).

Figure 1: Prediction
Graph of Rutting Based on the Developed Model
(Source: Mohd Isa et al., 2005)

4.1 Data Collection

The data was collected at all Federal Road Route One (FT 01) in Selangor.
There has divided into three districts, which is Hulu Langat (Kajang – Semenyih -
Beranang), Gombak (Rawang – Selayang Baru – Batu Caves – Jalan Ipoh –
Jalan Kucing) and Hulu Selangor (Serendah -Ulu Yam- Kuala Kubu Baharu –
Kalumpang - Tanjung Malim (border Selangor and Perak)). The road has been
observed about 92 kilometers. Location FT01 in Selangor has been chosen
because earliest federal road in Malaysia and Selangor is highest having
industries locations compare to other states. Every type of damages and the
seriousness of type’s damages were identified followed the Distress Identification
Manual for the Long-Term Pavement Performance Program (2003). The

seriousness of types of damages is depending on their severity levels, which are
low, moderate and high. Usually, every type of damages has their own
measurement (unit of measure) to define their severity levels.


From the observation along the districts of Hulu Langat, Gombak and Hulu
Selangor, the major distress always occurs are fatigue crack, followed by
longitudinal cracks and patch deterioration. From the 194 locations of
observations around these three districts, it shows the cracking is grouping
damages that almost happen on the federal road which 58.2%, followed by
patching and potholes (19.6%), surface deformation (11.9%) and surface defect
(10.3%). Table 4 show the frequency and percent of groups of damages.

Table 4: Frequency and Percentage Group of Damages on FT01 (Selangor)

Group of Damages Frequency Percent (%)
Cracking 113 58.2
Patching and Potholes 38 19.6
Surface Deformation 23 11.9
Surface Defect 20 10.3
Total 194 100.0

Table 5 below shows the frequency and percentage of types of damages

around FT01 around Selangor. It shows that, fatigue cracking is the highest
percent of damages which 33.5%, followed by longitudinal cracking (21.6%) and
patch deterioration (16.5%).

Table 5: Frequency and Percentage Types of Damages on FT01 (Selangor)

Types of Damages Frequency Percent (%)

Fatigue cracking 65 33.5
Longitudinal cracking 42 21.6
Patch Deterioration 32 16.5
Rutting 23 11.9
Polished Aggregate 12 6.2
Ravelling 8 4.1
Potholes 6 3.1
Edge cracking 4 2.1
Transverse cracking 2 1
Total 194 100.0

From the observations, fatigue cracks always occur at left lane which vehicle
that bring the heavy loads usually used this road. This mean that, fatigue crack
can occur due to repeated heavy vehicles. Longitudinal cracking and patch is
badly deteriorated especially on Hulu Langat District compare to Hulu Selangor
and Gombak. Longitudinal cracking always occur at the middle of between two
tires (not in the wheel path). At several place, longitudinal crack occurred along the
way. The primarily causes of longitudinal cracking are improper construction
practices, followed by a combination of heavy load repetition and loss of
foundation support due to heave caused by swelling soil ( Ardani et al. , 2003).

Mostly patch at Hulu Langat has a high severity level. Figure 2 until Figure 4
shows the high seriousness types of damages occurred at several locations
around Hulu Langat, Gombak and Hulu Selangor.

(a) (b)
Figure 2: High Seriousness of Fatigue Cracking happened at Hulu Langat District (KM 340)

(a) (b)
Figure 3: High Seriousness of patch deterioration happened at Gombak District
(KM 410)

(a) (b)
Figure 4: High Seriousness of fatigue cracking, potholes and rutting happened at Hulu
Selangor District (KM 456)

The characteristics for seriousness of damages are divided into three scales
which are low, moderate and high. Table 6 below shows the percentage of
seriousness types of damages around the FT01 around Selangor. It shows the
major level of severity types of damages occurred is moderate which is 65.5%.
However, 21.1% high seriousness types of damages still happened at several
locations especially at left lane and stopping point such as junctions and traffic

Table 6: The Seriousness Types of Damages along the FT01 (Selangor)

Seriousness of Damages Frequency Percent (%)

Low 26 13.4
Moderate 127 65.5
High 41 21.1
Total 194 100.0

From the observation along the Federal Road Route One at Selangor, it can
be concluded that the major distress always occurs are fatigue crack, longitudinal
cracks and patch deteriorations. From the literature review, the possible causes
of these damages are because of repeated loading, poor construction and less
quality material used.

Mostly, the seriousness of the types of damages on the FT01 at Selangor is

moderate. However, at several locations, there has high severity level especially
possible place vehicle drive slow and stop such as at left lane (left lane usually

for heavy vehicle), curve, junction, u-turn junction, traffic light and bus stop.
From the literature review and observation, the effects of increased of legal
load limit on Malaysia Federal Road will increase the possible of damages.
However, the damages can be reducing by an increase number of tires per axle
(tandem axle) and increase the number of axles because greater surface contact
area between tire and pavement will reduce pavement damages. Also, increase
the speed will be decreasing the load duration.

The authors expressed their gratitude to Malaysian Logistics Council (MLC)
for initiated the grants awarded by the Ministry of Higher Education (MOHE). The
authors would also like to express their appreciation to Mr. Osama Mahmoud
Yasenn, Mr. Elmi Alif Azmi and Mrs. Nuryantizpura Mohd Rais for their
assistance in observation types of damages along Federal Road Route One. The
authors acknowledge with thanks the kind permission of the Director of
MITRANS, Prof Madya Sabariah Mohamad for this paper to be published and

American Association of State Highway and Transportation Officials (1993),
“AASHTO Guide for Design of Pavement Structures”, Washington, D.C.

Ardani, A., Hussain, S. & LaForce, R. (2003). Evaluation of Premature PCC

Pavement Longitudinal Cracking in Colorado . Proceedings of the 2003 Mid-
Continent Transportation Research Symposium, Ames, Iowa.

Austroads (1987). A Guide to The Visual Assessment of Pavement Condition .

Sydney NSW 2000 Australia.

Berita Harian. (2010, July 30). Tiada Kompromi Lori 6 Tayar Masuk Lebuhraya, pp.

John, S. M. & William, Y. B. (2003). Distress Identification Manual for the Long-Term
Pavement Performance Program (4th revised ed.). U.S. Department of
Transportation, Federal Highway Administration, 6300 Georgetown Pike.

Marshek, K. M., Chen, H. H., Connell, R. B. & Saraf, C. L. (1986). Effect of Truck
Tire Inflation Pressure and Axle Load on Flexible and Rigid Pavement
Performance . Transportation Research Record N1070, Transportation Research
Board, Washington DC.
Mohd Isa, A. H., Ma’soem, D. M. & Hwa, L. T. (2005). Pavement Performance
Model For Federal Roads . Proceeding of the Eastern Asia Society for
Transportation Studies, Vol. 5, pp. 428-440
New Strait Times. (2010, June 22). Fewer Potholes with New Type of Asphalt, pp.
Paul, C. & David, C. (1997). The Design and Performance of Road Pavements (3rd
revised ed.). New York: McGraw-Hill 1997.
Pavement Interactive (2010). Pavement Distress . Retrieved from Error! Hyperlink
reference not valid.

Peters, K. & Tim, D. (2008). Effects of Increasing Truck Weight Limit on Highway
Infrastructure Damage . Retrieved from http://road-transport -

Querycat (2011). Querycat Question . Retrived from

Rani, O. A. (2007). The Effectiveness of Pavement Rehabilitation at Kuala Lumpur

Karak Highway . Master thesis, Universiti Teknologi Malaysia.

Runyan, J. (2010). Transportation Productivity Coalition Hopes to Increase Weight

Limits for US Trucks, but Law Likely Depends on Passage of Full Highway Bill.
Retrieved from

Thailand Infrastructure Annual Report (2008) Retrieved on 12 April 2010 from
1177475763598/3714275-1 234408023295/5826366-123440810531 1/chapter2-
transport-sector. pdf

The Star. (2010, July 27). Heavy Vehicle Banned , pp. 12.

The Star. (2010, May 5). Tolled Roads are Better Maintained , pp 11.

Washington Asphalt Pavement Association, WAPA (2011). Alligator Cracking.

Retrieved from


Osama Mahmoud Yassenn 2

, Assoc. Prof. Dr. Mohamed Ahmed Hafez,

Dr. Md Yunus Ab Wahab, Dr. Intan Rohani Endut,

And Assoc. Prof. Ir. Bahardin Bin Baharom

ABSTRACT Increasing axle load limit is one of the important issues nowadays in Malaysia.
This paper takes a case study which is an across section at Lingkaran Luar highway to study
the increasing of damage caused by increasing axle load. This paper uses axle load factor to
describe damage increasing since axle load factor compare the damage caused by the
arbitrary load to the standard axle load. This paper introduce the method of using finite
element analysis to calculate axle load factor, it also suggest using of falling weight
deflectometer to determine the modulus of elasticity for the analysis. It was found that
increasing axle load limit will cause a rapid increasing in axle load factor. This increment in
the damage percentage can reach four times of the increment of axle load limit increment

Keywords: axle load limit, damage ratio, finite element analysis.

2 Student , Universiti Teknologi MARA, Malaysia Institute of Transport, 40450 Shah Alam, Selangor,
MALAYSIA Tel: +60193692471 , E-mail:


Transportation is very important. It plays a role similar to blood circulation system in

the human body. When the efficiency of transportation system increased the whole

society will get the benefit and if its efficiency decreased the whole society will be


In Malaysia, it is very obvious that traffic volume and traffic loading is increasing due

to the economic development of the country. On the other hand the pavement

damages and maintenance cost is also increasing. To sustain this development it is

very important not only to plan for constructing new pavement, but also to study,

preserve and upgrade the existing pavement. One of the current issues related to

this, is that the local container hauliers want to increase axle load limit to decrease

their transportation cost. This will allow them to meet the high maintenance costs

and turnaround of their vehicles. Unfortunately this will increase the damage caused

by their vehicles to the pavement. This paper will concentrate on a case study at

Lingkaran Luar highway, to predict the damage increasing due increasing axle load

limit on this highway.

Finite element method has been proven to be useful in pavement analysis; this

paper will use the finite element method to determine the value of axle load factor.

This factor is being used to describe the damage caused by certain loading

comparing to the damage caused by standard axle load. The axle load factor that

has been determined is based on rutting and fatigue criteria.

Another objective of this paper is to compare the axle load factor determined by

using different rutting and fatigue models. The models that will be compared are

American Institute model, Shell Model, and the Indian model.


The objectives of this paper are

1- To introduce using of finite element method in pavement analysis to

determine axle load factor for different axle loads for a certain pavement
cross section.

2- To compare the axle load factor determined based on American Institute

model, Shell model and the Indian model.

Flexible pavement is a complex structure contains different materials with
different behaviour. This explains why it is not easy to have a perfect
mathematical description to model its behaviour.

Pavement structural analysis started with linear elastic theory that

has been developed by Boussinesq [1]. It is a quite simple analytical model
where soil is assumed to be an elastic material, and can be described
completely by modulus of elasticity and Poisson’s ratio. Then this theory
was extended by the work of Burmister to multilayer elastic model [2]. In this
model pavement is considered to have different layers with different elasticity
modulus and Poisson’s ratio. Burmister developed his solution first for a two-
layer system and then extended them to a three layers system [3]. This
model can be used for low stress and strain levels. In this paper the
multilayer elastic theory has been used for the analysis of the pavement
because of its suitability since the stress and strain that will be measured is
so small.


The conventional methods to measure the modulus of elasticity for pavement

materials are (California bearing ratio) test, triaxial compression test,

and in-situ plate loading test. The determination of elasticity modulus using

triaxial compression test or in-situ plate loading test is time consuming and

may involve significant expenses. Furthermore, there is a limitation for the

number of the samples that can be obtained from the site [4].

For the CBR test, there are many experimental equations have been

developed to use the CBR value to determine the elasticity modulus of the

material. Some of these equations are:

(Shell relationship)
(Poulsen & Stubstad , 1980)[5]
(Powel et al., 1984)[6]

stands for the modulus in MPa.

Regardless of these studies, it was proven that there is no obvious

relationship between and , moreover, the test was not suitable for

granular materials with large particle sizes. Therefore, the use of

correlation for this type of materials can lead to large errors [4].

Due to the above mentioned reasons, it is suggested that the falling weight

deflectometer is used to determine the moduli of pavement layers. Falling

weight deflectometer is a non-destructive testing device used for pavement

properties evaluation including elastic moduli for the different layers. The

falling weight deflectometer is designed to transmit on a load pulse to the

pavement layers by dropping a large weight on pavement surface. Deflection

sensors will measure the deformation of the pavement in response to the

loading. By a method called "backcalculation”, the elasticity modulus can be

Figure 1. Falling weight deflectometer [Dynatest]

Table 1 modulus of elasticity for the studied cross section

The modulus
Layer of elasticity in Layer thickness
MPa in mm

AC layer 797 195

Base layer 174 379

Subbase layer 116 227

Subgrade Layer 34 102



The pavement layers in the finite element method are considered to be a solid

continuum. The solid continuum domain is divided into sub domains. Each sub

domain is divided into a number of finite elements. These finite elements assemble

to the whole problem during the analysis. These elements are connected by

common nodes at their common ends. The analysis gives an approximate solution

of the problem for different boundary conditions and under various loading types

using energy or stiffness formulation [7]. Finite element models have been used

intensively pavement analysis and design. There are three types of models which

have been used for multilayered pavement structure analysis and design: three-

dimensional (3-D), plane strain, and axisymmetric [7]. For our case study,

axisymmetric model was used, since it is adequate and to reduce analysis time.

Even though using the finite elements three-dimensional models can solve all the
problems that can be solved using two-dimensional models, three dimensional

models are very expensive to be used in terms of computational time and data

preparation [8].

Figure 2. Finite-Element Discretization of the pavement cross section (Axisymmetric)

Figure 3. pavement cross section geometry and boundary conditions


One of the important parameters for pavement analysis is the tire-pavement

contact area. Tire-pavement pressure distribution is usually recognized to be

complex and the distribution is affected by the type of the tire. There are a lot of
inconsistencies in the data collected from different experimental studies

evaluating the contact pressure distribution between pavement and tire [9].

However, the area of contact may be represented by two half circles with a

rectangle between them as shown in Figure 7, this shape, which consists of a

rectangle and two half circles, is converted to a rectangle as recommended by

Huang, having an equivalent area of 0.5227 L2 and a width of 0.6 L.

Unfortunately this area is not axisymmetric and cannot be used for axisymmetric

model[10]. Since our model is axisymmetry, a circular contact area having an

equivalent area of 0.5227L2 was used. This assumption has been used for the

multilayer theory. Although it is not correct, the error incurred is believed to be

small [10].

Figure 4. tire-pavement contact area [?J


Flexible pavement is a structure that consists of at least three layers; asphalt

surface layer, unbound aggregate base layer, and natural soil subgrade layer. Under

usual conditions, the stress distribution will be similar to the one in Figure 8. Rapid

shifting from compression to tension in the top and bottom of the pavement is

caused by moving wheel load. The repetition of this action develops tensile strain

will cause fatigue cracks. In accordance with elastic layer theory, the maximum

strain located at the bottom of the surfacing layer [9]. Thus most pavement design

models are based on strain at the bottom of the asphalt layer to forecast the
performance of fatigue cracking. In addition, rutting is caused by cumulative vertical

strain on the paving layer and the subgrade [11].

Figure 5. Flexible pavement response affected by uniform load P; successive wheel

positions A and B cause cyclic loading that result in compressive stress and strain
on subgrade, and tensile strain and compressive strain in pavement [12].

After finite element analysis complete there are two response parameters that will

be considered. The first one is the maximum horizontal tensile strain at the bottom

of asphalt layer which is responsible for fatigue failure. The second is the maximum

compressive strain at the top of the road base layer which is responsible for rutting

failure. The figures below shows the horizontal strain and vertical strain distribution

for Lingkaran Luar highway cross section studied.

Figure 6. Horizontal strain distribution

Figure 7. vertical strain distribution


The ratio of (number of repetitions to failure of standard load and pressure )

to (number of repetitions to failure of arbitrary load and pressure )with the same

material is the Equivalent Axle Load Factor (EALF)


= number of repetitions to failure of standard load = number of

repetitions to failure of arbitrary load = maximum tensile strain at
the underside of asphalt layer under the

standard single axle load

= maximum tensile strain at the underside of asphalt layer under arbitrary Load
= fatigue coefficient. In AI model = 3.291 ,in Shell model = 5.671, in Indian model
= 3.89.

The table below shows the calculated axle load factor based on fatigue criteria for
the case study. We can see that axle load factor will increase rapidly by increasing
axle load. This indicates that increasing axle load will increase the fatigue failure by
a percentage higher than increasing axle load percentage.

Also it can be seen that for fatigue criteria the shell model will give the highest axle
load factor value and American institute model will give the lowest values.

Table 2 axle load factor based on fatigue criteria for Lingkaran Luar highway cross section

8.16 ton 9 ton 10 ton 11 ton 12 ton

AI model
1.00 1.18 1.37 1.58 1.74

Shell model
1.00 1.33 1.73 2.20 2.59

Indian model
1.00 1.22 1.46 1.72 1.92


= number of repetitions to failure of standard load.

= number of repetitions to failure of arbitrary load.
= maximum compressive strain under the standard single axle load.
= maximum compressive strain under arbitrary Load.
= fatigue coefficient. In AI model = 4.477, in Shell model = 4, in Indian model =

Table 3 axle load factor based on rutting criteria for Lingkaran Luar highway cross section

8.16 ton 9 ton 10 ton 11 ton 12 ton

AI model
1.00 1.32 1.76 2.26 2.81

Shell model
1.00 1.28 1.66 2.07 2.52

Indian model
1.00 1.33 1.77 2.28 2.85

From the above table it can be concluded that increasing axle load limit on
Lingkaran Luar highway will cause increasing of rutting failure. Axle load factor will
be increased. This increment of axle load factor will be much higher than the
increasing percentage of axle load. And by increasing 50% on axle load the axle
load factor and the rutting damage will increase by more than 150%.

Also it can be seen that, although Indian model will give the highest axle load factor
based on rutting criteria all other models will give a very close value for the axle load
factor based on rutting criteria.


The axle load factor will be taken as the highest between the axle load factor based
on fatigue and based on rutting criteria. And worth mentioning that although the axle
load factor based on Indian model gives the highest values, and the shell model give
the lowest, the axle load factor determined by the three models will almost be the
same, even though when specific criteria considered like fatigue only, they will be
Table 4 axle load factor based on rutting and fatigue criteria for Lingkaran Luar highway
cross section
8.16 ton 9 ton 10 ton 11 ton 12 ton

AI model
1.00 1.32 1.76 2.26 2.81

Shell model
1.00 1.33 1.73 2.20 2.59

Indian model
1.00 1.33 1.77 2.28 2.85

Figure 8. Increasing of axle load vs. increasing of axle load factor

In can be seen from the figure above that increasing axle load will increase axle load
factor rapidly. By increasing axle load 47% the damage due to the loading by 180%
according to AI and Indian models and 160% according to shell model. This will
increase the maintenance cost of the pavement rapidly.
Increasing axle load limit on the pavement would cause a rapid
AI model increasing
shell of model
model Indian
fatigue and rutting. This was shown by the rapid increasing of axle load
factor based on ratting and fatigue criteria. The increasing of the damage
could reach four times of the damage caused by standard axle load. So
0% 10% 20% 30% 40% 50%
increasing axle load limit on Lingkaran Luar highway is not advisable since
the maintenance cost can increase rapidly because of this action.

The highest axle load factor based on fatigue criteria will be given by Shell
model. The highest axle load factor based on rutting criteria will be given by
the Indian model. And overall the axle load factor based on both criteria will
have almost a similar value for Indian and American models, and the lowest
values for Shell model.

The writers would like to thank DYNATEST ASIA PACIFIC SDN. BHD . in
Malaysia for their help and for the data provided.

1- Boussinesq, J. (1885). Application des potentiels á l'étude de l'équilibre
et du mouvement des solides élastique. Paris: Gauthier-Villard.
2- Burmister, D. M. (1943). The Theory of Stresses and Displacements in
Layered Systems and Applications to the Design of Airport Runways.
Proceedings, Highway Research Board, Vol. 23, 126-149.
3- Burmister, D. M. ( ). The General Theory of Stresses
Displacements in Layered Soil Systems. Journal of Applied Physics, Vol.
4- Lacey, G., Thenoux, G., & Rodríguez-Roa, F. (2008). THREE-
Journal for Science and Engineering, Vol. 33, 1B–66-61B–66.
5- Sukumaran, B., Willis, M., & Chamala, N. (2005). Three Dimensional
Finite Element Modeling of Flexible Pavements. Paper presented at the
Advances in Pavement Engineering (GSP 130), Austin, Texas, USA
6- Poulsen, A., & Stubstad, R. N. (1980). fastlæggelse af E-moduler for
underbundsmaterialer ved målinger i marken. Interne Notater
108,statens vejlaboratorium,vejdirektoratet.
7- Cho, Y.-H., McCullough, B., & Weissmann, J. (1996). Considerations on
Finite-Element Method Application in Pavement Structural Analysis.
Transportation Research Record: Journal of the Transportation Research
Board, 1539(-1), 96-101.
8- Fang, H., Haddock, J. E., White, T. D., & Hand, A. J. (2004). On the
characterization of flexible pavement rutting using creep model-based
finite element analysis. Finite Elements in Analysis and Design, 41(1),
9- Helwany, S., Dyer, J., & Leidy, J. (1998). Finite-Element Analyses of
Flexible Pavements. Journal of Transportation Engineering, 124(5), 491-
10- Huang, Y. H. (2004). Pavement analysis and design: Pearson/Prentice
11- Lay, M. G. (2009). Handbook of Road Technology: Taylor & Francis.
12- Mulungye, R. M., Owende, P. M. O., & Mellon, K. (2007). Finite element
modelling of flexible pavements on soft soil subgrades. Materials &
Design, 28(3), 739-756.


Amiruddin Ismail 1 and Mohammad Hesam Hafezi 2 and Omran Kohzadi
Seifabad 3
Sustainable Urban Transport Research Centre (SUTRA)/

Department of Civil and Structural Engineering

Faculty of Engineering and Built Environment
Universiti Kebangsaan Malaysia
43600 UKM Bangi, Selangor Darul Ehsan, Malaysia

ABSTRACT: Operating time and cost are the main parameters used by both passengers
and operators of a land public transportation. A relative good public transportation bus line
network is to provide easy access and cheaper in cost to the users. This article represents
survey behaviour of passengers and operations of a bus line network. The data obtain in city
of Tehran, Iran was used in this study. Data such as waiting time, fare, and number of buses
used in the operation were compared. Waiting time included wait time in bus stop for bus
arrival and time block in bus. Fare is payment of each passenger from original station to final
destination. Operator viewpoint prefer to use minimum number of buses for service with
highest efficiency. High performance level has been shown by using statistical analysis for
the above-mentioned parameters in a bus line network as long as a specific lane has been
devoted for bus services. It was found operators of transportation service want to reduce
their operating cost and increase their service performance. It has been highlighted
passenger who used bus lane pay less fare and less time for waiting and journey time.
Keywords: Bus line network, efficient bus lane, operating cost, waiting time


Buses are the most widely used in transit technology today because bus
networks are easily accessible and cheaper than other kinds of public
transportation. They are operated in nearly all cities with transit service
and in a majority of them are the only transit modes. There exist two types
of public transportation in Tehran: Bus network and Subway network. Due
to the limited subway lines, passenger demand for using the bus network
is higher. Studies of improving the performance of bus services are
important because of the increasing demand patterns of passengers. On
the other hand, operating cost of bus services is lower than other types of
public transportation. Verifying the case of cost is important in
optimization public transportation. Generally, cost is divided into two
parts: direct costs and indirect costs (Ibeas et al. 2010). Furthermore, the
cost of bus service includes operating cost and the cost of users which if
they are minimized can increase efficiency (Wirasinghe & Vandebona
2010), an objective for all societies to achieve (Meignan et al. 2007).
Overall, the evaluation of bus service can conducted from three points of
views: travellers, operators and authorities (Chen et al. 2009). One of the
specifications to improving performance is the devotion of a specific lane
for their services. And, this can be implemented in three major ways:
mixed traffic lanes (MTL), regular bus lane (RBL), and exclusive bus lane
(EBL) (Vuchic 1981). In MTL, bus routes are dedicated to urban mixed
traffic. In other words, buses move alongside other motor vehicles and
non-motorized vehicles in a street. In RBL, there is a specific lane for bus
services during peak period traffic where the route is separate from other
vehicles. Besides this, a contra-flow bus lane can allow buses to travel in
the opposite direction to other vehicles (RBLs). Finally, in EBL there is an
exclusive bus lane for the use of the buses separated from other traffic at
all times of a day. In this paper we present the result of survey behaviour
of passengers and operations of a bus line network in MTL and EBL.


Tehran is the capital city of Iran with about 8,791,378 citizens in an area
of 730 km2. The administrative structure of Iran is focused in Tehran
which has 22 regions and 112 districts. With such a vast area, the trip rate
of Tehran is about 12,850,000 million per one day and the share of buses
network in public transportation is about 20/3 percent which is the largest
share of public transportation in Tehran. As a result of this extremely high
demand of passengers using it, the bus line to be studied in this paper is
about 10,950 m from west to east where the transfer of passengers from
suburban areas to the city centre takes place. The total number of
passenger transfers with this line is 18,700 per day.

Figure 1. Schematic map of bus line in Tehran

It has 15 bus stops and two terminals where the origin and destination
line terminals were used only for parking buses and crews rest and not for
boarding and alighting passengers, as shown in Figure 1. To organize
passengers, a metal framework in each bus stop is used. Some bus stops
are shared stops with other lines. These stations are larger compared to
the non-shared ones. While, the first bus stop is for boarding passengers
from the origin line, the last bus stop is for alighting passengers from
buses in the destination line. The total number of buses is 22 and all of
them are of one make. Each bus has 45 seats and the total capacity is 75
people. The routes of this bus line are composed of two sections: the first
has mixed traffic lanes with a distance of 6,240 m served by 8 bus stops
and the second is the exclusive bus lane with a distance of 4,710 m and 7
bus stops. Generally, the traffic route is divided into three periods: first,
peak-hour traffic in the morning, second, traffic hours during the day, and
third, peak-hour traffic in the afternoon. For convenience, passengers
have two ways for paying their fare: by using the touch-n-go card when
boarding the bus and by paying cash to the bus-driver when alighting.


3.1. Mixed traffic lanes and exclusive bus lane

Usually, when the width of the street is less than 2 lanes for traffic, bus
service is implemented alongside other vehicles duration their mission. All
classes of vehicles are allowed use on the street. The location of bus
stops is designed outside of the route on sidewalk because when the bus
stops at the station, it blocks one of the lanes and other vehicles can use
the other lane to carry on with their journey.

The travel direction is one-way and there will be more delays of bus
service during peak-periods traffic. For street tracks width that are wide
that allow traffic congestion, the exclusive bus lanes are used. The bus
service is implemented in independent alignment. Vehicle classes using
this alignment are buses and emergency vehicles for their mission such
as: ambulances, fire trucks and police vehicles. Roadways are physically

separated from other traffic with fences, curbs or greenery. For
alignments on a side street, the location of bus stops is on the sidewalk
and for those alignments located in the middle of the street such as Bus
Rapid Transit (BRT), bus stops are embedded inside exclusive bus lanes.
Usually, the travel direction is two-way.

From the route information of the bus line to be studied in this paper, the
distance and layout table between bus stops given in Table 1 can be
derived. This bus line is composed of two parts.

Table 1. Distance and layout of bus stops

Stop number Distance (m) Layout
1 0 MTL
2 950 MTL
3 650 MTL
4 450 MTL
5 360 MTL
6 900 MTL
7 600 MTL
8 380 MTL
9 1400 EBL
10 550 EBL
11 400 EBL
12 550 EBL
13 1100 EBL
14 330 EBL
15 380 EBL
16 1100 MTL
17 850 MTL

Route between origin station to bus Number 8 and also route between
bus Number 16 to destination station is Mixed Traffic Lanes. And the
remainder of the route which is between bus Number 9 and 15 is the
exclusive bus lane.

3.2. The effective components in bus service

To achieve the effective components, several components have to be
satisfied. The first is the passengers. It can be said that passenger
behaviour in bus mission is the most important effect component.
Passenger behaviour includes: first, walk to the bus stop, second, wait at

the bus stop for the arrival of bus, third, wait time for boarding and
alighting and fourth, bus transportation (Meignan et al. 2007; Yan & Chen
2002). If distance between bus stops is equal or near equal, it has some
advantages, such as: passengers who walk to bus stop can choose the
nearest bus stop. Also, optimizing bus stop spacing can minimize social
costs (Ibeas et al. 2010). The waiting time for passengers at the bus stop
for the arrival of the bus follows from two important issues: period
frequency bus mission from first bus stop (origin) and delay function along
highway. Generally, the origin and destination bus mission took more time
for boarding and alighting passengers from bus. But, during peak-hours
traffic this issues can cause large delays. Due to passenger demand for
using the bus network is higher and often the capacity of the bus is near
full. Finally, passengers after boarding the bus should wait for arrive to
their destination. The journey time is a complex function that will include
many terms that are out of the scope of this paper.

The second effect component in bus service is operating. It includes

some terms such as: observance real-time departure from origin station,
follow from designed path for mission, proper stop in each bus stop and
preservation of the average speed designed for mission. Usually for
improving bus service, the bus companies prepare a schedule for it.
Normally, observance of real-time departure of all buses existing in origin
station according schedule can reduce delays in their mission
(Oudheusden & Zhu 1995; Lam et al. 2009). In addition, following the
real-path and not stop out of bus stop spacing on the route can reduce
delays too. Moreover, proper stop in each bus stop has some advantages
such as: facilitates passengers for boarding and alighting passengers,
wasting less time for load passengers, more safety for passengers who
are alighting. Finally, for a reduction in total delays the bus operator can
adjust the bus average speed along different sections of the route.

The third component is fleet buses. For this the following have to be
noted: adjustment to number of bus needed, organization of the fare
system and optimization of bus stops. The fleet size is important factor in
operating costs. Optimum use from existing buses at different times of the

day for their mission can increase bus service reliability (Chen et al.
2009; Lam et al. 2009). By using the touch-n-go card, paying fare time
can be reduced, and subsequently, decrease the total delays in bus
mission. Optimization of the bus stops includes spacing and framework.

Figure 2. Simulation results for Tehran bus-network, measure of the load of passenger.

Finally, distance on foot and travel time is related to the

locations of bus stop. Increase distance between bus stops causes
increase in the distance on foot. On the other hand, this issue
decrease travel time in the vehicle. To compensate part of the
operating cost advertisement at the bus stop can be carried out.

Having discussed the three effective components in a bus

service above, the next section will discuss these components
using numerical examples.

3.3. Quantitative Approach

3.3.1 Number of passengers

For selecting the number of buses needed for a bus line, some
conditions have to met, for instance, the geographical and length of
route, bus type and passenger demand. One way of lowering the
waiting time for the bus arrival on long routes is to increase the
number of buses. Another way is use buses with more capacity, such
as, articulated bus or Double-decker bus. During peak-hour traffic
periods, passenger demand for using the buses is more than during
non peak-hour traffic. Figure 2 plots the simulated distribution of one
bus load of passenger for a day.

This measure results in one month of the bus trips. Obviously, the
peak periods are at 07:30, 13:30 and 17:30 hours which are
commonly observed in urban traffic. This measure allows the location
of overloaded and unused buses. To solve these problems, a
rescheduling of the bus service during peak periods is conducted.

3.3.2 Journey time of bus mission

The fundamental elements in scheduling the bus service are distance,

speed and time. Distance refers to the time the bus enters and exits
from the terminal and the distance of the bus stops. For speed,
allocate the average speed. Total time is the time of departure
between bus stops and the time block in each bus stop for boarding
and alighting passengers.

is the journey time for bus mission from origin to destination station, can
calculated by below formula:


The above parameters are defined below: is the number of bus stops

is the average speed on the total of route is the average time block in
each bus stop

is the distance between origin terminal to first bus stop
is the distance between destination terminal to last bus stop

Figure 3 represents the actual measure of journey time for bus

mission from origin to destination station. Variance between actual
and virtual for this measure is around 0/56 percent. To wit, the actual
journey time around 0/56 percent is more than the virtual journey time.
Average running time (including travel time and stop time) is around
5/15 and 2/87 minute for MTL and EBL ways, respectively. Different
reasons have caused an increase in this value. They are traffic jams,
traffic lights and large volumes of passengers.

Figure 3. Simulation results for Tehran bus network, measure of journey time

This figure indicates the time of loading and unloading passengers
at the first and last bus stops is more than at the intermediate stations
due to the higher number of passenger for boarding and alighting,
respectively. Furthermore for MTL ways, the average journey time
between each bus stop is more than EBL ways.

Figure 4. The average speed of bus in the peak-hour traffic

3.3.3 The average speed of bus mission

The speed of bus is variable in different parts of the route. Many

factors have influenced this: traffic congestion, distance between
stations and physical route (Hwe et al. 2006).

The average speed of bus in different parts of the route during
peak-hour traffic is shown in Figure 4. At period-hour traffic, the
average speed of bus is reduced significantly. According to the
statistics, the average speed of bus at peak-hour traffic is reduced to
around 62% as compared to non peak-hour traffic. The average speed
of the bus is around 8 and 13 kilometres per hour for MTL and EBL
ways, respectively. Considering the relationship between speed and
velocity, if bus stop spacing is of equal amounts, the bus operator can
run with symmetric average speed in different parts of the route.
Usually, the direct routes in designing bus routes are used not
because of diminished acceleration of buses. This figure shows that
average speed in EBL way is more than that of MTL way.

3.3.4 Delays of bus mission

Overall, the behaviour function of delays is linear. The waiting time for
bus arrival at the last bus stop is sum total delays occurring in the past
bus stops. Delays of bus mission include cases such as delay at
departure from first station, traffic congestion and passenger volume.
The bus departure from the first station has to be according to
schedule. Rest crews should be planned, so they finished before
starting on the new mission.

Figure 5. Delays between two bus missions

The average speed of bus is reduced at peak-hour traffic causing

more delays. Increasing passenger demand at bus stops can cause
an increase in time block at bus stops for the bus mission. Delays
between two bus missions are as shown in Figure 5. When departure
period from the first station is 15 minutes, the average of delays in
EBL way is less than that in MTL way.

The average of travel time delay in each bus stop is between 151
and 31 seconds for MTL and EBL ways, respectively.


In this paper, survey behaviour of passengers and operations of a bus

line network in mixed traffic lanes and exclusive bus lane has been
presented. Statistical analysis combines volumes of passengers, bus
journey time, average speed and delays of bus mission during peak-hour
traffic in the morning. We have shown that the use of the exclusive bus
lane layout can estimate more bus service reliability with a decline in total
delays . Moreover, to improve bus service in crowed areas the mixed
traffic lanes and exclusive bus lane can be used together. With increasing
average speed of bus in EBL ways and decline running time between bus

stops, can compensate delays at the last stations. This approach reduces
total delays in bus mission . This approach has been applied and
validated on a real case study in a bus-network in the city of Tehran.
Probably, it is also a useful tool for many other bus companies in large
cities of the developing world.

The main perspective of this work is to evaluate bus service reliability.

They are useful to regulate the scheduling of bus service. The location of
EBL in the bus route is an interesting challenge (Bermond & Ergincan
1996). Forthcoming works will consider increasing the reliability of bus
mission schedule at peak- hour traffic.

We thank the Fara Tarabar Mahdi bus company for providing the test
data and their valuable opinions.


Bermond, J.-C. & Ergincan, F. Ö. (1996). Bus interconnection networks.

Discrete Applied Mathematics 68(1-2): 1-15.

Chen, X., Yu, L., Zhang, Y. & Guo, J. (2009). Analyzing urban bus service
reliability at the stop, route, and network levels. Transportation Research
Part A: Policy and Practice 43(8): 722-734.

Chuanjiao, S., Wei, Z. & Yuanqing, W. (2008). Scheduling Combination and

Headway Optimization of Bus Rapid Transit. Journal of Transportation
Systems Engineering and Information Technology 8(5): 61-67.

Hwe, S. K., Cheung, R. K. & Wan, Y.-w. (2006). Merging bus routes in Hong
Kong's central business district: Analysis and models. Transportation
Research Part A: Policy and Practice 40(10): 918-935.

Ibeas, Á., dell'Olio, L., Alonso, B. & Sainz, O. (2010). Optimizing bus stop
spacing in urban areas. Transportation Research Part E: Logistics and
Transportation Review 46: 446-458.

Lam, S.-W., Tang, L.-C., Goh, T.-N. & Halim, T. (2009). Multiresponse
optimization of dispatch rules for public bus services. Computers &
Industrial Engineering 56(1): 77-86.

Meignan, D., Simonin, O. & Koukam, A. (2007). Simulation and evaluation of
urban bus-networks using a multiagent approach. Simulation Modelling
Practice and Theory 15(6): 659-671.

Nagatani, T. (2001). Interaction between buses and passengers on a bus

route. Physica A: Statistical Mechanics and its Applications 296(1-2): 320-

Oudheusden, van, D. L. & Zhu, W. (1995). Trip frequency scheduling for bus
route management in Bangkok. European Journal of Operational
Research 83(3): 439-451.

Vuchic, V.R., “Urban Public Transportation“. Facsimile ed. 1981: Prentice

Hall College Div.

Wirasinghe, S. C. & Vandebona, U. (2010). Route layout analysis for express

buses. Transportation Research Part C: Emerging Technologies. (in

Yan, S. & Chen, H.-L. (2002). A scheduling model and a solution algorithm
for inter-city bus carriers. Transportation Research Part A: Policy and
Practice 36(9): 805-825.


Amiruddin Ismail1 and Omran Kohzadi Seifabad2 and Mohammad Hesam

Sustainable Urban Transport Research Centre (SUTRA) /

Department of Civil and Structural Engineering

Faculty of Engineering and Built Environment
Universiti Kebangsaan Malaysia
43600 UKM Bangi, Selangor Darul Ehsan, Malaysia
UKM PH.D Student, Lecturer of Faculty of Engineering

Azad Islamic University, Shoshtar, Iran

Error! Hyperlink reference not valid. y

Nowadays, encountering different issues of civilization and living in cities and towns have
created a lot of problems. One of the essential factors in formation, development and
survival of cities and one of the necessities of civilization is the seriously paying attention to
the needs and requirements of human in cities. Among the important requirements of
human, walking and walkways are most important factors which have been forgotten these
days. In fact it is because of paying more attention to the city transportation and its
development. Paying attention to organizing and providing spaces for passers-by particularly
in cities centre, is the most important factors in forming and developing social life of cities.
This indicates that face-to-face encounter between people as passers-by in cities are
regarded as an important issue for such plan. This research considers the method of
sampling including number of passers-by and personal characteristics such as sex and age.
The research methodology is based on observation, questionnaires, comparison of results
obtained, and the analysis of source study and life history. The article tools include
observation, interviews and life history. The conclusion of this study will be about the
identifying, improving and organizing the factors influencing the problem of passers-by in
cities centre which are as the quality of walkway. Furthermore, the role and efficiency of land
use related to walkway and the role of roadway accessing to walkway are considered in

Keywords: Organizing, Improving, Walkway, Passers-by, Roadway, City Centre, Land Use,

The city is the place where people live, communicate and interact. In this place
feeling safe in street is considered to be the understanding of social respect
among people. Therefore the safety of people as passers-by is more important
than the damages caused by accidents. All people of city, either those who use
personal cars or those who use public transportation as the mode of transport
and sometimes walking are also important (Behbahani 2001). So paying
attention to passers-by means respect to all members of the society and not
limited numbers of people. Facilitating the transportation system is not

emphasized, but providing facilities for public transporting and organizing the city
life needs of them are the important objective and goal of planning and designing
the cities roads. Getting information about the passers-by characteristics are
necessary for designing various transporting networks and the services which
should be provided for them. If the buildings have more access to continuously
connected networks which are light, safe and secure, nice, different and distinct,
the passers-by will be encouraged and there will be more people who replace
walking instead of using cars or any public transportation. This fact will cause
less traffic and fewer problems in public transportation.

The continuous human encountering creates these issues in mind that is

there any right for passers-by to have some spaces in the city? And which
section is responsible for dedicating such spaces and improving the city
transportation? Is it possible to improve the quality and quantity of walkway
means of providing correctly acceptable models? To answer such questions, it is
necessary to pay attention and to evaluate the dimensions of the issue (Alfanzo

Before the industrial revolution, people built the cities by the rivers and the
cities shape and texture were based on the geography and the direction of the
rivers. The size of cities also was based on the people ability for travelling and
movement of animals. Therefore all cities and roads were formed based on
social and psychological needs of their people. As the cities become
industrialized and the human need for spaces in cities was ignored, people as
passers-by faced a lot of problems that caused western countries to solve these
problems and to study about it (Soltanzadeh 1992). From 1905 to 1910 a French
engineer called “O. Henard” put forward the plan of making “unleveled
intersection” and various subways and bridges which become popular afterward
(Razaviyan 2002).

In 1928 another civil engineer called “K. Stein” used the theory of
neighborhood unit in more complete form in designing Radburn and suggested
the complete separation of roadway and walking spaces (Ahmadi 2004). An
Iranian citizen imagine street as a straight direction which is covered with
asphalt. Street with canal shape for car movement is not similar in meaning with
a place called city. The consequences of this issues leads to this fact that
walkway and passers-by not to be considered as a important factor in the
urbanity and this is a main problem in cities of Iran (Doroor 2002).

In recent two decades planning and organizing of walkways have been
received great attentions by urbanity engineers. Although the western countries
have studied about passer-by and walkways since last five decades and
unfortunately this studied has been started since one decade in Iran (Ahmadi
The 10th section of designing and planning regulations in Iran set by the
ministry of civil and buildings in 1997 which is called “The walkways” including a
series of rules for applying in city designing. These rules are useful for
improvement of walkway. In 2009 J. Spilkova and M. Hochel, published a paper
about economy of walkway movement. They studied that movement principle is
depends to the particular layout of harmony between actively of land use and
passers-by. Also, powerful attractors such as shaping mall and environmental
quality of walkway are necessary to fluent movement. Also walking is
economically and environmentally the best way of commuting, but the needs of
passers-by are not well addressed (Spilkova & Hochel 2009).

The walkway cannot be attracted by passers-by; as an example, the

passers-by do not use the walkway in city center of Yasuj properly. Most of
people who walk in the city center face problem such as: Lack of having proper
access to educational centers, shops, work places and other city center
activities, lack of safe walking for the purpose of shopping, sport, fun and free
times and lack of the harmony between spaces and activities and the weakness
of actions. These problems are the major ones for each person who want to walk
at least 5 minutes a day. The lack of proper walkway has caused a lot of people
prefer to use automobiles even for short ways.

Today people unwontedly use their personal cars for doing their works in
their city; this will bring environmental, social, economical and health problems
all together. Therefore it is important to understand such issue and make the
public understand it too and try to find solutions (Hosseiniyan 2004).

In this study, some of the important and determining factors in organizing

spaces for walkways are qualitative and quantitative. The experiences show that
if the internal and physical factors are considered along with social, cultural,
psychological and behaviourist factors of people, the organizing and designing
program of walkways will be successful (Behbahani 2001).

The leading objectives of this study in organizing and movement of the

people as passers-by are as follows: The first leading objectives is by well
organizing and improving the walkway to achieve the physical and internal

appearance of city, decreasing pollution also increasing safety and developing
green area. The second leading objective is to identify the influential and
practical factors in movement of passer-by in walkway by determining its social
aspect and its capacity with defined targets. The third leading objective is to
analyze and investigate the problematic issues and to introduce the solutions for
attracting the passers-by. The forth leading objective is to pay attention to the
city centres which are the main parts of cities. Therefore the quality of walkway
are important in seeing and understanding the socially internal life. In fact the
more increase in quality of services and spaces of the passer-by, the better
understanding and discovering people’s psychologist, culture and visions. Thus,
for providing above such conditions for the movement of people as passers-by
and also for improving and using the people’s cultural and visual values, the
centre of Yasuj (one of cities in Iran) has been selected. The areas which are
going to be studied are limited to the walkway of Motahhari and Taleghani
streets which are between Jomhoori Street (late Municipality) and the Clock
square as shown in Figure 1.

in this area for going to work, to do shopping, having fun and walking. The city has been built around thes

The aim of this study is to focus on the quality of walkway in Yasuj city
centre, because the quality of walkway is important in seeing and understanding
the socially internal life. The low quality of walkway in particular the lack of
harmony between walkway and the passer-by is most important to scrutiny in
this study. Four basic and important principles for evaluation and the study of
communication and coordination between passers-by and walkways are
necessary. These important principles that studied are expressed as the
following Table 1.

Table 1. Most important factors which are effective in quality of walkway

General terms Special terms Subset 1 Subset 2

Air pollution
Protection against bad
weather Contamination of floor
Pollution of surfaces
Green space on foot
crossing Side pollution
Lightness of walkway
Pollution of walkway Obstacles in walkway
Human scale of green
Crossing off a part Necessity of Passers-by safety space and city
of life passers-by health furniture
Human scale of Limitation of visibility
on walkway
walkway and speed of passers-
b y

The proportion of
walkway to

Activities of passers- Essential activities
by Optional activities
The amount of space Width of walkway
to stand Useful width of
Capacity of The amount of space
walkways walkway
to sit
The amount of space
for walking

Removing of walkway obstacles and organizing of city furniture

Variety of activities
variation in
Variety of space
Visual variety
Sensory perception Vision or visual perception
improved of Hearing perception
Smelling sense
Tactile sense

1.1. Characteristics of Yasuj city centre
The study area is divided into to five parts as shown in Figure 2. Beginning
and end of each part is apparent with special marks. Concepts of field
studies have been observed locally with the highest accuracy as possible.
Concepts of field studies in five parts are explained in five aspects including
walkway width, the coverage of walkway floor, land use neighbouring
walkway and traffic volume and the level of performance of passers-by.

Figure2. Categorized to five parts in city centre of Yasuj, Iran (A to B)

a)Walkways width

Width of the walkway is one of the important factors affecting the number of
passers-by using walkway. Walkway width doesn’t have same value in different
parts which is ranging from the 3.90 to 5.20 m in the beginning to the end. Width
of walkway space is divided to both live and dead space and this is a very
important topic.

b)The coverage of walkways floor

Floor covering is one of the important factors in attracting or preventing passers-
by for walking. Floor covering walkway can be effective on the city physical
landscape. Floor covering walkway must also be used according to climatic
conditions and the type of adjacent land use.

c) Land use neighbouring (land use spectrum)

In the axis of walkway there are varieties of land use that may be in harmony in
some cases. Land use usually are scattered with inattention to rules of
neighbourhood. Land use the eight have been divided to 8 categories including
1 - Food 2 - commercial Integrated 3 - different stores 4- Miscellaneous services
5 - banks 6- Educational Building 7-office buildings 8- residential buildings.
Distribution to Land use is not uniformly and it is not in order along the walkway.

d) The level of performance and the number of passers-by in the walkways

There are close relationship between performance level with walkway width,
coverage of the floor and the variety of Land use. According of Table 2, the
volume of traffic on the walkway (in research area) during the morning peak
hours were 626 to 2730 passers-by and in the evening peak hours along the
walkway of the 515 to 2965 people have been ranged.

Table2. The volume of pedestrians in research area

Volume passes-
Local gather Time period
Taleghani Avenue north sidewalk of the 9 to 10 am 2730
junction, Hafte-Tir Square 17.45 to 18.45 pm 2965
Taleghani Avenue South sidewalks, 9.15 to 10.15 am 685
crossing Hafte-Tir Square 18 to 19 pm 650
Taleghani Street width, intersection Hafte- 9.45 to 10.45 am 1075
Tir Square 16.30 to 17.30 pm 875
Taleghani Avenue Width, between 7.45 to 8.45 am 1211
vegetable sales street and Sahebaalzaman
10 to 11 am 626
16.45 to 17.45 pm 1203
18 to 19 pm 474
Taleghani Avenue Width, between 9.15 to 10.15 am 690
Sahebaalzaman Mosque Street and
17.45 to 18.45 pm 515

e) Street Lights and traffic lights

All intersections have no walkway crossing lines. Yasuj generally doesn’t have
specific programs written for lining the streets and walkway and traffic signs on
the floor. In the direction walkways studied only traffic lights have been used in
the beginning (point A) and end of that path (point B), which guide only vehicles
and no attention has given to passers-by. In this section, we explained the some
characterizes of Yasuj city centre, and in the next section we will presentation

the methodology of study. The results of the current study could provide
appropriate approach to find out the solutions for the problems studied.


Referring to the subject and the different in human races, cultural, economical,
social, physical and internal dimensions this study will be conducted by
analytically and descriptively. Field study includes observation, surveys, map
consideration, data collection and questionnaires. Library study It is the aim to
analyze, describe and to identify the data by SPSS software and Goller-
scorecard to recognize the functions and relations of factors. Although people as
passers-by are the major focus but there are also some margins to be

2.1. Analysis and discussion about important factors studied

Evaluation of study will be made in two ways.
a)Observational methods (field study) and the comparative method (study
of documents).

In this way, field observations and perceptions of the effects of each factor
based on this study is compare what is existing and what it should be taken.
It also discusses about the results obtained from this study using
international experiences and insights based on urbanity and standards
considered for valuation of the factors.
b)A questionnaire method
In this approach through a questionnaire about important factors filled by
passers-by the factors affecting and its effects on walkways and passers-by
studied has determined and valued.

2.2. How to evaluate important factors

After the observation and investigation of area conditions with the collection
of questionnaires, one of the mathematical way called matrix Guler ( Guller-
scorecard) approach was used for evaluation and determination of
significance coefficient for each factor. Based on the matrix method, each
horizontal row and each column represents a particular rule indicates how to
use vertical rules with a range of degrees including very good, good,
moderate, weak and very weak with the points of five to one respectively

have been graded. The matrix method provides the possibility to
determination of the significance coefficient for each factor. The
multiplication of the degree of importance of each factor in the place of
observation (based on the average of most people and researchers)
determines the points of each factor regarding to its importance. This trend in
each of the specific factor and their subsets was performed and finally the
desired level of harmony between walkway and walking will expressed. In
order to clarify how the determination of the significance coefficient, a
sample of matrix created is presented Table 3.

Table 3. Significance coefficient determination of walkway factors and others

Opinion of people Researcher Opinion Average of
Degree and importance the
factor (PO) (RO) significance
High Average Low High Average Low
(3) (2) (1) (3) (2) (1)

Climate protection * * 2

* * 3
Providing green space
* * 3
Providing immunity

Elimination of * * 3
* * 3
Observation scale

In this table average of the significance coefficients equal with people

opinion plus researcher opinion divide two.
asc = (po+ro)/2 (1)

2.3. Quantitative Approach

In this section according to the principles and fundamental issues raised also
subsets of basic issues, the assessment of all issues which included 20,
important and effective factors were conducted with the same matrix Guler
and with the determination of significance coefficient for each factor, the
percentage of every factor conducted in walkways was 41.7%, for total of 20
important factors (percent of finale result line Table 4). With consideration of
significance coefficient in the walkway and provided good conditions the total
points required for obtaining a good degree of walkway are 240 of 20 factors
whereas Degree of observance factors studied are 102, so this register that

there are more problems about environmental quality of walkway in this
research area.

Table 4. Degree of observance factors affecting the quality of pedestrian environment

Degree of observance factors
Factors studied

5 4 3 2 1
Weather protection 2 * 4 10 40
Provided of green space 3 * 9 15 60
Removal of pollution 3 * 6 15 40
Provided of light 3 * 6 15 40
Obstacle removal 2 * 6 15 40
Attention to human scale 3 * 9 15 60
Limited of visibility and speed 2 * 6 10 60
Total percent observe factors 47.6
Capacity sidewalks 3 * 9 15 60
Afoot activities (elective &
1 * 1 5 20
The amount of space to stand 1 * 2 5 40
The amount of space to sit 3 * 3 15 20
The amount of space for walking 3 * 6 15 40
Street obstacles and organizing
2 * 6 10 60
Total percent observe factors 40
Variety of activities 2 * 4 10 40
Space diversity 2 * 4 10 40
visual diversity 2 * 4 10 40
Total percent observe factors 40
vision or visual perception 3 * 3 15 20
Understanding Hearing 2 * 4 10 40
sense of smell 3 * 6 15 40
tactile perception 2 * 4 10 40
Total percent observance factors 37.5
finale result 47 102 240 41.7

With regard to assessing and implementing the desired axis and according to
the evaluation factors and results and also the status quo and understanding
can be concluded that the assumptions considered correct and numerical results
obtained from each of the factors in its assumptions are true rate.


It is important to consider the situation of all passers-by in providing access.
Organizing and improving according to the walkway’s accessibility for people
should be done according to the different situations of weak people of the
society including old people, children, pregnant women and disabled and those
who carry kids. Walking is the really natural right of each human being including
all ages and situations (Mirdamadi 1995).
The assessment of all issues which included 20, important and effective
factors were conducted with the same matrix Guler and with the determination of
significance coefficient for each factor, the percentage of every factor conducted
in walkways was 41 .7%,for total of 20 factors ( percent of observance studied
factors in finale result line table x ) . With consideration of significance coefficient
in the walkway and provided good conditions the total points required for
obtaining a good degree of walkway are 240 of 20 factors whereas Degree of
observance factors studied are 102, so there are more problems about
environmental quality of walkway in this research area.

There should be some facilities for walking. The people as passers-by

expect safety, security, proper width, proper floor and convenient condition. This
research concentrates on the reconsidering human and on the spaces
(walkways) he needs for walking in the city. In other word it concentrates on
issues such as security and safety, convenience and conformability which were
not r egar ded as a kind of necess it ies for human. This st udy has
Recommendations regarding with exist problems of environment quality that will
be effective to solve them as the following:
- Keep walkways from the bad weathers
- Organizing and order to give the green spaces and trees.
- Organizing furniture urban route study - collected itinerant peddler
- Provide adequate lighting for walkways
- Elimination of barriers to walkway path, modify sex floor directory, use the
appropriate slope use of colour, form, shape, appropriate size
- Removal and prevention of traffic vehicles and designed to Taleghani

- The street as a commercial and recreational space (only certain hours).

- Walkways signpost installed in compliance with standard
- To remove the body of visual crowding passages (cooler and chimneys
- Transfer of some activities and Land use to other parts of town.

- Location and construction of health services over the walkway without
direct access to the walkway.

- Considering the desirability factors, and capacity adjustment factors.

- Use walkway lining to reduce the speed of cars.

- Provide access for passers-by with respect to all of them.

- Get to work on the bridges of communication channels open and the gutter.
There is not one universal remedy for solved walkway problems because there
are varieties of individual, group, regional, and physical environmental factors
come into play.

4. References

Alfonzo, M. (2005). To walk or not to walk? The hierarchy of walking

needs.Environment & Behavior, 37(6), 808-836.

A h m a d i , M . ( 2 0 0 4 ) . An investigation in problem of passers-by in

Tehran .Dissertation, Islamic Azad University of Tehran.

Behbahani, H. (2000). A guide line to city streets . Arkan Publications.Esfahan.

Behbahani, H. (2000). The advanced systems for organizing, managing and

maintaining the city roads and streets . Journals of Traffic, Summer.

Doroor, A. (2002). The cover of streets . Technical & Social Journal, 21.

Hosseiniyon, S. (2004). View and vision in the highways . Municipality Journals, 48.

Mehdizadeh, J. (1999). The concepts and methods of making side walks .

Municipality Journals, 19.
Mirdamadi,M. (1995). An introduction to road making in Iran , Tehran.
Razaviyan, M.T. (2002). The programming of city land functions . Monshi
Publications. Tehran.

Soltanzadeh, H. (1986). An introduction to the history of city and urbanization in

Iran. Abi Publications. Tehran.

Soltanzadeh, H. (1991). City atmosphere in the historical texture of Iran . Tehran:

The Office of Cultural Research of Municipality.

Spilkova, J. & Hochel, M. (2009). Toward the economy of walkway movement in

Czech and Solvak shopping malls . Environment & Behavior, 41, (3), 443-445.

Tavasoli, M. (2000). City designing, concepts and principles . Abadi Journals, 25.

Mukti Advani, 2B.Kanagadurai

Scientist, 2Head – RDM Division


Central Road Research Institute, New Delhi,

Present study focuses on pedestrian crossing infrastructures which are majorly constructed
on the basis of number of commuters crossing the road and the delay caused to motorised
vehicles due to such crossings. However, most road crossing infrastructures developed for
pedestrian crossings increases the crossing time of pedestrians no lesser than 3 to 4 times
compare to at grade signalised road crossing time. A methodology has been evolved to
evaluate the decision regarding the choice among different pedestrian crossing facilities
to be provided. For this evaluation, sustainability of different types of vehicles also has been
considered to give them priority. Three different scenarios have been developed based on
different sustainability factors which are based on vehicle occupancy and space occupied on
road. Study includes careful observation of existing facilities provided for crossings on a
corridor of Delhi and applied to test the methodology.

Keywords: Pedestrians, Sustainabil ity factor, infrastructure

1 Introduction
Increase in number of vehicle population has lead to traffic congestions along with
the delays at intersections. In order to facilitate the faster movement of vehicles in
Delhi, road infrastructure facilities such as flyovers, signalised junctions, subways,
and foot-over-bridges are being developed. In the past eight years 37 flyovers have
been completed and currently more than 16 flyovers and rail over bridges are under
construction. Apart from these; another 20 flyovers and over-bridges are in the
planning stage. The major arterial roads are being made as through roads with less
number of intersections in order to provide faster movement and reduced the traffic
congestions. While planning for such road infrastructure facilities to decongest, it is
necessary to provide sufficient space and accessibility facilities for sustainable
modes such as cycle and walk trips. Indian Road Congress (IRC 103-1988)
provides guidelines for pedestrian facilities. This mentions two types of grade
separated pedestrian facilities i.e. (i) Pedestrian subways and (ii) Foot over bridges
(FOB). To make the movement of pedestrians more comfortable on such roads, the
concept of foot over bridges (FOBs) with escalators was also being introduced in
eight such FOBs currently under construction and seven more at the planning stage
(DelhiLive, 2010).

Since, each trip made by public transport involves walking; facilities provided for
walking along the roads (i.e. footpaths) and crossings at the interchanges are the
most important part of the road infrastructure. However, UTTIPEC (2009) mentions
that in Delhi; 40% roads do not have any footpath. Further, the study on walkability
for pedestrian facilities by Parida at. el, (2008) shows that at least 40% of the
studied pedestrian facilities are not suitable for walking and would need
improvement. On the other hand the accident data shows that nearly half of fatal
accidents in Delhi involve pedestrians. This indicates that there is a need of
reviewing existing facilities to identify the gap between facilities provided and their
suitability for commuters. Present study includes the suitability of pedestrian
crossing facilities provided on the study corridor.

2 Crossing facilities for pedestrians – Study corridor of Delhi

Present study analyse a stretch of 7 km; part of inner ring road (Mahatma Gandhi
road) in Delhi. Land use of this area mainly involves commercial and institutional
and partially residential in nature. Figure 1 shows the corridor on a clip of Google

Figure 1: Study corridor and surrounded land cover

The study corridor includes 12 grade separated pedestrian facilities i.e. 2 foot-over-
bridges, 8 subways and 2 more crossing facilities for pedestrians lying over the
underpass in this corridor. This corridor is full of different types of road infrastructure
and specially to provide crossing facilities to the pedestrians for crossing the road.
More details of this have been shown in Table 1.

Table 1: Different road infrastructure in study corridor
Sr. Structure type Chainage Chainage Pedestrian Remark
No and name at the at the end and cycle flow
. start per 5 minutes
1 Foot over 0 50 meter 170 Stairs, ramp and
bridge ‘A’ pedestrians, 3 escalator provided.
Maharani bagh cyclists No at-grade
crossings unless
heavy congestion
observed - 56
pedestrians at grade
2 Foot over 900 meter 950 meter 40 pedestrians Stairs and ramp
bridge ‘B’ (26 school provided. No at-
children) grade crossings
possible due to
3 Lajpat nagar 1.4 km 2.6 km
4 3-arm 1.7 km - 73 pedestrians 28 pedestrians and 5
Subway ‘A’ (67 school cyclists at grade
5 3-arm 1.9 km - 9 20 pedestrians at
Subway ‘B’ grade
6 3-arm 2.3 km - nil 18 pedestrians and 8
Subway ‘C’ cyclists at grade
7 Subway ‘D’ 2.9 km - 92 pedestrians No at-grade
and 1 cyclist crossings possible
due to railing
8 Moolchand 3.3 km 3.8 km
9 Crossing facility 5 pedestrians No at-grade
above under crossings possible
10 Crossing facility 3 pedestrians No at-grade
above under crossings possible
11 Subway ‘E’ 4.0 km 4.1 km 70 pedestrians No at-grade
+ 2 cyclists crossings possible
due to railing
12 South ex. 4.2 km 4.8 km No at-grade
Flyover1 crossings possible
due to railing
13 Subway ‘F’ 5.1 km 5.15 km 303 No at-grade
pedestrians crossings possible
due to railing
14 South ex. 5.5 km 5.6 km
15 Subway ‘G’ 5.9 km 6.95 km 104 No at-grade
AIIMS pedestrians crossings possible
due to railing
16 AIIMS 6.1 km 6.7 km
17 Safdarjung 7.0 km 7.05 km 77 pedestrians

Table 1 shows that there are 12 crossing facilities provided by design to the
pedestrians and/or cyclists. However, pedestrian counts at these locations shows
that wherever at-grade crossings are possible; pedestrians have chosen this for
crossing the road compare to the subway crossing. Pedestrian flow at Foot-over-
bridge at Maharani bagh location has observed as 170 ped/5 min. Crossings at this
location has been restricted through railing on both the sides of FOB. This railing
has been broken and at the time of congestion pedestrians observed to be crossing
through this opening. All 3-arm subways provided are at the interchanges below the
single 1.2 kilometre long elevated flyover. In one of these 3 subways; it has been
observed no pedestrians at all. However, at the same location there is a pedestrian
flow of 18 along with the 8 cyclists has been observed in 5 minutes duration. This
indicates that the first preference for crossing the road remains at-grade level.
However, through enforcement which has been applied by designing railings,
flyovers and underpasses. This shows that existing facilities can be further modified
through design and location to make them more suitable for the commuters. Since,
each public transport user is also a pedestrian; facilities for pedestrians will further
help in higher public transport use.

3 Impact of infrastructures on sustainable modes

There are bus stops on ring road which have been shifted after construction of
flyovers. This makes an extra meter of walking involving physical discomfort to the
commuters. Bus stop location should be on need based and it should be facilitated
with proper crossing facilities. Wherever choice between public and private vehicles
facility occurs; public transport has to be prioritised without any doubt. Table 2
shows that constructed infrastructures are motorised vehicle oriented. Focus of this
infrastructure is primarily for car and two-wheeler users. While reducing the travel
time for these private vehicles either ignores the facility to be provided to the
sustainable modes (i.e. walk, rickshaw, cycle and public transport) or wherever
provided is not suitable. This is an alarming stage of road infrastructure
development, since lots of money resources have been used to build such
infrastructure; it is very essential to check their effect in all respect. Level of Service
(LOS) measure for pedestrians and bicycles depends on ‘Event Delay’ in HCM,
2000. However, it’s not only the delay but more importantly is the physical effort
(human energy) attached to this non motorised (sustainable) way of making a trip.

Table 2 presents a broad comparison of impact of different infrastructures on
private, public and other non mortised traffic modes.
Table 2: Impact of road infrastructure on sustainable modes
infrastructure Effect on cars and two- Effect on pedestrian, cyclist, cycle
wheelers rickshaws and bus commuters
(Sustainable modes)
Flyover and Under Red signal avoided for Extra physical effort for
pass main stream commuters walking and riding bicycle on
i.e. primarily motorised flyover
private vehicles
For perpendicular movement,
limited access points
Shifting of bus stops which
may result in longer walking
distance for bus commuters
Clover leaf flyovers smooth driving at major No at-grade crossings by
intersections design; increase in crossing
Foot over bridge Smoother driving Crossing with additional
physical effort
Subway Smoother driving Crossing with additional
physical effort
An attempt has also been made to conceptualise a methodology to evaluate
different options of the road infrastructure facilities.

4 Methodology for evaluation of pedestrian crossing facilities

Road infrastructural projects usually have been justified based on cost-benefit

analysis and the prime concern for this analysis used to be the vehicle operating
cost savings. Elvik (2000) mentions that impacts of road schemes specifically
designed for pedestrians and cyclists are not fully included in current cost benefit
analyses. IRC (1988) provides the guidelines for pedestrian crossing facilities.
According to IRC (103, 1998),
“Provision of a grade separated pedestrian facility may be warranted at locations
where one or more of the following conditions exist:

i) Volumes of pedestrians and vehicular traffic are so large that insertion of

an exclusive pedestrian phase will increase the cycle time for traffic
signals beyond 120 seconds
ii) Vehicular traffic demands uninterrupted flow as associated with major
arterial roads and expressways;

iii) Control at-grade pedestrian crossing decisively fails to mitigate the

problems of pedestrian-vehicle collision. Viability of a grade separated
pedestrians facility must be checked again delay costs for both

pedestrians and vehicle drivers/users including increase in vehicle
operating costs inflicted by increased delays .”

(IRC-103, 1998)
As mentioned above; two conditions (i and ii) indicates that grade separated
pedestrian facilities should be provided for uninterrupted flow of vehicular flow.
However, third conditions include the delay costs for both pedestrians and vehicle
drivers/users. Present study develops a methodology to compare this delay cost for
pedestrians and vehicle users. This also involves the sustainability factor for
different types of vehicles. Sustainability factors have been calculated based on
vehicle occupancy and space occupied by different vehicles on road. However, time
delay remains the prime focus of evaluation in all developed scenarios. Total time
taken for crossing by each commuter has been estimated to drive an optimum
solution for a given facilities. Pedestrian crossing flow and main road vehicle flow
has been observed first and then the total time saving has been calculated for this
traffic flow. To check the suitability of different crossing facilities, these two have
been compared. Finally, one with the higher time saving has been chosen for priority
infrastructure. This methodology assigns higher weight-age for sustainable modes
through sustainability factor (SF) consideration. TSMRC = Time Saving of Main
Road Commuters

TSMRC = Time required to cross the distance without any stopping (FOB or Subway
provided) - Time required crossing the distance in case of at-grade signalised

where; n = car, motorised two-wheeler, bus, auto, taxi

TSCC = Time Saving of Crossing Commuters
TSCC = Time required to cross the distance through Subway or FOB - Time
required crossing the distance in case of at-grade signalised crossing

where; n = pedestrians, cyclists, cycle rickshaw

To include higher weight-age for sustainable modes i.e. pedestrians, cyclist, cycle
rickshaws and public transport; multiplying factors have been added i.e. SF =

Sust ainabi lit y f act or s. TSM RC =

i = vehicle type i.e. car, motorised two-wheeler, bus, auto, taxi
j = number of vehicles of type i
SF=Sustainability Factor
i = vehicle type i.e. pedestrians, cyclists, cycle rickshaw
j = number of vehicles of type i
SF = Sustainability Factor
If TSMRC > TSCC; Provide Subway or FOB
If TSCC > TSMRC; Provide safe at-grade crossing facility
Further, these equations have been modified based on sustainability factor
(scenario 1) and as a which is a function of vehicle occupancy (scenario 2) and
space occupied on road (scenario 3).

5 Data analysis and results

CRRI (2010) provides the observed traffic volume count on ring road at Defence
colony near Andrews Ganj flyover. It has been assumed that traffic flow in study
corridor is same as of at defence colony which is km away from the study area.

Since, time required for road crossing in case of signalised intersection depends on
the actual signal design and signal cycle timings for each arm of the intersection;
delay time may vary accordingly. HCM (2000) states that when pedestrians
experience more than a 30 seconds delay, they become impatient and engage in
risk taking behaviour. Therefore, in present case assuming that signals have been
designed ideally i.e. pedestrian waiting time is no more than 30 seconds and the
average are 15 seconds. Similarly, for motorised vehicles, average wait time is 15
seconds and delay time due to signalised intersection including acceleration-
deceleration time is 10 seconds. Therefore, total delay is considered as 25 seconds.

The estimation of time saving for traffic on selected corridor (i.e. ring road) for main
road traffic (TSMRC) and road crossing commuters (TSCC) has been calculated for
different scenarios which are based on different sustainability factors as listed

Scenario 1: Time delay calculation considering the sustainability factors as: Non
motorised vehicles=1, Private Motorised vehicles =0, Public Motorised vehicles=0.5
Table 3 presents the calculation for scenario 1 regarding the time saving of main
road commuters.

Table 3: Estimation of time saving for traffic on ring road (TSMRC

*Traffic Traffic Time
Sustainability saving
*Vehicle type Counts counts saving
Factor (SF) considering
(16 hours) (5 minutes) (seconds)
Small car 64659 337 8425 0 0
Big car 43129 225 5625 0 0
Scooter 9179 48 1200 0 0
Motor cycle 63734 332 8300 0 0
26488 138 3450 0 0
Bus 7274 39 975 1 975
Mini bus 542 3 75 1 75
LCV and
3543 18 450 0.5 225
goods auto
HCV & MCV 2315 12 300 0.5 150
bicycle 507 3 75 1 75
84 1 25 1 25
and other
Total 28900 1525
*Source: CRRI (2010)

Table 4 presents the time saving details of crossing commuters (TSCC) at different
locations within the study corridor.

Table 4: Comparison of time saved by main road commuters (TSMRC) and crossing
commuters (TSCC)
Crossing Time based on
plus cycle Sustainability TSMRC
Structure saving TSCC
flow (5 Factor (SF) value of
Foot over
At grade
bridge ‘A’
173 2595 1 2595 crossing is
At Maharani
Foot over
40 600 1 600 -
bridge ‘B’
Subway 'A' 73 1095 1 1095 -
Subway 'D' 93 1395 1 1395 -
Subway ‘E’ 72 1080 1 1080 -
Subway ‘F’ At grade
South Ex 303 4545 1 4545 crossing is
area proposed
Subway ‘G’ At grade
South Ex 104 1560 1 1560 crossing is
area proposed
Subway 'H' 77 1155 1 1155 -
Scenario2: Time delay calculation considering the sustainability factors as a
function of vehicle occupancy as: Non motorised vehicles=1, Private motorised
vehicles = (vehicle occupancy)/vehicle capacity, Public motorised vehicles = more
than 20=1, between 20 and 10=0.5, less than 10=vehicle occupancy/capacity. Table
5 presents the calculation for scenario 2.
Table 5: scenario 2 based on vehicle occupancy
Average Vehicle TSMRC
*Traffic Traffic Vehicle capacity Time
Time Factor (SF)
*Vehicle Counts counts occupancy saving
saving (vehicle
type (16 (5 considering
(seconds) occupancy/vehicle
hours) minutes) SF
Small car 64659 337 8425 1.3 5 0.26 2190
Big car 43129 225 5625 1.3 5 0.26 1462
Scooter 9179 48 1200 1.2 2 0.26 312
Motor cycle 63734 332 8300 1.2 2 0.6 4980
Auto 2.5 4
26488 138 3450 0.62 0
Bus 7274 39 975 43.2 54 0.8 975
Mini bus 542 3 75 29.6 37 0.8 75
LCV and - -
3543 18 450 0.5 225
goods auto
HCV & MCV 2315 12 300 - - 0.5 150
bicycle 507 3 75 - - 1 75
Rickshaw - -
84 1 25 1 25
and other
Total 28900 9639
*average load factor considered for bus is 0.8
Scenario 3: Time delay calculation considering the sustainability factors as a
function of space occupied on road by different vehicles and vehicle occupancy.
Table 5 presents the calculation sustainability factor for the scenario no 3.

Table 5 presents the calculation of sustainability factor

Average space Average
Space Factor
Vehicle Type occupied on vehicle
occupied on (normalised
road ( ) occupancy
road per person space/person)
= 1-(1.41/1.54)
Small car 1.83 1.3 1.41
= 0.08
Big car 2.00 1.3 1.54 (max) 0
Scooter 1.10 1.2 0.92 0.40

Motor cycle 1.00 1.2 0.83 0.46
Auto rickshaw 1.35 2.5-1 (driver) 0.9 0.42
Bus 24.25 60 0.40 0.74
Mini bus 18 29.6 0.61 0.60
LCV and goods 0.50
- - 0.5
HCV & MCV - - 0.5 0.50
bicycle - - 1 1
Rickshaw and 1
- - 1
Table 6 presents the time delay calculation for scenario 3.
Table 6: scenario 3 based on space occupancy on road and vehicle occupancy
*Traffic Traffic Time
*Vehicle Counts counts Sustainability saving
type (16 (5 Factor (SF) considering
hours) minutes) SF
Small car 64659 337 8425 0.08 674
Big car 43129 225 5625 0 0
Scooter 9179 48 1200 0.40 480
Motor 0.46
63734 332 8300 3818
Auto 0.42
26488 138 3450 1449
Bus 7274 39 975 0.74 721.5
Mini bus 542 3 75 0.60 45
LCV and 0.50
goods 3543 18 450 225
HCV & 0.50
2315 12 300 150
bicycle 507 3 75 1 25
Rickshaw 1
84 1 25 25
and other
Total 28900 7612.5

In present study, at three locations at-grade crossing facility has been

recommended as shown in table 4 i.e. at Maharani Bagh and two in the area of
South Extension according to scenario 1.. Since, escalator provided at Maharani
Bagh has not been used efficiently for both the sides; facility suitable to crossing
commuters has been required to promote pedestrian movements. Provision of
escalator in all the directions will reduce the total time for more road crossing
pedestrians. In this case crossing time will lesser than the present time required for
crossing the road.

Subways provided at south extension are most used crossing facilities and therefore
more commuters are putting higher effort for crossing the road. It has been

suggested to provide an at-grade crossing facility which will reduce the physical
effort requirements of the sustainable modes i.e. primarily pedestrians.

In case of scenario 2 i.e. sustainability factor as a function of vehicle occupancy,

TSMRC is 9639. This do not justify for at-grade signalised crossing facilities on
studied corridor. Similarly TSMRC value of 7612 of scenario 3 also does not
recommend at-grade signalised crossing facilities for present traffic flow. However, it
is important to note here that increased crossing commuters’ flow leads to higher
TSCC and which demands at-grade crossing facilities. However, in practice this is
completely reversed i.e. wherever higher crossing flow is available, other than at-
grade facilities used to be recommended.

Developed methodology can be further modified to include different type of people

crossing the road i.e. school children, elderly people, disabled, bus commuters
crossing the road, etc

6 Conclusion
Suitability of road infrastructure for sustainable modes of travel is at the priority
concern. Easy, comfortable and safe movement for pedestrians and non motorised
vehicles can be provided by proper design of road environment. Since; pedestrians,
cyclist and cycle rickshaw trip involves physical energy, comfort and limitations of
human body have to be incorporate by design. Each alternate way of crossing the
road should be compared to the direct crossing at road. Present study develops a
methodology to do such comparison among different alternatives of road
infrastructures. This methodology is based on scenario of vehicle occupancy and
space occupied on road. However, this can be further modified to include different
group of people based on gender, age, disability to prioritise sustainable modes.

CRRI (2010) ‘Traffic studies and estimation of travel demand for phase III of
Delhi metro’ Final Report.

DelhiLive (2010) ‘More flyovers for Delhi’ accessed on 15th March

• Elvik (2000) ‘Which are the relevant costs and benefits of road safety
measures designed for pedestrians and cyclists?’ Accident Analysis and
Prevention, Volume 32, page: 34-45.

Highway Capacity Manual (2000). National Research Council, Washington,


IRC (1988), Indian road Congress – code 103, Guidelines for Pedestrian
facilities’, New Delhi, India, 1988.

Parida P. (2008) ‘Design and safety of pedestrian facilities’, Urban Mobility

India, New Delhi, India.

UTTIPEC (2009) ‘Pedestrian Design Guidelines’ Unified Traffic and

Transportation Infrastructure (Planning and Engineering) Centre, Delhi
Development Authority, New Delhi, India.


Zohreh Asadi Shekari 1

and Muhammad Zaly Shah 2*

Ph.D. candidate in transportation planning. Department of Urban and Regional


Planning, Faculty of Built Environment, Universiti Teknologi Malaysia, 81310 Skudai,

Johor. Email:
Corresponding author. Senior Lecturer. Department of Urban and Regional

Planning, Faculty of Built Environment, Universiti Teknologi Malaysia, 81310 Skudai,

Johor. Email:

ABSTRACT: Streets are prominent indicators for liveability of our communities and they
should be for all people. But lots of them are designed just for cars and selected users. The
intention of this paper is to encourage planners and engineers to design complete streets, or
streets that consider the needs of all urban residents. This article presents several indicators
from various guidelines which consider different pedestrian needs to introduce complete
street level of service for pedestrian and finds out existing streets failures. This evaluation
method can be used for different types of streets and shows their designing weakness for
pedestrian. So this unique implemental pedestrian level of service (PLOS) is useful for
improving existing streets and produces a complete guideline for designing new streets in
case of pedestrian.
Keywords : pedestrian, indicators, level of services, urban streets


A measure for describing existing conditions for each user in streets can be defined
as level of service (LOS). These conditions can be travel time, speed, traffic
interruptions, convenience, free maneuver, safety and comfort. Highway capacity
manual (HCM, 2000) introduced pedestrian level of service as follows. PLOS A is
free flow and maximum speed, PLOS F is no movement and PLOS E is congested
situation. Highway capacity manual is criticized by many researches for considering
pedestrians behaviour like vehicles (Landis et al, 2001; Mcleod, 2000). Comfort,
aesthetic and convenience are considered by various researches in case of PLOS
(for instance, Bradshaw, 1993; Jaskiewicz, 2000; Khisty, 1994; Landis et al., 2001;
Leslie et al., 2004; Hines and Steinman, 2004). Ease of Crossing Street, choosing
suitable speed and convenience for passing others, climate control, weather
protection, transit shelter, arcades, walking distance, grades, directional signing,
sidewalk ramps and pathway directness can be defined as pedestrian level of
service (Henson, 2000). Philips, et al (2001) considered pedestrian level of comfort
and safety instead of expected density of the facilities, width and pedestrian volume.
This paper presents level of service for pedestrian, using several indicators from

various guidelines which consider different pedestrian needs and finds out existing
streets failures. Different types of streets and their designing weakness for
pedestrian can be evaluated by this unique implemental model. So this pedestrian
level of service (PLOS) is useful for improving existing streets and produces a
complete guideline for designing new streets in case of pedestrian.

2.1 Data and case studies
The data that are used in this study were selected from 20 guidelines (1- City of
Ottawa (Canada, 2008), 2- Calgary City (Canada, 2008), 3- street design guidelines
for Landcom (Australia), 4- city of Whittlesea (Australia, 2009), 5- New South Wales
(Australia, 2001), 6- city of Charles Sturt (Australia, 2009), 7- city of Chicago (USA,
2007), 8- Portland pedestrian design guideline (USA, 1998), 9- city of Ashland
(USA, 1999), 10- design guideline for streets and sidewalk (Minneapolis, USA,
2008), 11- Charlotte’s Streets (USA, 2007), 12- subdivision and development street
standards (Pima County, USA, 2005), 13- neighborhood street design guidelines
(Oregon, USA, 2000), 14- Aurora Urban street standards (USA, 2007), 15- street
design guidelines for healthy neighborhoods (USA), 16- accessible sidewalks and
street crossing (USA), 17- pedestrian design guidelines (India,2009), 18- street
design manual (New York,USA,2009), 19- Mn/DOT road design manual
(Minnesota,USA,2010) and 20- mobility master plan bicycle and pedestrian design
gu idel ines(Tacoma, USA,2009)) in various developed countries. This selection from
different parts of world can help to evaluate pedestrian indicators in different socio-
economic context. This study also uses 20 factors as pedestrian indicators. Two
case studies are used in this research. One of them is a street in Setia Tropika area
in Johor Bahru Malaysia and the other one is Canberra road in Sembawang area in
Singapore. Both of these case studies are collector urban streets. 1790 m of Setia
Tropika and 724 m of Canberra road were observed in this research.

2.2 Analysis
This study evaluates (1) slower traffic speed, (2) buffer and barriers (curb and
furnishing zone), (3) fewer traffic lane, (4) shorter crossing distance (curb
extension), (5) mid block crossing, (6) social space (café), (7) landscape and tree,
(8) facility (fire hydrant), (9) furniture (trash receptacle), (10) footpath pavement ,(1 1)
marking (crosswalk), (12) pedestrian refuge and median, (13) corner island, (14)
sidewalk on both sides, (15) advance stop bar, (16) width of footpath, (17) driveway,
(18) lighting, (19) signing and (20) bollards as pedestrian indicators for PLOS. So
mathematically PLOS can be defined as follows. Refer Eq. (1)

PI= Pedestrian indicators

This formula is right when it is assumed that all indicators have the same priority for
PLOS. But to have more reliable results for PLOS, priority of indicators should be
considered. So the formula can be changed as follows. Refer Eq. (2)

= coefficient of each pedestrian indicator

This coefficient can show the strength of each pedestrian indicator in case of PLOS
so the priority and importance of each indicator can be indicated by pedestrian
indicator coefficient. This coefficient can be calculated by evaluating the importance
of indicator for different guidelines. So mathematically this coefficient can be defined
as follows. Refer Eq. (3)

D = Depth of evaluation

= Number of guidelines that evaluate indicator i with depth of evaluation j

Depth of evaluation can show the quality of evaluation for each indicator in different
guidelines. For instance guideline A just suggests that indicator I should be
considered in sidewalks but guideline B has more research about this indicator and

presents some descriptions and guideline C indicates standards and complete
instructions for indicator I. So depth of evaluation should be different for each
indicator in various guidelines. Depth of evaluation can be defined as follows.

= 1(not complete) = 2 (semi complete) = 3 (complete)

Table 1 presents depth of evaluation for each indicator and guideline (D chart) and
Table 2 shows number of guidelines that evaluate indicator i with depth of evaluation
j( . So, the coefficient of each pedestrian indicator can be defined as in

Eq. (3) (refer also to Table 2):

= (1×7) + (2×0) + (3×10) = 37, = 38, = 15, = 28, = 32, = 14, = 38,
= 17, = 16, = 32, = 22, = 32, = 15 = 39, = 9, = 56,
=23, = 31, = 24, = 18

Now, all coefficients of pedestrian indicators are recalculated so just s are

needed. The revised PLOS based on Eq. (1) is as in Eq. (2):

can be achieved by comparing existing street condition with standards of

guidelines that are combined together for each indicator. This combined guideline
can be defined as complete street guideline in case of pedestrian. is a number
between 0 and 1. The best fitness between existing condition and pedestrian
complete street guideline can achieve the highest point that is 1 and no fitness is 0.
There are also some points for conditions that have semi fitness between 0 and
1.Table 3 indicates how is calculated for each indicator. Percentage of PLOS
can be indicated as follows. Refer Eq. (4)

And PLOS degree can be defined as follows.

A: 80-100 B: 60-79 C: 40-59 D: 20-39 E: 1-19 F: 0

2.3 Analysis results

The results of case studies analysis are drawn in this part so each section of this
part evaluates each indicator for both case studies and finally percentage of PLOS
and PLOS degree are estimated. All pedestrian indicators for both case studies are
evaluated by Table 3 and fitness between existing condition and pedestrian
complete street guideline can be achieved by Table 3. The highest point that is 1
and no fitness is 0. There are also some points for conditions that have semi fitness
between 0 and 1. All that are calculated for both case studies are shown in
Table 3. For first indicator average speed can be calculated by Radar Gun data but
in Canberra road using Radar gun was not possible so vehicles speed is estimated
by using test car method and average speed was less than 50 Km/h so in case of
speed this street has good fitness with standards.

Now all coefficients of pedestrian indicators and s are calculated so PLOS can be
evaluated by PLOS formula. Refer Eq. (2) and Eq. (4)

So for Setia Tropika PLOS can be defined as follows.

PLOS= (37 × 1) + (38 × 0.79) + (15 × 0.5) + (28 × 0) + (32 × 0) + (14 × 0) + (38 ×
0.62) + (17 × 0) + (16 × 0.25) + (32 × 0.78) + (22 × 0) + (32 × 0) + (15 × 0) + (39 × 1)
+ (9 × 0) + (56 × 0.78) + (23 × 1) + (31 × 0) + (24 × 0.5) + (18 × 0) = 244.72

And = 536 and

So PLOS% = (244.72 / 536) × 100 = 46 Therefore PLOS grade for this street is C

For Canberra PLOS = 351.45 So PLOS% = (351.45 / 536) × 100 = 66 Therefore
PLOS grade for this street is B

Streets are prominent parts of our society and they belong to all people not just to
cars and drivers. This study tries to introduce complete streets for pedestrian. This
implemental evaluation of pedestrian level of service (PLOS) can help to evaluate
factors that influence walking trips based on different guidelines. This study
considers lots of pedestrian indicators and 20 guidelines in developed countries so
more indicators and guidelines can enhance accuracy of PLOS that is achieved by
this method. This PLOS presents existing street failures and also indicates
improvements issues to achieve higher PLOS degree. Therefore this method is a
way to have complete street that consider all needs of pedestrian. Improvements
issues for case studies are defined as follows.

Improvements for pedestrian in Setia Tropika Street can be defined as follows. At

least for 980 m of street furnishing zone adjacent to the curb should be considered
and for 1232 m of street the wide of furnishing zone adjacent to the curb should be
increased from 1.5 m to 1.8 m. One travel lane on both side of street should be
converted to on street parking and enough curb extension should be considered. At
least 14 mid block crossing should be considered. There should be some social
spaces like café in this street. Distances between trees should be decreased to 9 m
and trees should be removed from standard distance limitation at 2 intersections.
Trash receptacles should be provided for 2680 m of street. Footpath width should be
increased to 1.8 m. At least 36 crosswalks with standard marking and advance stop
bar should be provided. At least 20 pedestrian refuges and medians should be
considered in this street. All Corner Islands should be converted to standard
condition. All length of street should be covered by standard pedestrian lighting.
Transit and crossing signs should be provided. At least 112 rows of bollards should
be considered. Improvements for pedestrian in Canberra road can be defined like
Setia Tropika by using scores weaknesses to achieve higher PLOS degree.
Based on coefficient of each pedestrian indicator there are some factors that have
more impacts on PLOS in this method for instance width of footpath has the most
effective coefficient for PLOS and advance stop bar has the least. Therefore
improvements for indicators should be prioritised from the greatest to the lowest

effective coefficients. Figure 1 indicates the coefficient for each pedestrian indicator
in this study.
60 50 40 30 20 10 0

Figure 1. Coefficient of pedestrian indicators chart

Table 1. D chart

Guidelines Indicators
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20
1 1 2 1 2 2 3 1 1 1 3 2 1 2 1 1
2 1 2 2 1 1 1 3 1 1
3 1 1 1 2 1 1 3 3 1 1 1
4 2 2 1 1 2 1 2 2 1 2
152 3 4 3 1 1
5 6 7 8 9 1011121314151617181920 3 1
6 3 1 1 1 3 2 2
7 1 3 1 2 of
Coefficient 2 pedestrian
2 2 2
indicators 3 3 2 1 2
8 3 3 3 1 2 3 3 3 1 1 1 3 3 3 2 3 2
9 3 3 1 2 2 1 3 1 1 3 3 3 1 2
10 3 3 1 3 3 3 2 3 2 3 3 1 2 3 3 3 3 3 3
11 3 2 3 3 2 1 3 2 3 2 3 2 3 2 2 1
12 3 2 1 3 2 3 3 2 1
13 1 1 2 3 2
14 3 2 2 2 1 1 1 1 1 2 3 1 3 1 1
15 1 1 1 1 2 1 2 3 3 1
16 1 2 2 2 1 1 1 3 1 2 2 3 1
17 3 3 3 1 3 3 1 2 3 3 3 2 3 3 3 3 3 3 2
18 3 2 2 3 2 2 3 1 2 3 1 2 3 2 1 3 1 2
19 1 2 1 2 3 1 2 1 2 2 3 3 3 3 3 2 1 1
20 3 3 3 2 3 1 2 1 1 3 3 3 2 3 2 3 2 3 2 2

D = 3 (Complete) D = 2 (Semi complete) D = 1 (Not complete)

Table 2. Nij

Depth of Indicators
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20
1 7 4 4 4 3 7 3 9 5 6 7 3 2 1 2 5 11 4
2 8 1 6 7 2 10 1 4 1 3 7 5 1 4 3 7 2 7
3 10 6 3 4 5 1 5 2 1 8 3 5 1 13 2 16 5 4 3

Table 3. PI i data for Setia Tropika, Johor Bahru and Canberra Road, Sin gapor

(1) Slower traffic speed (speed) Set i a t r opi ka

A: 38.01 Km/h PI1=1 speed <=50
Standard: 50-60 km/h Canberra road 50 Km/h
speed limitation PI1=1
speed <=50

(2) Buffer and barriers Set i a t r opi ka

CL=3444m N1=3444m
Standard: 1- Min width for curb 0.15m, min height 0.10- CI=1
0.15 m C1=0 C2=2464x1.5=3696m2
Q: standard curb length? W1=0 W2=1.5
A:CL N2(1 )=490x2x 1.8=1764
N1= length of curb that street needs FI1=0/1764=0
CI= CL/N1 N2(2)=(1062+170)x2x1 .8=4435.2
2-Furnishing zone between 1.8 and 2.4 m width adjacent to FI2=3696/4435.2=0.83
the curb FI=(FIC1xL1+FIC2xL2)/3444=
Q: Area of furnishing zone adjacent to the curb? (0 x 980+0.83x2464)/3444=0.59
A: C PI2=(1+0.59)/2=0.79
W=Wide of furnishing zone adjacent to the curb Canber r a r oad
CL=1367m N1=1367m
C1=280 x5.5=1540m2
W1 =5.5 W2=1 .3
N2(1)=280 x5.5=1540m2
FI1=1 N2(2)=1087x1.8=1956.6
FI=(FIC1 xL1 +FIC2xL2)/1 367=
(1x 280+0.72x1087)/1367=0.78
PI2=(1 +0.78)/2=0.89
(3) Fewer traffic lane (number of travel lane) Set i a t r opi ka
Standard: 2-3-4-5 lanes Number of lanes=4 PI3=0.5
Canberra road
Number of lanes=4 PI3=0.5

PI3= 1 No lane<=2
(4) Shorter crossing distance (curb extension) Set i a t r opi ka
No on street parking
Standard: 1-For any crosswalk where on street parking is
No curb extension
considered whenever possible in pedestrian oriented areas
2-Should not block the bike line and turning bus and truck PI4=0
Canberra road No on
street parking
No curb extension

(5) mid block crossing Set i a t r opi ka

n1= 490/120=4
Standard: 1-Crosswalk frequency :various but not farther n2=1062/120=9
apart than 60-90 m and not closer than 45 m and do not n3= 170/120=1
prohibit crossing for more than 120 m N=4+9+1=14
2- Typical width of mid block crossing is 3 m but where the C=0
sidewalk is wider than 3.7m the crosswalk maybe wider than P=0/14=0
standard width to match the sidewalk PI5=0
Q: Number of standard mid block crossing? A:C Canber r a r oad
n1= 256/120=2
C=2 P=2/5=0.4 PI5=0.4
(6) Social space (Café) Seti a tr opi ka There is
Standard: 1- Café :for sidewalk 3.6m or narrower should have no café in street PI6=0
min 1.2 m of clear sidewalk for sidewalk wider than 3.6 m
Canber r a r oad
C=1 N=1

(7) Landscape and tree Seti a tr opi ka C= ((1

790×2)-1 79) ×9/15=2040.6
(shade, reduce run off and attractive place) N=(1790×2)-179)=3401
Standard: 1- Tree branches should have a vertical clearance P1=2040.6/3401=0.60
of at least 2.4m F=2040.6-0=2040.6
2-At least 7.6m far from intersection P2=2040.6/3401=0.60 NI=2
3-Trees no more than 9 m apart to provide for continuous
tree canopy P3=2/3=0.67
4-Trees should be planted on both sides of streets PI7=(0.60+0.60+0.67)/3=0.62
If distance between trees > 9
Q: ((Length of street with tree-total length of intersections and
Canber r a r oad
C1 = 559.6×9/20=251 .82
N1=559.6 C2=765.6
F1 =251.82-0=251.82
I=3 P3=0/3=0
PI7=(0.77+0.71 +0)/3=0.49

If distance between trees is various
P2=F1+F2.../ N1+N2+...
NI=number of intersections with second condition
I= number of total intersections
P3=NI/I PI7=(P1 +P2+P3)/3
(8) Facility (Fire hydrant) Set i a t r opi ka
There is no need for fire hydrant
Standard: 1-Located in furnishing zone when zone is 1.2m or PI8=0
behind the sidewalk with a min of 1.8 m clear for pedestrian Canberra road C=9
P I8=0.64

(9) Furniture (Trash receptacle) Set i a t r opi ka

C= 500+400=900m
Standard : 1-Centered in furnishing zone when zone is 0.9m N= 1790×2=3580m
wide or greater PI9=900/3580=0.25
2- 9m far from intersection
3-Min 0.6 m from the face of curb Canberra road C=1448
N= 724×2=1448m

(10) Footpath Pavement Setia tropika C1=490×2×1.18=1

156.4m 2
Standard: 1-Should be stable ,anti skid smooth and C2=(1062+170)×2×1.5=3696m 2
continuous W1=1.18 W2= 1.5
2-Surface discontinuities may not exceed 1.25 cm and N1=490×2×1.80=1764
vertical discontinuities between 0.6 cm and 1.25 cm N2=1232×2×1.80=4435.2 PC1=1
Q: Area of standard pavement? 156.4/1764=0.65
A: C PC2=3696/4435.2=0.83 PI1
W=wide of footpath 0=(0.65×980)+(0.83×2464)/3444
If W< 1.80 m =0.78
N= (length of street (both sides) - length of Canber r a r oad
C=(1 367× 1 .5)-(1 2×1 .5×2) =2014.5m 2

W=1.5 N=1367×1.80=2460.6

(11) Marking (crosswalk) Setia tropika C=0

Standard: Two types of cross walk a and b: P=0/36=0
a:Ladder or longitudinal : PI11=0
1- Min cross walk width3m but desirable width is 5 m
2-Space between strips 0.3-1.5 m
Canberra road C=25

P=C/N PI11=1 P>=1 PI11=P P<1
(12) Pedestrian Refuge and Median Set i a t r opi ka
C=0 There is no crossing section
Standard:1- N=20 P=0/20=0
A:Refuge: min 1.8 m width and 6 m length PI12=0
B:Median: min width for median 1.8 m but 2.7 m preferred
and vegetated medians shall be a min 3 m wide ,max width Canberra road C=1
N=4 P=1/4=0.25

(13) Corner island Set i a t r opi ka

Standard: 1- The pedestrian crossings should be at 90 C=0 N=8
degrees across the turn lane PI13=0/8=0
2-Do not place benches ,...or other features in the center
island which may attract pedestrian
Canber r a r oad
C=12 N=12

(14)Sidewalk on both sides Set i a t r opi ka

l1= 1722
Standard: Two sidewalks N1=1790-68= 1722
Q: length of sidewalk in one side? P1=1722/1722=1 n=1
A:l1 l2=1722 N2=1722 P2=1722/1722=1
N1 =length of street – length of intersections in one side m=1
P1=l1/N1 PI14=(1+1)/2=1
Canber r a r oad
l1= 250+430=680
N1=680 P1=1 n=1
l2=256+431=687 N2=687
P2=1 m=1 PI14=(1+1)/2=1

(15) Advance stop bar Setia tropika C=0

Standard: 1-On streets with at least two travel lanes in each P=0/48=0 PI15=0
direction in advance of crossings
2- Stop and yield lines can be used from 1 to15m in advance
Canber r a r oad
C=26 N=32

(16) Width of footpath Seti a tr opi ka

C1=490×2×1.18=1 156.4m 2
Standard: Footpath zone min 1.8 m but 2.4 m width is C2=(1062+170)×2×1.5 =3696m 2

recommended W1=1.18 W2= 1.5

Frontage zone various but min 0.15 m width N1=490×2×1.80=1764
C= Total footpath area N2=1232×2×1.80=4435.2
W=wide of footpath PC1 =1156.4/1764=0.65
If W< 1.80 m PC2=3696/4435.2=0.83
N= (length of street (both sides) - length of PI16=(0.65×980)+(0.83×2464)/3444
intersections)×1 .80 =0.78
If W>=1.80 m Canber r a r oad
C=(1 367× 1 .5)=2020.5m 2
W=1.5 N=1367×1.80=2460.6

(17) Driveway Set i a t r opi ka

There is no Driveway PI17=1
Standard: 1- Is not wider than garage
2- 9 m far from major street intersection 6m from another Canber r a r oad
There is no Driveway PI17=1

(18) Lighting Setia tropika C=0

N=(1 790-68)×2=3444
Standard: 1-enough light P=0/3444=0
2-Pedestrian scale is important PI18=0
3-Should be centered a min of 0.9 m off the face of curb and
from any accessible structure such as shelter
4-Full cut off fixtures which focus light downwards and Canberra road C=0
N=680+687= 1367

(19) Signing Set i a t r opi ka

Transit ,public facilities , crossing and way finding P1=0 P2=1
Q: Enough transit signs? P3=0 P4=1
A:yes P1=1 A:no P1=0 PI19=(0+1+0+1)/4=0.5
Q: Enough public facilities signs?
A:yes P2=1 A:no P2=0
Q: Enough crossing signs? Canber r a r oad
P1=1 P2=0 P3=1
P4=1 PI19=(1 +0+1

(20) Buffer and barriers (bollards) Setia tropika C=0

N=(36+20)×2=1 12
Standard: 1- Removable and lockable bollards across both P=0/112=0
entrances spaced not more than 1.5 m apart to control PI20=0
vehicle access and also 1.2 m to cross wheelchair user 2-
Height varies but around 0.75-1.2 m
3-Allow 0.45m clear space from the front the Curb

N= (total crosswalk + total median crosswalk sections that P=0/70=0
street needs)*2PI20=0
P=C/N PI20=1 P >=1 PI20=P P <1

Canberra road C=0



Bradshaw, C. (1993). Creating and Using a Rating System for Neighborhood

Walkability: Towards an Agenda for “Local Heroes” . Paper presented at the
14th International Pedestrian Conference

Henson, C. (2000). Levels of service for pedestrians. ITE Journal .

Jaskiewicz, F. (2000). Pedestrian Level of Service Based on Trip Quality . Paper

presented at the Urban Street Symposium.

Khisty, C. J. (1994). Evaluation of pedestrian facilities. Beyond the level-of-service

concept . Washington: National Research Council.

Landis, B., Vattikuti, V., Ottenberg, R., McLeod , D., & Guttenplan , M. (2001).
Modeling the roadside walking environment: A pedestrian level of service .
Washington, D.C: National Research Council.

Leslie, E., Saelens, B., Frank, L., Owen, N., Bauman, A., Coffee, N., et al. (2004).
Residents’ perceptions of walkability attributes in objectively different
neighbourhoods: a pilot study. Health & Place .

McLeod, D. S. (2000). Multimodal Arterial Level of Service . Paper presented at the:

4th International Symposium on Highway Capacity.

Phillips, Rhonda, Karachepone, J., & Landis, B. (2001). Multi-Modal Quality of

Service Project : The Florida Department of Transportation.

Steinman, N., & Hines, K. (2004). A Methodology to Assess Design Features for
Pedestrian and Bicyclist Crossings at Signalized Intersections . Paper
presented at the 83nd Annual Meeting of the Transportation Research

Transportation Research Board. Highway Capacity Manual 2000 . (2000).

Washington, D.C: National Research Council.
Angelalia Roza1, Bayu Martanto Adji2,
Raja Syahira Raja Abdul Aziz3, Mohamed Rehan Karim4

Center for Transportation Research, Civil Engineering, Engineering Faculty,

University of Malaya, Kuala Lumpur, Malaysia

Abstract: Bus is one of transportation modes in intercity traveling. Many reasons are laid behind the
preference of bus an intercity transport, such as income, trip purpose, travel, travel time. This paper
presents the outcome of a study on intercity bus passenger travel characteristics with particular
emphasis on university students. A case study was conducted for the intercity bus services between
Kuala Lumpur and Penang. The stated preference method was employed in this study and a special
questionnaire was developed for the purpose. The survey was held on intercity terminal of Parit
Buntar and Butterworth. Regression analysis was applied in the data analysis to get the logit model
between bus preferences with the variable that influence the transport mode preference. The variable
used in this study are travel time and fare. Sensitivity analysis also applied in this study to get the bus
preference sensitivity toward the attribute value of changes.

Keywords: intercity bus, travel characteristics, stated preference, fare, travel time attribute, logit
model, sensitivity.


Intercity public transport service in Malaysia is served by three types of transport mode
service; intercity bus, intercity train (express train) and intercity flight (domestic flight). In
Malaysia, bus becomes popular due to its flexibility compare to the rail transport, in terms of
time departure schedules, frequencies and the bus routes coverage. Schedule of departure for
bus is more flexible and the network is wider than the train. Another reason reported is
because of its cheap fare compare to air transport, which makes bus more affordable to those
with low income.

However, bus transportation service today starts to face high competition against the rail
transport. The main reason is both of air and rail transport offer many interesting factors to
attract the intercity travelers. For example, Keretapi Tanah Melayu Berhad (KTMB)
improved their level of service by shorten its travel time using Electric Train Services
between Ipoh-KL Sentral-Seremban. Besides, air transportation service like AirAsia tries to
eliminate people’s perception of high cost air transportation by selling cheaper ticket at the
off peak times. V. Correnti at al. (2007) argued that in the main complexity in dealing with
fare, especially by air, lies in handling the large number of ticket categories that is part of the
revenue management strategy adopted by airlines Moreover, lack of safety and comfort level
will gradually make the number of users choosing intercity bus reduced time by time. To
face the competition, Malaysia government supports the intercity bus transport development.
It is proven by the construction of intercity terminal in Bandar Tasik Selatan which is
integrated with the feeder transportation in order to increase the accessibility of that terminal.
Now, Intercity bus transport management in Malaysia is seems to be an interesting issue
since The Land Public Transport Authority (SPAD) just have been implemented in 2011. In
the past year, without any authority control, it is difficult to set the rules and control. Unlike
rail transport which is managed under Keretapi Tanah Melayu Berhad (KTMB), and air
transport which is managed under Malaysia Airports Holdings Berhad (MAHB).

This paper presents the stated preference analysis of the mode choices probability of two
types of intercity land transport in Malaysia; intercity bus and rail. Generally, bus or rail
mode has its different specific characteristics that affect the intercity travelers’ choice of
their transportation mode. Train is preferred by some people because of the accessibility to
the train station is better than bus, and quite comfortable. However, the travel time is longer
than the bus because of its track. Moreover, the track does not cover some cities as an Origin
or Destination in Peninsular Malaysia such as Kuantan, Pahang. On the other hand, bus
routes have access to all cities in Peninsular Malaysia and the fare price is also affordable for
long journeys. It offers more comfort and less travel time than the train. But accessibility to
the terminal is sometimes poor. This is because the bus terminal is usually located around the
border of the city.

By looking at the characteristics of these two modes of land transport, it can be seen that
general factors that can support the preference of a transportation mode for intercity traveling
are: the travel time, costs, accessibility to the terminal/station, safety and comfort. Intercity
travel mode choice models are based on the utility maximization hypothesis which assumes
that an individual’s mode choice is a reflection of underlying preferences for each of the
available alternatives and that the individual selects the alternative with the highest
preference or utility (Chandra R. Bhat, 1995).

Therefore, this research is necessary to study the intercity bus users’ characteristic and to get
the specific understanding on bus preference (probability and sensitivity) against the train as
its competitor in terms of cost and travel time.


The preference of mode and the trend of intercity travel characteristic could be known by
studying the behavior of individual users associated to the modes. A survey was conducted
to study the intercity bus services between Kuala Lumpur to the north, corridor of Malaysia
to derive the passenger intercity travel characteristics in Malaysia. The destination in the
survey is KL-Parit Buntar, Penang corridor. The survey was held on intercity bus terminal in
Parit Buntar and Butterworth, Penang, Malaysia.

A special questionnaire was developed and evaluated for the purpose of collecting the data.
Primary data is used in this research. Random intercity bus passengers were interviewed with
several questions about their socio-economic characteristic and travel characteristic. The
passengers’ travel characteristics analysis includes; gender, marital status, monthly income,
monthly expenditure for transportation, expenditure for intercity transport, intercity travel
frequency, feeder service, feeder transport travel time to terminal, intercity mode service
preference, intercity bus mode choice reason, and intercity travel trip purpose.

Socio-economic data obtained is used to describe the effect of socio- economic characteristic
of intercity bus preference, since the authors argued that there is some influence of user’s
socio economic characteristic in preference of bus. In line with Chen (2002), he discovered
in his research that some information such as passengers’ age, personal income, number of
travel companions, in-vehicle time and cost, reliability of air transport, the comfort and
safety of the railway, and the quality of the intercity coach influenced passengers’
transportation mode-choice behavior when considering a long distance travel.
A research was carried out by Rao et. al, (1998) about the choice of access mode to rail using
the Artificial Neural Network (ANN) model and Multinomial Logit (MNL) model.
Throughout the research, they found that the passengers’ information of gender, age,
household income, household size, travel allowance, waiting time, travel time, and travel
cost significantly influenced commuters’ mode-choice behavior.

Steven et al. (1999) stated that the mode choice usually determines how people travel. In
their analysis six mode choice options were considered including car alone (driving privately
operated vehicles), motorcycle (riding in privately operated vehicles), train (express train),
bus (express Bus), plane, and others.

In this research, authors had conducted an analysis of mode choice between intercity bus and
intercity train with nine scenarios of both travel time and fare. A field survey was conducted
and a stated preference method was employed, emphasizing on the Coridor of Kuala
Lumpur-Penang. Another survey was conducted at the inter-city bus terminal of Parit Buntar
and Butterworth, Penang to identify the probability and sensitivity of the bus preference on
the specific OD. This survey are done in 4 days includes weekdays and weekend.

In selecting suitable model to forecast travel demand, two different criteria are taken into
consideration: (1) the number of individuals represented by the model and (2) the data used.
For the first criteria, two classes of model can be produced by using the number of
individuals represented by the model. The first is called aggregate or first generation models,
which are also known as zonal level models. The first generation models represent the
behavior of more than one individual, or perhaps an ‘average’ individual. The second is
called disaggregate or second generation models, which are sometimes also referred to as the
behavioral approach. The second-generation model had been applied by Ortuzar and
Willumsen (1994) to represent the behavior of each individual.

In a disaggregate logit model, the expected maximum utility (accessibility) is capable of

capturing travelers short term welfare change due to induced travel (rational mode/route
choice and destination choice) based on utility maximization (Ben-Akiva and Lerman,
1985). In this study, the authors used disaggregate model to forecast the travel behavior of
intercity bus user in Malaysia. According the authors, the model is suitable to predict
people’s behavior in preference of bus according their individual characteristic.

Subsequently, a set of questionnaire was given to the respondents to analyze their preference
towards bus in some given scenarios by offering some parameter that can influence in
deciding intercity travel mode. Several evaluated parameters that mostly affect the users’
preference towards transport mode for intercity travel are travel time, fare, terminal
accessibility, safety and comfort level. Froidh, (2008) in Y.-H Cheng (2010) stated that uses
have been proven to be an alternative for price-sensitive passengers. A price war can
probably have a greater impact on bus operators. Authors are keen to know about the
capability of the bus when entering the price competition with the train in the existing travel
time and vice versa (travel time competition with the existing fare). Thus, regression analysis
for transport mode choice was described by using the stated preference or the probability of
bus preference toward train. The scenarios given in the questionnaire were exploited to
estimate the bus sensitivity toward the attribute changes made. The data processing
performed by Binnomial Logit Model, and the attributes involved emphasizes on travel time
and travel costs.


Respondent's Gender
38.89 %
% Male

Female train and bus) service competition in

The current situation of public land transport (intercity
KL- Parit Buntar, Penang corridor services are illustrated by fare and time below. There are
almost 67 bus companies operating and offering intercity and urban passenger services to
almost all destinations throughout Peninsular Malaysia. Based on the field survey conducted,
several bus companies provide services for Penang-KL corridor are Plusliner, Etika Express,
Nice Bus/ Nice+/Super Nice, Bandar Express, Konsortium Melayu Klang, Transnational,
Era Mesra, Eagle, Samisha Express/Samisha Holiday, Transnational. The first/highest
service bus coach is around RM53-65. The second/standard coach fare is around RM3 1-42,
and third/lowest coach service fare is around RM27-30.
Train service from Kuala Lumpur to Parit Buntar,Penang, was served by electrified train and
KTMB intercity train (express langkawi senandung utara/malam, express rakyat, expess
sinaran and special north train). The fare ranges from RM34 (U$D 11.12) for third class/asc,
RM67 (U$D 21.90) for second class/ Afc, and RM1 14 (U$D 37.18) for Premier /first class
with sleeping coach. The train travel time is 6 to 7 hours with the route track around


5.1. Intercity Bus User Characteristic Result and Discussion

The result shows that majority of the respondents aged between 20-25 years old (94.44%)
and the majority of status marital among the passengers is single (83.33%). From Figure 2, it
is found that 61.11% of the intercity travel passengers were employed by female. Figure 3
shows the biggest portion of the bus users (88.89%) was student.

Figure 1 : Age of respondents Figure 2 : Gender of respondents Figure 3: Marital status of

Figure 4 shows that the monthly income level of the respondents (which were student)
is varying from the range less than RM 1000 to RM 5000. The result shows 72.22% of
the respondent has income lower than RM 1000 per month, another 16.67% of them
have monthly income between RM 1.001-3.000 and only 11.11% of them earn between
RM 3001- 5000. This shows that the majority of intercity traveler using buses are
from low income group. Most of the respondents (61.11%) argued that they only
allocate less than RM 100 of their income for transport budget (two way traveling),
while 33.33% of them spent between RM 100 to 300 for their transport budget, and the
rest 5.56% allocate RM 300 to RM500 (Figure 5). From Figure 6, it was found that
most respondents (38.89%) allocate their transport budget for intercity traveling less
than RM 30, while 44.44 % of them allocate RM30-45, whereas 11.11 % of them
allocate RM45-70 and the rest 16.67% of them spend RM85-1 05.
Figure 4: Monthly Income Figure 5: Monthly Transport Budget

Figure 6 : Monthly budget allocation for Figure 7 : Intercity travel cost

Intercity transport

Figure 6 shows that 44.44% of the respondents usually spend less than RM50 for one way
intercity travel, 38.89 % of them spend RM30-45, and 16.67% of them allocate more than
RM100 to 150 for intercity traveling. From Figure 7, it is found that 3 8.89% of them usually
spend less than RM30 for one way traveling, but 44.44% of the respondent usually spends
RM30-45 for one way traveling. The rest of them spend between RM45 up to RM105.

Based on intercity transport expenditure of the respondent in Figure 6, and fare of intercity
bus in Figure 7, an interesting point was found is although mostly intercity travelers have
limit of transport budget; 44.44%% of them are willing to take the bus with the fare around
RM30-45, which was implied that mostly bus user (students) enjoyed the standard coach bus
services (second class service).

Traveling frequency has a strong relationship with traveling purpose. Work travel and non-
work travel are the two major groups of intercity traveler based on objectives demand of
traveling between cities. The behavior of the traveler for each group in their trips is slightly
different. Based on High (2009), non-work travel includes travel for personal and family
business, school activities, religious activities, health care, and social and recreational
activities. Moreover, working purpose travel characteristics usually does not take long
enough time (only 1-2 hours), daily, and traveler is usually the people who live in the rural
areas. Recreational and social purposes trips such as visiting family in the village yard or
celebrating religious events are usually performed as weekly and seasonally events. This
type of trip is typically done during the weekend, school holidays, public holidays and
festive season holiday. In line with V. Correnti et al (2007), people usually do intercity
traveling during weekend or festive season. The demand will be higher during holidays, such
as school holidays and public holidays. Otherwise on weekdays it will go off peak.
Figure 8: Intercity travel frequency Figure 9: Reasons of intercity bus preference
Respondent's reason of Intercity Transport
The respondent Intercity
belongs to Travel Frequency
category of non-work travel, related to the Preference
answer on their
intercity travel frequency. Figure 8 shows that 33.33% of the respondents travels monthly,
27.78% of respondent travels
38.89% 27.78%weekly, while 38.89% of the respondent
11.11%travels twice a year
(or more). Many study had prove that travelers behavior in doing long-distance trip
substantially differ from routine trip patterns. The different appear not only in the set of its
available modes, but also in the travelers characteristics. Long-distance travel accounts a
small number of total trips for a substantialEvery
share of all passenger Km-traveledMoreand
the Comfortable Travel Time F
emissions created. month 55.56%
33.33% twice a year (or more)
Intercity land public transport in Malaysia is dominated by bus and train. Author have
argued, these two types of land transport (bus and train) are constantly compete against each
other in mainly four reason i.e, comfort, travel time, fare, or safety. Figure 9 shows that the
reason why choosing intercity transport mode. The most important reason to consider is
travel time (55.56%). The second important reason to consider is comfort (22.22%). The fare
and safety just being considered after comfort and they were in the same portion of important
(11.11%). The travel time of intercity bus is shorter than train. That is becoming the
considerable reason to make bus likely prefer comparing to the train. In relation with fare, V.
Correnti (2007) noted that price elasticity varies by the purpose of travel, being higher
for leisure travelers.

Respondents’ attitudes and perceptions of transport modes may affect their preference and
choices (Outwater et al. 2003) in Lang Yang et al. (2009). Feeder transport service to the
intercity terminal also affects the intercity traveler interest on choosing their intercity
transport mode. 83.33% of the respondent argued to state that feeder service is important in
deciding what intercity mode they prefer. Crisalli (1999) stated that once they decide to use
public transport, they have to consider the access/egress mode, the access/egress time,
waiting time, the number of transfer, transit fare, and transit pass ownership, and the
schedule flexibility. If public transports were not available for the trip, they should take their
private car. However, for long trip, driving own car for intercity traveling was not so
desirable for certain people, since they should consider fee, toll, transit pass ownership,
schedule flexibility and the number of people who shared the trip with them (Lang Yang et
al, 2009). In fact, waiting time is one of important factors in influencing the passenger mode
choice. However, Crisalli (1999) state in his modeling that waiting time is not a well-defined
attribute for public transportation compared with frequency, and its coefficient is
insignificant. The number of transfers is also insignificant.
Figure 11
(a) Feeder transport access mode (b) Feeder transport access mode (c) Feeder transport average
from intercity bus terminal to Intercity Bus Terminal access/egress time to/from
intercity terminal

In this study case of KL- Parit Buntar, Penang corridor, the feeder transport access mode is
dominated by public transport. Figure 11(a) proved that intercity bus user prefer
commuter bus (22.22%) and motorcycle (22.22%), private car (16.67%) and taxi
(16.67%) as their feeder transport for access to the intercity bus terminal. The rest 22.22%
of respondent used not sure and prefer to answer others. It is also similar with feeder
transport egress mode in Figure 11(b), many respondents choose public transport such as LRT
(50%), commuter bus (33.33%), or taxi (5.56%). Only small amount of them prefer to wait
and be picked up by private car (5.56%) or motorcycle (5.56% ).

Related to the access and egress transport service based on the respondent’s perception,
most of respondents argued that the accessibility of intercity bus is sometime important
(44.44%), 3 8.89% of them think it is important and 11.11 % of them state it is very
important (Figure 12).

Figure 12: Intercity traveler Perception on distance Figure 13: Intercity Bus Respondent’s Preference
effect toward intercity mode choice ( Intercity Bus Competitors)
Figure 13 shows that the train is the competitor of intercity bus in Malaysia. 61.11% of the
respondents prefer train as an alternative transport compare to bus, while 27.78% of
respondents prefer to use the car alone. Terry Dinan and Austin (2004) stated that as the
travel distance increases, private cars lose their competitive advantage. When the traveling
distance is over 300 km, air transportation possesses irreplaceable advantages over the other
modes. Due to lower prices and the services provided on bus (video games, onboard movies,
magazines, and newspapers), intercity buses have about a 30–40% market share in Taiwan,
even with their unreliable traveling time.

In this study case of Kuala Lumpur to Parit Buntar-Penang corridor, the distance is around
326.7 km. In line with Lee and Chang (2006), it was a common belief in transport that rails
are dominant mode in long distance travels whilst the automobile is a significant travel mode
for journeys of short to medium. To deal with this condition, intercity bus and rail service in
Malaysia get the greater market share than the car (Figure 13). 5.56% of the passengers
prefer to use motorcycle or plane as their alternative mode.

5.2. Stated preference, result and discussion

The stated preference method was employed in this case study to see the probability of mode
choice between intercity train and intercity bus. Regression analysis was applied in the data
analysis to get the logit model between bus preference with the variable influenced the
transport mode preference. The variables were fare (X 1) and travel time (X 2).

The existing condition was reviewed. Travel time for intercity bus (executive class) for OD:
Kuala Lumpur to Penang is 5.5 hours; Bus Fare is RM 63 (U$D 20.60). Travel time for
intercity train is 8 hours; Train (i.e., Senandung Malam Train) fare is RM 49 (U$D 16.03).
From the regression analysis, the value for the utility was derived.

Y= 0.351761-0.029664372 X 1 -0.321124793 X 2
X1 = Fare (RM)

X2 = Travel Time (Hour)

Y = Utility

The purpose of sensitivity analysis is to determine how the sensitivity and probability of
intercity bus preference toward fare and travel time changes.

In Figure 14, it can be seen Bus sensitivity toward cost attribute for Kuala Lumpur – Parit
Buntar,Penang corridor. From Figure 14, there were compared some scenarios for certain
bus travel time different without any fare changes. Figure 14 shows the set of travel time
different scenario, vary from (-)4 hour to 4 hour. The ‘sensitivity toward travel time
attribute’ figures below figure out bus probability and the sensitivity response.

Nowadays, KTMB (train) as the bus competitor has been encouraging some efforts to
enhance its level of service i.e., toward the construction of Rawang-Ipoh Electrified Double
Tracking Project and express train service. Abdul M. et al. (2008), KTMB was reported to
continue its effort to turn rail travel as a preferred mode of public transportation. With the
completion of the Rawang-Ipoh Double Tracking Project and ongoing project on: Ipoh -
Padang Besar, Seremban-Gemas, Gemas Johor Baru Double Track, further improvement
would be made on the quality of services in terms of speed and reduced the travel time with
smooth and comfortable journey. Train has a possibility to attract more intercity bus
passengers to move from their current mode. Figure 14 implied when the scenario of travel
time different between bus and train can be set (-)4 hours, the probability of bus preference
would be rising until 0.772232189. But when the scenario of travel time different between
bus and train can be set 4 hours, the probability of bus preference would be decreasing until
0.20619 1023.
Fare Travel
N Different Time Probability
o. (RM) Different (bus)

1 14 -4 1.220958531 0.772232189

2 14 -3 0.899833738 0.71091 5334

3 14 -2 0.578708944 0.64077028

4 14 -1 0.2575841 51 0.56404233

5 14 0 -0.063540642 0.484120182

6 14 1 -0.384665436 0.405002145

7 14 2 -0.705790229 0.330529712

8 14 3 -1.026915023 0.263682628

9 14 4 -1.348039816 0.206191023

Figure 14: Sensitivity toward Travel Time Attribute

In Figure 15, it can be seen the Bus sensitivity toward travel time attribute for Kuala Lumpur
– Parit Buntar, Penang corridor. From Figure 15, there were compared some scenarios for
certain bus fare different without any travel time changes. The ‘sensitivity toward cost
attribute’ figures below figure out bus probability and the sensitivity response.

Fare Travel
N Probability
Different Time Utility
o. (bus)
(RM) Different
1 -25 -2.5 1.896181839 0.869458774

2 -20 -2.5 1.74785998 0.85168268

3 -15 -2.5 1.599538121 0.831953821

4 -10 -2.5 1.451216263 0.810185547

5 0 -2.5 1.154572545 0.76034512

6 10 -2.5 0.857928828 0.702227746

7 15 -2.5 0.709606969 0.670314309

8 20 -2.5 0.561285111 0.636749838

9 25 -2.5 0.412963252 0.601798198

10 40 -2.5 -0.032002324 0.492000102

Figure 15: Sensitivity toward Cost Attribute

Figure 15 implied when the scenario of fare different between bus and train can be set (- )
RM25, the probability of bus preference would be rising until 0.869458774 . But when the
scenario of fare different between bus and train can be set (-)RM25 ,the probability of bus
preference would be decreasing until 0.492000102.

Y.-H Cheng (2010) stated that a price war can probably have a greater impact on bus
operator. In this study, based on the two sensitivity graph; toward travel time and fare, it was
found that in this case of study, the the travel time seems to be more sensitive than fare. It is
relevant with the respondent’s statement in Figure 9 that travel time reason (55.56%) is more
important than fare (11.11%).

Thus, conducting further experiment is also a crucial issue for any intercity passenger-
transportation company. As in many other fields, there is a need for ways of experimenting
with alternative policies or operating strategies without being forced to test them in the 'real
world' (Lardinois, C., 2010). Moreover, K. Kottenhoff and C. Lindh (1995) had informed
that the improvement of on a railway line (a new train concept by improved time table and
lower fare) in the south of Sweden in 1992 , has replaced the combination of old rail cars and
standard coach bus. But after 18 month when a new high comfort bus serviced were
introduced can be a good complement to the train time table to increase the number of
departure. Learning from the experience of train and bus experience in Sweden would be
also profitable to encourage Malaysian better in future intercity transport strategy.


It was found that; first, the mainly trip purpose of intercity bus user in this study is social
visit purpose, such as visiting family in the village yard or celebrating religious events
travel. It usually has characteristic as weekly and seasonally trip. It means this journey
usually slightly increases during weekend, school holidays, public holidays and festive
season; otherwise, during weekday, it will go off peak. The respondents are non-working
travel group. The dependence of public transport mode for egress and access transport to
terminal is higher than private transport in this survey , KL- Parit Buntar, Penang corridor.
In users’ perception, terminal accessibility is sometime important in influencing their
intercity mode choice. In this study, it found that the main reason of choosing intercity bus
is travel time, followed by comfort and fare. it was found that in this case of study, the the
travel time seems to be more sensitive than fare. The stated preference result is relevant
with the respondent’s statement that travel time reason is more important than fare.


Abdul, M., Ibrahim, Y., & Hun, G. O. H. M. (2008) Case 5, Keretapi Tanah melayu
Bhd, Group, 1(2), 163 - 182

Bricka, Stacey., 1999, Variations in Long-Distance Travel. TRB Transportation

Research Circular E-C026—Personal Travel: The Long and Short of It,
Conference Proceedings Washington, D.C., ISSN 0097-85 15, pp. 197-206

Ben-Akiva, M. and Lerman, S. R. (1985) Discrete Choice Analysis: Theory and

Application to Travel Demand, MIT Press Series in Transportation Studies,
The MIT Press, Cambridge, MA

Cheng, Y.-hsiang (2010) High-speed rail in Taiwan: New experience and issues for
future development. Transport Policy, 17(2), 5 1-63

Crisalli, U. (1999) User's behaviour simulation of intercity rail service choices.

Simulation Practice and Theory 7, pp. 23 3-249

Cullinane, S., Kevin Cullinane, K. (2003) Car dependence in a public transport

dominated city: evidence from Hong Kong, Transportation Research Part D 8,
Clever, R and Mark M. Hansen (2008) The Interaction of Air and Rail in Japan,
Transportation Research Record, 2043, 1-12

Clever, R. (2006), Airport and Station Accessibility as A Determinant of Mode

Choice. Civil Engineering

Chandra R. Bhat (1995), A heteroscedatic extreme value model of intercity travel

mode choice, Transportation Research Part B Vol. 29, no.6, pp. 471-483

Hurtubia, R., Polytechnique, E., Atasoy, B., Lausanne, D., Glerum, A., Lausanne, D.,
et al. (2010), Considering latent attitudes in mode choice: The case of
Switzerland, (n.d.). 12th WCTR, July 11-15, 2010-Lisbon, Portugal, 1-17

H. -L. Chang, C. -C. Yeh (2005), Factors affecting the safety performance of bus
companies- the experience of Taiwan bus deregulation, Safety Science 43, pp.

Justin S. Chang, (2010), Estimation of Option and non-use values for intercity
passenger rail Service, Seoul national university, Journal of Transport
Geography 18, pp 259-265

K. Kottenhoff and C. Lindh (1995), The value and effects of introducing high
standard train and bus concept in Blekinge, Sweden, transport Policy. Vol.2,
No. 4, pp. 235-241

Lang Yang, Charisma F Choudhury & Moshe Ben-Akiva, Massachusetts Institute of

Technology, João Abreu e Silva & Diana Carvalho, Instituto Superior Tecnico
(2009), Stated Preference Survey for New Smart Transport Modes and
Services: Design, Pilot Study and New Revision, Research Domain: Intelligent
Transportation Systems Research Project: Smart Combination of Passenger
Transport Modes and Services in Urban Areas for Maximum System
Sustainability and Efficiency (SCUSSE)

Lardinois, C. (2010), Simulation, Gaming and Training in a Competitive,

Multimodal, Multicompany , Intercity Passenger- transportation Environment,
J. Operational Res. Soc. Vol. 40, No. 10, pp. 849-86 1

Steven E. Polzin, Xuehao Chu, and Joel R. Rey (1999), Mobility and Mode Choice
of People of Color for Non-Work Travel, TRB Transportation Research
Circular E-C026—Personal Travel: The Long and Short of It, Conference
Proceedings Washington, D.C., ISSN 0097-85 15, pp. 391-412.

V. Correnti et al (2007), The potential of rotorcraft for intercity passenger transport,

Journal of Air Transport Management 13, pp 53-60

Yao, E., & Morikawa, T. (2005), A study of an integrated intercity travel demand
model, Transportation Research Part A 39 (2005) 367–381
Bayu Martanto Adji, Angelalia Roza , Raja Syahira Binti Raja Abdul
Aziz, Mohamed Rehan Karim

Center for Transportation Research, Faculty of Engineering, University

of Malaya, 50603 Kuala Lumpur, Malaysia, Fax: +603-79552182

Potential of cycling for daily activities in and around campus is very high, but weather, other
traffic flow, lack of cycling facilities, culture and social norm will be the typical constraints that
would discourage cycling in campus. Bicycling would face challenges that may reduce its
attractiveness as compared to motorized. All of these constraints must be considered in
efforts to encourage students in university campus to cycle. On the other hand, a lot of
factors can encourage student to cycling, such as the limit of bus route in campus, travel
time and cost. This paper presents the results and analysis on cycling preference as
transport mode in campus with the University of Malaya in Kuala Lumpur, Malaysia as a
case study.

Key words : bicycle, cycling, transport mode, traffic congestion.


There are 12 residential colleges in campus housing more than 10,000 students out
of a total of 25,000 students for the whole university. Students who live in campus use
motorcycles, cars, public bus (Rapid KL bus) which goes to the city, university bus
which goes around within the campus, with some who walk and a much smaller
number who use bicycles.

The public bus fare is RM 1 (about USD$ 0.30) with average frequency of one in
every 30 minutes. The university bus is free but frequency in very low, thus not very
attractive to students,

1.1 Dependence on private vehicles

In University of Malaya, the both of bus do not cover all roads in the campus
network. This is why students who live in residential colleges prefer to use private
vehicles. High dependence on private vehicles subsequently causes a very bad effect
on traffic and environment in the University of Malaya campus, such as traffic
congestion, air pollution and accidents (Karim, 1992).

Through university pro-active educational milieu, college campuses are privileged

places to communicate sustainability and to help reshape society’s transportation
patterns (Balsas, 2003). Balsas also argue that university campuses can constitute a
laboratory for testing and implementing various alternative transportation strategies,
reducing infrastructure costs and minimizing their impacts on surrounding areas.

The significant benefit of reducing car usage in campus is for better future
environment. For example, reduction in the number of cars used means a decrease
for parking area, so that the remaining area can be planned for other facilities that are
more useful (Shannon et al, 2006).

1.2 Cycling for doing activities in campus

In last decade, one of the most sustainable forms of transport is bicycle. Bicycle has
no fuel consumption and brings good health to their users and others as well since it
does not produce harmful smoke (Gatersleben et al, 2007). Besides, Gatersleben et
al. (2010) also states regular bicycle users also gain benefit of low cost of travel.

Gatersleben et al (2010) argue that there are four types of bicyclists on English roads:
(1) responsible bicyclists who use a bicycle safely and responsibly; (2) lifestyle
bicyclists who are keen bicyclists who spend time and money on bicycling; (3)
commuters consists of professionals who use the bicycle to commute to work
regardless of the weather; (4) hippy-go-lucky bicyclists who use their bicycle for
their everyday life and leisure activities.

In fact, there are very small cyclists exist in University of Malaya. Several
discouraging factors affect students to evade using bicycle in campus. These factors
include aggressive car driver, bicycle availability, land topography and lack of
cycling facilities. When asked about the reason people are unlikely to use a bicycle,
they often refer to traffic safety, heavy traffic, inconsiderate drivers, pollution, bad
weather, distance and travel time, gradient, not being fit enough and social pressure
(Bannister, 1988; Davies, Gray, Gardner, & Harland, 2001; Gatersleben & Appleton,
2007; Kingham, Dickenson, & Copsey, 2001; McClintock & Clearly, 1996; Newby,
1993; Wardman, Hatfield, & Page, 1997).

On the other hand, students have high potential for cycling and a lot of factors can
encourage them to cycle. Students are usually more environmentally conscious and
receptive to new ideas. Those who are physically fit, have restricted budgets, live
close to campus and already own a bicycle can easily be attracted to start cycling in
campus (Balsas, 2003). Certainly, public transportation system, environmental
concerns, limited budget and their dynamic activities among youth can encourage
student for cycling in campus. Shannon et al (2006) argue that it is not hard to
encourage student for cycling as a lot of factor could trigger the process, such as
limited car parking permit for undergraduate students. The key factors that can
motivate them are health, affordability, environmental concerns, time and pleasure
(Bonham et al, 2010). Garrard et al (2006) also state similar statement that
motivators for cycling include physical and mental health, fitness, sustainability and
affordability, in line with Cavil et al (2007).


The most widely used transportation mode from residential college is motorcycle
(38.10%), followed by car (28.99 %) and public transport (10.87 %) as shown in
Figure 1.
Transportation mode from
car, 28.99%
college to classes

Public transport,

Walking, 15.94%

cycling, 1.69%

Figure 1 : transport modes from residential college to classes

Most faculties act as centre of activities in University Malaya, and the mobility is
covered by public bus and University bus. However, the route from residential
colleges as the origin is only covered by University bus which serves long time
headway. This explains the high dependence of motorcycle and car among students
who live in residential colleges compared to using public transport. Students also
prefer walking (15.9 %) than using public transport if the distance to the destinations
is not very far.

Concern then arises as the high dependence of motorcycle and private car (71.5%)
will cause a few bad consequences, such as traffic congestion, air pollution and
parking availability. Other than traffic congestion effects, parking availability
become one of the current main problem in University of Malaya. Shoup (1997) and
Dober (2000) argued that the major problem with automobile is the amount of
parking it requires. In college campuses parking is a common problem with different
slants (Balsas, 2003). Keniry (1995) state a university is a group of administrators,
faculty and students held together by a common grievance over parking.

The current system in University of Malaya is implementing restriction of car usage

among undergraduate students and parking management program which only allow
specific registered vehicles to park. However, there are still many cases of limited
parking availability among students and staffs especially in faculties. Therefore the
authors suggest that the university will develop cycling environment and facilities in
order to cope with these problems. In line with Brown et al (2001) who state that
bicycles increase student’s access to housing and employment, reduces the cost of
attending college, and increases transportation equity.

Table 1: Most frequent transport mode among students in campus

Mode from residential college Transport mode for activities in campus
to classes car motor walking cycling public transport
Motorcycle - 56% 25% - 19%
Car 43% - 57% - -
Public transport - - 25% - 75%
Walking - 29% 71% - -
Cycling - 33% 67% - -
A survey was conducted to study the origin-destination activities in campus. The
destination of the trip is the most frequently visited places among students which are
the Sport Center, the Perdana Siswa, IPS, Student Clinic (Residential College 12),
University of Malaya Central Library, mosque and Chancellery. The origin is taken
from selected faculties in University of Malaya which are Faculty of Engineering,
Medical, Science, Accounts and Islamic Studies. The distance from origin to
destination is calculated based on the path traversed by car or motorcycle. The
nearest distance from origin to destination is 265.2 meters (Table 2), from
Engineering Faculty to Main Library. The farthest distance is 3,187.5 meters, from
Islamic studies to DTC and Chancellery. Based on the distances calculated, all of
Origin-Destination has potential for cycling.

Table 2 : Travel distance matrix (m)

Origin/ Sport Perdana IPS Student’s UM Main Mosque DTC &
destination Center Siswa Clinic Library Chancellery
(College 12)
Engineering 1805.6 501.2 1386.6 1386.6 265.2 1428.6 572.2
Account 2039.9 735.5 1620.9 1620.9 499.5 1662.9 806.5
Science 1356.0 286.0 987.0 987.0 296.0 1213.4 487.0
Islamic 629.7 3116.5 881.6 881.6 2880.5 1979.7 3187.5
Medical 2566.3 1261.9 2147.3 2147.3 1025.9 2189.3 1332.9

Table 3 shows the most visited destination by the student in student activities is
Perdana Siswa Complex, followed by Main Library and Sports Center.

Table 3 : Most frequented destinations in campus

Origin/destination Sport Perdana IPS Student’s UM Mosque DTC &
Center Siswa Clinic (College Main Chancellery
12) Library
Engineering 10.9% 41.2% 7.7% 7.6% 17.9% 6.3% 8.2%
Account 11.6% 41.9% 2.3% 2.3% 32.6% 2.3% 7.0%
Science 35.6% 24.4% 11.1% 2.2% 22.2% 2.2% 2.2%

Obviously, biycle is not a new mode of transportation for students. 84.7 % of student
said that they had experience in cycling (figure 2), although 41.5 % of them stated
that they had used bicycle the last time few months ago, 29.9 % few years ago and
17.7 % long time ago (Figure 3). Many factors are taken into consideration by the
students in making the bicycle as a mode of transportation. Some of them are time,
cost, safety, comfort, and the environment. However, there are those who think that
bicycles are not transporation mode, but are only used for recreation purpose,
workout, or just for kids to play.
Have ever cycled

60% Yes, 84.4% Yes, 84.9% Yes, 84.7%

No, 15.6% N o, 15.1% No, 15.3%
female male total

Figure 2 : Have ever cycled

Figure 3 : The experience in cycling

From Analytical Hierarchy TheProcess, as seenof

last time in cycled
Table 4, the most constraint for
few years ago,
cycling in campus29.9%
is rainy day, higher than the hot weather. The author argues that, if
it rains students will certainly not be using the bicycle, they are not comfort to use
rain coat in cycling and road conditions will also very dangerous. Although
long time ago,
conditions when in rain fall in tropical countries is not too extreme
than when the
falling snows in winter but at least have the same characteristic that will reduce the
number of bicycle users and road conditions are used is also very dangerous for
cycling. yesterday, 3.4%

Hot weather in tropical countries it can also reduce the number of bicycle users, but
in certain condition,
Few the area which is lack of public transport route and full of
months ago,
pedestrian, it 41.5% few weeks
will encourage student more choose ago,because they feel too hot
5.4% few days ago, 2.0%
to walk.

Bicycle availability, the aggressive of car driver and hot weather are in four top
ranking of cycling constraint as seen in table 4.
Table 4 : Constraint cycling in campus
Ranking Constraints
1Rainy days
2 Bicycle availability
3Inattentive or aggressive drivers
4 Hot weather
5Terrain (steep hills)
6 need a car for carrying something
7Lack of bicycle facilities at destinations
8 Concerned about bicycle theft
9Concerned about personal appearance
10 Not comfortable to use bicycle
11Do not enjoy physical activities
12 Do not know how to use bicycle

If we want to increase cycling it is necessary to make infrastructural changes (e.g.,

providing more cycle lanes) as well as societal changes (changing work hours and
attitudes to parenting) and individual changes (e.g., attitudes towards driving and

From the survey results are shown in figure 4, by 83% said they would use the
bicycle if it provided a bicycle facilities on campus. Related to the distance, by 36.2
% of the student would use a bicycle with the distance of 5 – 6 km (figure 5).
Figure 4 : Willingness to cycling related with bicycle facilities availability

Willingness to cycling related with bicycle

Facilities available

No, 17%

Yes, 83%
Willingness to cycling related with distance
6 -7 km, 1.7%
5 - 6 km, 36.2%

1 - 2 km, 19.0%

2 - 3 km, 22.4%

4 -5 km, 6.9%
3 -4 km, 13.8%

Figure 5 : Willingness to cycling related to distance

The most reason of students who will use a bicycle if there are bicycle facilities is
because health (20.7%), followed by the environment by 19.1 %, fun ( 13.6 %) and
faster (12. 2 %). While others, such as too hot to walk, cost, fun and willingness to
place parking are hane smaller percentage than the three reasons above ( figure 6).
Figure 6 : the reason for cycling
The reason
The survey done via the internet showed thatfor cycling
the interest of men to respond to the
Parking area Faster, 12.2%
questionnaire is higher
avaibility and
than women. Figure 7 shows the
Fun, 13.6%
percentage of men and
women asaccessibility,
the respondents and in figure 8 is shown the willingness to cycling by
gender. 13.2% Too hot for
walking, 9.1%

Cycling indeed is a physical activity that is usually preferred by male. In line with
Dickinson et al, 2003, cycling culture is male, there is more attention to attract
Cost, 12.2%
woman to cycling. Substantial gender differences in cycling participation in Australia
and other English speaking countries have led some researchers to suggest that
women are not interested in cycling (Merom et al., 2003). This is not the case in
Health, 20.7%
several western European countries, where utilitarian cycling rates are high, and
women cycle more frequently
19.1%than men (Garrard, 2003).
The Gender of Respondent

Female, 39.1%
80% Female, 55.6%Female, 51.8%
40% Male, 60.9%
Male, 44.4% Male, 48.2%
live out side live inside total

Figure 7 : The Gender

Figure 8 : The Related gender with willing to cycling

Traffic safety concerns have beentoidentified
cycling as arelated with
major constraint on cycling in
countries with low rates of cycling, high rates of car use, and large gender
differences in cycling (Garrard et al., 2006; Goldsmith, 1992). These concerns appear
to have a differential impact on women, perhaps because they are more risk averse
than men (Byrnes et al., 1999).
100% Female,
Female respondents
80%in an on-line Female,
survey of 2403 cyclists
42.90% in Melbourne, Australia, in
2005 were more60% 66.70%about cycling in traffic’
likely than males to report that ‘concerns
and ‘aggression 40%
from motorists'
Male,were constraints on cycling (Garrard et al., 2006). In
a telephone survey
20% Male, 33.30%
of 1880 adult Australians conducted females showed a preference
for using off-road0%paths rather than roads with no bicycle facilities ( Garrard et all,
2006). Willing to use Not willing to use

Dependent of student for activities in campus is still very high. There are a lot of
constraints that are being considered by students to choose to cycle around campus.
However, if proper bicycle infrastructure is designed by the university, students are
willing to make bicycle as their transportation mode inside the campus. The gender
perception in the society also influences the female students to refrain from cycling
even if the facilities are provided.


Balsas, C., 2003. Sustainable transportation planning on college campuses,

Transport Policy 10, 35–49
Bannister, C. (1988). Travel to work patterns in England and Wales for
pedestrian and cyclists-their policy implications. Occasional paper,
University of Manchester, Department of Planning and Landscape.
Bergstrom, A., Magnusson, R., 2003. Potential of transferring car trips to bicycle
during winter, Transportation Research Part A 37, 649–666
Byrnes, J.P., Miller, D.C., Schafer, W.D., 1999. Gender differences in risk
taking: a meta-analysis. Psychol. Bull. 125 (3), 367–3 83.
Cavill, N., Watkins, F., 2007. Cycling and health: an exploratory study of views
about cycling in an area of North Liverpool, UK. Health Education 107,
Cullinane, S., Kevin Cullinane, K., 2003. Car dependence in a public transport
dominated city: evidence from Hong Kong, Transportation Research Part D
8, 129–138
Davies, D., Gray, S., Gardner, G., & Harland, G. (2001). A quantitative study of
attitudes of individuals to cycling. TRL report 481. TRL, Crowthorne. B.
Gatersleben, H. Haddad / Transportation Research Part F 13 (2010) 41–48
Dickinson, J., Kingham, S., Copsey, S., Pearlman , D, 2003. Employer travel
plans, cycling and gender: will travel plan measures improve the outlook for
cycling to work in the UK?, Transportation Research Part D 8, 53–67.
Dober, R., 2000. Campus Landscape, Wiley, New York.
Garrard, J., 2003. Promoting cycling among women. Health Promot. J. Aust. 14
(3), 213–215.
Garrard, J., Crawford, S., Hakman, N., 2006. Revolutions for women: increasing
women's participation in cycling for recreation and transport, summary of
key findings.
Garrard, J.,Rose, G., Kai Lo, S., 2008. Promoting transportation cycling for
women: The role of bicycle infrastructure, Preventive Medicine 46, 55–59
Gatersleben, B., Appleton, K., 2007. Contemplating cycling to work: Attitudes
and perceptions in different stages of change, Transportation Research Part
A 41, 302–312.
Gatersleben, B.,Haddad, H., 2010. Who is the typical bicyclist? . Transportation
Research Part F 13, 41-48.
Goldsmith, S., 1992. The national bicycling and walking study case study no. 1:
reasons why bicycling and walking are not being used more extensively as
58 J. Garrard et al./ Preventive Medicine 46 (2008) 55–59
Hagman, O., 2003. Mobilizing meanings of mobility: car users constructions of
the goods and bads of car use, Transportation Research Part D 8, 1–9.
Karim MR (1992). Traffic accidents in university environment, I.T.E. Journal
(USA), July 1992, pp. 30-34
Keniry, J., 1995. Ecodemia—Campus Environmental Stewardship at the Turn of
the 21st Century, National Wildlife Federation, Washington, DC.
Kingham, S., Dickenson, J., & Copsey, J. (2001). Travelling to work: Will people
move out of their cars. Transport Policy, 8, 15 1–160.
Malmberg, T., 1980. Human Territoriality. Survey of Behavioural Territories in
Man with Preliminary Analysis and Discussion of Meaning. Mouton
Publishers, The Hague.
McClintock, H., & Clearly, J. (1996). Cycle facilities and cyclists’ safety.
Transport Policy, 3, 67–77.
Merom, D., Bauman, A., Vita, P., Close, G., 2003. An environmental
intervention to promote walking and cycling—the impact of a newly
constructed rail trail in Western Sydney. Prev. Med. 36, 235–242.
Nankervis, M., 1999. The effect of weather and climate on bicycle commuting,
Transportation Research Part A 33, 417-43 1.
Newby, L (1993). On the right tracks: cycle planning best practice and its
potential in Leicester. Research report No 3. Best Practice Research Unit.
Leicester: Leicester Environment City Trust.
Nilsson, A., 1995. The potential for replacing cars with bicycles for short
distance travel. Thesis 84, Department of Traffic Planning and Engineering,
Lund Institute of Technology, Lund, Sweden.
Shannon, T., Giles-Corti. B., Pikora, T., Bulsara ,B., Shilton, T., Bull, F., 2006.
A active commuting in a university setting: Assessing commuting habits
and potential for modal change, Transport Policy 13, 240–253
Shoup, D., 1997. The high cost of free parking. Journal of Planning Education
and Research 17 (1), 3–20.
Tengstrom, E., 1992. The use of the automobile. Its implication for man society
and the environment. Report 14:1992, The Swedish Transport Research
Board, Stockholm.
Wardman, M., Hatfield, R., & Page, M. (1997). The UK national cycling
strategy: Can improved facilities meet targets? Transport Policy, 4, 123–
Irin Caisarina 1,2* and Assoc. Prof. Dr. Hassim bin Mat 3*
PostGraduate Student, School of Housing, Building and Planning,

Universiti Sains Malaysia, Pulau Pinang, Malaysia

Engineering Faculty, University of Syiah Kuala
Banda Aceh, Indonesia
School of Housing, Building and Planning, Universiti Sains Malaysia,

Pulau Pinang, Malaysia

*E-mail Address: irene1 ,


Pulau Pinang is the third biggest city in Malaysia. It has enjoyed a relatively rapid
development in various sectors such as infrastructure and economy. In addition to this,
Pulau Pinang is also ranked as the third most escalated car traffic city, especially the
growing number of car ownership and usage. The increasing car ownership and usage in
Pulau Pinang for the mobility of its resident has led to car dependency and it has caused
serious traffic congestion problem.
The launching of public transport system under new entity “Rapid Penang” has given a new
enlightment for public transport system in Pulau Pinang, this is hoped to provide a fresh
outlook on modes of transport choices apart from private vehicle (particularly car). Rapid
Penang, likewise, is also expected to be able to provide services on passenger travel
demand, like intensified frequency, quicker travel time, wider route coverage, punctuality,
and the like However, despite the availability of its recent armada of public transport, the
hope to transfer car passengers to Rapid Penang can not be fully met. This can be seen by
the high rate of car usage as favourite modes of transport for the mobility of Pulau Pinang
This paper is trying to analyze the factors that might cause the Penangites depends very
much on their car, the types on favourite mode of transport choice for their mobility, Rapid
Penang service and how effective it is implemented, as well as also try to find the reason
why the Penangites own a car. Besides, this paper also tries to seek alternatives and
necessary actions that can be considered to reduce traffic congestion and car dependency in
Penang Island.

Keywords : congestion, car dependency, public transport, modes of transport.


The use of private vehicle is normally considered as a personal use and this
is actually based on two concepts, namely mobility and accessibility. As has
been pointed out by “The Independent Commission of Transport” that the true
meaning of mobility or destination of transportation is accessibility (Tolley &
Turton, 1995). This definition, in the context of transportation, might be
described as a transportation system which can connect two locations in such a
quick period of time, with specific distances and with a low cost price (Hassim
bin Mat, 1995).

In industrial countries, the level of mobility and accessibility is closely linked

to public transport and private vehicle (car) ownership, whereas in developing
countries, the quality of infrastructure and individual and family income serve as
the most influential factors for mobility.

There are 77% of commuting urban population are using automobile like
private vehicle, truck or van (O’ Sullivan, 2003). Automobile is one of the most
popular transport mode and it is now being used by almost 90% population in
the US. This symptom creates three problems of transportation: traffic
congestion, air pollution and highway accidents. Problems of transportation in
big cities and traffic congestion has given the rise for seven aspects to give
effect (Thomson, 1977), as illustrated by the following picture

Figure 1 7 Problems of Transportation in Big Cities

Traffic congestion is the most common problem encountered by big cities in the
world and it still can not be Traffic
effectively, though, even the efficient and smart
and traffic movement

transport system has been implemented. This problem is usually driven by the increasing
number of operating vehicle,
car park particularly private vehicles
accident (car). The “European
Conference of Ministers of Transport (ECMT, 2007) ”, agreed to have a definition of
traffic congestion as follows::
“Congestion is the impedance vehicles
Transport impose on each other, due to
Pedestrian Impact on
the speed-flow difficulties
relationship, in condition where the use of a transport

system approaches its capacity ”.

Reducing traffic congestion caused by private vehicle (car) to achieve

Uncrowded public transport outside of peak hours
sustainable transport has become the primary
peak hoursgoal of traffic policy (Cullinane,
transport in

2003). Human dependence on car must be controlled or lessened to avoid traffic

congestion. One main way of achieving sustainability is by reducing the
ownership and usage of private vehicle (car) and in the same time, promoting
and introducing the variety of public transport (Banister, 1999; Commision of the
European Communities, 2001).
Car dependency terminology is first introduced by Newman, P. Dan
Kenworthy, J (1999). This term is used to describe the situation faced by some
world big cities, where car is widely utilized as the main mode of urban transport.
Car has ruled out the possibility to select other alternative modes of transport for
mobility. Also, the culture of using car has gradually formed into a natural habbit.
According to Newman & Kenworthy (1999), “Automobile dependency” defined
“High levels of per capita automobile travel, automobile oriented land
use patterns and reduced transport alternatives”.

In accordance to vehicle data statistic in Malaysia between 2003 to 2009,

Kuala Lumpur Federal Territory was at the top position of the registered private
vehicle in Malaysian Peninsula and it has showed a significant improvement
whilst Pulau Pinang was at the third position (

Car Ownership in Pulau Pinang has raised to the average rate of 9,5 % per
year. The raise of car ownership rate has caused serious traffic congestion in
Pulau Pinang city centre. This traffic congestion has triggered delayed of travel
time, high levels of traffic accidents (injuries and death) as well as to

Transport infrastructure provision and design to fulfill the need of travel in

Pulau Pinang is very much concentrated on private vehicle provision, above all
the car provision. Malaysian Government and Pulau Pinang Local Authority
Policy are focusing on car needs such as providing enough space of highway
roads and spacious parking lot to accomodate cars. On the other hand, public
transport mode has received little attention. This is precisely as mentioned by
Mohammad, J and Kiggundu, A.T (2007) that the icreasing number of vehicle
has generated traffic congestion and has severely affect the need and use of
public transport.

Rodrigue, J.P., from Hofstra University (

eng/ch6en/conc6en/levelautomobiledependency.html) argues that the level of
automobile dependency is connected to the availability of transportation
alternatives. Therefore, it can be said that low dependency occurs when the
availability of transportation alternatives is high (less than 50% trip), medium
alternatives are around 50% to 75%, and it can be categorized as high dependency when automobile

Figure 2. Level of Automobile Dependency

number of private vehicle including its user every year.

ng is to fulfil the needs of commuters with comfortable, reliable and affordable service (Azhar Ahmad, 2007, N

Data of the new registered vehicle (Car and Bus) from 2007 until 2009 can
be seen on the table below (
Table 1. New Registered Vehicle in Pulau Pinang
2007 46,488 231
2008 51,565 87
2009 47,307 240

As The descriptive thesis, its elaborate the result of 263 distributed
questioners from 1030 populations at Pulau Pinang, consist of 3 types of
housing complex in urban and rural area. In urban area there are University
Height Condominium (high cost housing), Taman Pekaka (middle cost housing)
and Hamna Flat Sungai Dua (low cost housing), and the housing complex in
rural area are Bandar Baru Ayer Puteh in Balek Pulau which consist of high cost,
middle cost and low cost housing.

A non probability sampling technique is used in this research by assuming

that the entire respondent is car owner and it is impossible to get detail
information of private vehicle owner. Questions are directly asked to the car
owner to get an accurate answer and to avoid the bias. This questionnaire
analysed by SPSS using descriptive statistic method.

4.1 Statistic of Car Owner Characteristic
Result of research from respondent characteristic data who lives both in housing
complex at urban and rural area is shown in table 1 below.

Tabel 1. Respondent Characteristics

Respondent Items Percentage (%)
Sex Man 56.7
Woman 43.3
Age 18 – 25 years 25.5
25 – 35 years 27.8
36 – 45 years 20.5
46 – 55 years 16.7
56 – 66 years 8.0
> 66 years 1.5
Education SR 10.6
SPM 40.7
STPM 6.8
Diploma/College 14.4
University 22.8
Post- Graduate 4.6
Occupation Private Company 43.0
Government/Public Service 16.7
Entrepreneur 13.7
Non working/retired 26.6
Household Income per < RM. 1,500 46.6
RM. 1,501 – RM. 3,000 34.6
RM. 3,001 – RM. 5,000 13.3
RM. 5,001 – RM. 8,000 3.8
RM. 8,001 – RM. 12,000 0.8
> RM. 12,000 1.1
Period of ownership < 1 years 17.1
1 – 5 years 39.2
6 – 10 years 23.6
> 10 years 20.2

4.2 Why Owning A Car?

Highest result from questions given to the respondent about their reason to
posses a private vehicle is because they enjoy the freedom of mobility. Second
highest answer is the easiness to drop children to school or other activities and
the use of private vehicle to transport goods at the third place. Reasons of why
respondents choose to own private vehicle/s is shown in table 2.

From the question of number of car posses by respondent showed 71.1%

of respondents own 1 car; 18.3% respondents own 2 cars; 7.2% respondents
own 3 cars; 3% respondents own 4 cars while 0.4% respondents own 5 cars.
Cullinane (2003) stated that number of cars owned by a family is related to
number of adult (>18 years old) in that family. However, in this research, we
found that buying power based on household income and the easiness of
owning that supported by government and bank policy is also giving a big

From the income, most of respondent (46.4%) afford less than RM

1500/household/month. Nevertheless, the condition is not decreasing their will to
posses car/s in order to use it for their mobility. As mention by Wootton (1999),
by owning car/s shall increase numbers of journey substantially and changing of
mode of transportation from bus to car. The same argumentation is mentioned
by Goodwin, Begg, Banister and Dargay (Cullinane, 2003) who stated that when
a person in the household own car/s, it is not only luxury but a necessity will take
place and makes the owner to be more on cars.
Tabel 3. Reasons to own car/s
Importance Weighted Rank of
Reason Total
Ranking Score 1 Weighted
1st 2nd 3rd
Dislike public transportation (bus) 22 6 18 46 96 7
Useful to transport goods 43 35 40 118 239 3
Easiness to own car/s by credit 25 19 10 54 123 6
Easiness to drop children to school or
56 28 18 102 242 2
other activities
Unavailable public transport in their
6 13 11 30 55 8
To increase social status 2 4 4 10 18 10
Favour to mobility (freedom of
68 51 40 159 346 1
movement) by using car
Comfortness 19 52 45 116 206 5
Efficiency in time 18 56 69 143 235 4
Disable in family member 5 2 6 13 25 9
Corporate car’s 0 2 0 2 4 11
Calculation based on research at Hong Kong by Cullinane (2003). Where weighted
scored gain by adding: first rank times 3, second rank times 2 and third rank times 1.

4.3 Public Transport in Users Perspectives

In the first 5 months of operation of Rapid Penang, more than six million
passenger has been using its service. To make Rapid Penang become every
commuters choice of public transportation a lot of improvement of service were
made. But instead of that, the user’s opinion on the improvement and service
that they had are also important to know. This is as a comparison of the service
given by the rapid penang and the service in users perspective.

User’s satisfaction of the service shows almost equal percentage, where

47.1% of the commuters are satisfied and 52.9% not satisfied.

The unsatisfaction of the users are because of the waiting time at the bus
stop is unreliable (94.2%), bus fare categorised expensive (74.1%) and 43.2%
mentioned that bus route were not covered their journey destination.

It can be obviously seen here that the recent Rapid Penang service has not
yet performed the reliability and punctuality aspects. They can not fulfill the bus
arrival time, which means there are ongoing delays in departure and arrival time
at the bus stop/destination. This is also related to the problem of bus frequency
as the important factor in bus services. The higher the bus frequency, then the
lesser time would be spend by the passenger for bus waiting, so that they can
reach the destinations right in time. Secondly, route/network coverage does not
operate through passenger destinations which allows dissapointment from the
passengers. This might shift passenger to use another modes of transport,
predominantly private vehicle, to travel and cover their destinations.

Satisfied category is given to comfortability (43%), availability of seats

(39.5%), air conditioning (49.4%) and cleanliness (38.4%).
4.4 Car and Public Transport for Mobility.
This part will describe the percentage of car and bus usage of respondent’s
journey for specific purpose in their daily mobility. The responses to the
questionnaire show that 88.2% respondents always use car to go to office,
college/university and only 7.6% use bus for this purpose. For official affair,
87.1% of respondents choose to use car and 3.0% use bus. Most of respondent
(94.3%) always use car for visiting friends and 95.4% for visiting their relatives
and only 4.6% use bus to visiting friends and 3.4% for visiting relatives. For
shopping purposes, mostly they use cars (95.1%) and the percentage of bus use
for shopping is 3.8%. More than eighty percent (80.2%) of respondents always
use car for their sport activity and 4.6% use bus. For respondent’s journey to
cinema/leisure, 85.9% use car and 4.6% use bus. For journey to food
court/restaurant the data shows that 95.8% of respondent use car and only 2.3%
of them use bus.

Based on the data shown above, it can be seen that the respondents
often use private vehicle than bus for their daily activity. This answer matches
with the first reason of owning a car which is because
they like the freedom of mobility from having a private car. It is also because car
can provide the convenient and wider access as the owner demand, such as
carrying things, dropping and taking children to school or other activities. And if it
is calculated from the percentage of the high rate of travel using cars (more than
75% travels) and if it is linked to the figures in table 2, then it is apparent that car
dependency level in Pulau Pinang is considered as high, in which there are 75%
car user for mobility in compare to bus as the considerable alternative modes of

Hong Kong is also experiencing car dependence situation although car

ownership is very low compare to other cities in the developed world
and Hong Kong public transport is on the good level to serve the
commuters (Cullinane, 2003). The same situation is also faced by the
cities in the Western world. The trend of private cars use is highly dominating
compare to public transport use, while in poor countries (which represent the
majority of human population) most people cannot afford a private car.

In the dense urban area, cost of parking becomes one of the judgement
in decision to own a car, so walking, cycling, motorcycling or public
transport often become other options for their mobility.
A research that has been done in Netherlands by Kitamura (1989) shown
that by owning a car has lead a change towards cars trip, particularly
changing on transit use.

A regular pattern on the use of private vehicle for every penangites journey
purpose has lead to traffic congestion. Some questions were asked to have a
respond from respondent concern and awareness of traffic congestion in
Penang Island. In this survey it is found that 64.6% respondent are unwilling to
leave their cars at home and use bus to journey trip. They keep using car but
78.8% of them change the travel time and 77.2% of respondent choose another
route to avoid traffic congestion. To overcome the urban traffic problem in
Penang Island, 43.0% respondent agree to improve public transport quality to a
better level from what The Rapid Penang has now. Forty eight point three
percent (48.3%) agree that local authority is to practice parking control, 47.9%
respondent agree to provide pedestrian in shopping area and 57.8% respondent
very agree for local authority to provide other mode of transport, such as
monorail or Light Rail Transport (LRT). Small percentage of agree and disagree
shown in idea to restrict the entry of cars into the city center.

From the above explanation, some salient points need to be re-emphasized
here, namely:

1. The main motive of private vehicle (car) ownership is because a car allows a
wider range of accessible areas. This mobility and accessibility, plus shorter
travel time, spaciousness/roomy, from and to distant places for its owner as
compared to public transport have become principal reason to have a car.

2. The survey’s results have shown a condition where car ownership and
its usage is on the high level compare to public transport (bus) use
for every journey purpose. The advantage given by private car as being
discussed before (became the main reason of owning a car) are
providing widest access and covered region for users to use the car
for every journey purpose.

3. It is found that the level of car dependency in Penang Island is high. This is
indicated by types of trip where more than 75% inhabitants use private cars
for mobility, exactly as stated in Victoria Transport Policy Institute (2002) and
strengthened by its relation to public transportation alternatives availability in
this very area.
4. The Rapid Penang as a new hope for Penang’s c omm unit y in
public transport service has tried to give the best service, but some
sectors still need to be improved, such as wider coverage area, bus
frequencies and the need to add more bus services for exceptional
condition in peak hours at certain area (point).In this case, some action
should be taken seriously by the local authority to reduce car ownership &
usage, and to reduce traffic congestion at the same time.

5. A proposal to provide special provision for bus such as bus lane (as
applied in Asian countries like Hong Kong and Jakarta, Indonesia)
could be considered in order to achieve a better service, to save
travel time and to avoid traffic jam (where it became a crucial issue
today). Wider bus linkage and coverage area are also in demand to keep
people away from using car and for commuters to be able to reach their
destination by using bus instead of using a car. A special provision
for buses should be in higher priority. A restriction on car use
entering the city centre only on certain times, limited parking lot
and increasing parking price/rates in the city centre should also be
applied. It is also important to do a solid campaign which lead to
encouraging people to use public transport (such as: car free day, bus
day, bicycle day, etc).

6. In addition to some suggestions mentioned above, it is also suggested

to develop a creative and innovative idea to promote Rapid Penang to
be a sustainable public transport in Penang. This is to support the
government policy that has been applied and as a support to provide an
excellent public transport service.

Banister, D. (1999). Sustainable Urban Development and Transport – a
Eurovision for 2020 . Transport Reviews 20, 113–130.
Commission of the European Communities. (2001). WHITE PAPER European
Transport Policy for 2010: Time to Decide . Brussel

Cullinane, S. (2003). Car Dependence in a Public Transport Dominated City:

Evidence from Hong Kong . Transportation Reasearch Part D, 8: 129–138.

Cullinane, S. (2003). The Relationship between Car Ownership and Public

Transport Provision: A Case Study of Hong Kong . Transport Policy 9, 29 – 39.
Hassim bin Mat. (1995). Model Pengangkutan untuk Mobiliti dan Pembangunan
Kawasan Luar Bandar , Thesis Doktor Falsafah, Pulau Pinang-Malaysia.

Kitamura, R. (1989). A Causal Analysis of Car Ownership and Transit Use .

Transportation 16, 55 – 173.

Mohamad, J and Kiggundu, A.T. (2007). The Rise of The Private Car in Kuala
Lumpur, Malaysia – Accessing the Policy Option . IATSS Research Vol 31, No 1, 69–

Newman, P., Kenworthy, J. 1999. Sustainability and Cities; Overcoming Automobile

Dependency , Island Press, United States of America.

Newman, P., Kenworthy, J. (2000). Sustainable Urban Form: The Big Picture . In
Cullinane, S. (2003). Hong Kong’s Low Car Dependence: Lessons and Prospect,
Journal of Transport Geography 11 (2003) 25-35.

O’Sullivan, A. 2003. Urban Economic , Irwin, United Stated of America.

Tolley, R. S., Turton, B. J. (1995). Transport Systems, Policy and Planning: A

Geographical Approach . Longman Group Limited, England

European Conference of Ministers of Transport. (2007). Managing Urban Traffic


Thomson, J. M. (1977). Great Cities and Their Traffic: Transport Systems, Policy
and Planning: A Geographical Approach , pp 182.

Wootton. (1999). Replacing the Private Car . Transport Review, 19: 157–175. .
New Straits Times, 6th October 2007
098 Travelling Characteristics of University Students' Bicycle
Excursion to Cultural Heritage Sites: A Case of Lijiang, China

Xing Huibin and Azizan Marzuki

School of Housing, Building & Planning
Universiti Sains Malaysia
11800 Pulau Pinang, Malaysia

Email: &

Although China is notoriously known as “the Kingdom of Bicycle”, bicycle tourism is still in initial
stage. With the increasing enhancement of environment awareness and the sharp surge of leisure
fitness in China, bicycle tourism, especially short tour, has a rapid development with a very large
space for future development. Bicycle touring can be divided into four kinds: bicycle contest, bicycle
excursion, bicycle exploration and bicycle package, in which bicycle excursion has the largest number
of participants. As far as it goes, university students, white-collar clerk and retired person constitute
the main body of the Chinese bicycle excursion traveller, in which university students undoubtedly is
the main participants in bicycle excursion. Lijiang is one of the most famous and best tourism
destinations in China, which has been listed as the World Cultural Heritage Site by UNESCO in 1997.
Currently, there are two universities in Lijiang, but the city plan to have at least 6 new universities
before 2020 as a part of the second universities town in Yunnan Province and the numbers of
university students are expected to reach more than 100,000. Based on the prospect of bicycle tourism
development in Lijiang, this article will focus on the travelling characteristics of university students’
bicycle excursion. A network community of bicycle tourism in Tourism and Cultural College of
Yunnan University was selected as a research sample. Questionnaires are sent to the respondents via
email after their online approval to survey their attitudes on bicycle travelling desire, bicycle
excursion experience and bicycle excursion values. All data was analyzed from the whole process of
bicycle excursion: pre excursion, on excursion and after excursion. On this basis, three models have
been constructed to clarify the bicycle characteristics of university students during the three stages of
bicycle excursion especially to cultural heritage sites.

Keywords: Bicycle excursion, travelling characteristics, university student, Lijiang, China.


Bicycle has been beyond 220 years old since it was designed by Comte de Sivrac in 1790.
However, it didn’t officially board the stage of tourism until the 1970s. Since 1980s, people
have been gradually aware of the significances of bicycle touring, such as preventing from
premature senility, improving cardiopulmonary function, losing weight, keeping fit, relaxing
soul and mind, and so on. Since then, bicycle tourism has been springing up like a mushroom
after rain all over the world with the widening calls for reducing energy consumption and the
growing carving for prolonging life. Furthermore, local and state governments have take
many measures to facilitate bicycle tourism in formulating traffic rules, designing traffic
signs, propagandizing traffic regulation and improving traffic infrastructures. Meanwhile, all
kinds of bicycle associations appear like a rising wind and scudding clouds, which is a
tremendous impetus to bicycle tourism. Tour Cyclo-touriste International FFCT organised by
the Federation France of Cyclone Tourism annually since 1927 has attracted more than
14000 bicyclist come to France from all over the world in 2009. Nowadays, bicycle tourism
has commonly accounted for 5%-10% of the whole tourism market in USA and many
countries in Europe resulted from the positive participations and broad support from all
sectors of society.

Although China is notoriously known as “the Kingdom of Bicycle”, bicycle tourism is

still in the initial stage. With the increasing enhancement of environment awareness and the
sharp surge of leisure fitness in China, bicycle tourism, especially short tour, will has a rapid
development with a very large space for development and market potential in the future. As
far as it goes, university students, white-collar clerk and retired person constitute the main
body of the Chinese bicycle travellers. Generally bicycle touring can be divided into four
kinds: bicycle contest, bicycle excursion, bicycle exploration and bicycle package, in which
bicycle excursion has the largest number of participants. Bicycle excursion can be defined as
a special one-day trip in the cities or outskirts at weekends or on holidays, which is generally
no more than 10 kilometre and 24 hours. In China university students undoubtedly is the
most main and potential participants in bicycle excursion originating from it’s unmatched
advantages of less expenditures, fashionable form, healthy quality and enriching experiences.
Actually, in almost all Chinese universities, there are already several student associations in
charge of organizing these short bicycle touring.

Lij iang is one of the most famous and best tourism destinations in China, which has
been inscripted as the World Cultural Heritage Site by UNESCO in 1997. Now there are two
universities in Lij iang, and according to the issued city planning, before 2020, at least 6 new
universities will be built as one of important parts of constructing the second universities
town of Yunnan Province. At that time, the number of university students will reach more
than 100000 in Lij iang. But there are still many intractable problems and constraints which
will hinder the sustainable development of bicycle excursion in Lij iang. The article will put
an attention to the travelling characteristics of university students' bicycle excursion which is
very important for all stakeholders of bicycle tourism to improve the healthy development of
bicycle excursion.


2.1 The Definition of Bicycle Excursion

In a broad sense, bicycle tourism can be classified as a special tourism experience by means
of bicycle as the main transportation or an integral component of the journey (Lumsdon,
1996; Ritchie, 1998; Simonsen & Jorgenson, 1998; Sustrans, 1999). However, Simonsen and
Jorgenson (1998) point that racing cyclists should be excluded from bicycle tourist because
their travelling is more competition, not recreation. Meanwhile, Jabaudon (2003) opines that
cycling excursions such as to the shops, office or school should be excluded from any
definition of bicycle tourism. Simonsen and Jorgenson (1996) classify bicycle tourists by
using a continuum of cycle tourists from cycling enthusiast to occasional cyclist. Ritchie
(1998) divided bicycle tourists into 5 types according to travelling time and illustrates their
different characteristics. From the moment situation, bicycle touring consist of bicycle
contest, bicycle excursion, bicycle exploration and bicycle package.
In a narrow sense, bicycle tourism is just the same with bicycle excursion. From this
perspective, Ritchie (1998: p. 568-569) proposes a universal definition of bicycle tourism
which is used in this research:

“A person who is away from their home town or country for a period not less than 24
hours or one night, for the purpose of a vacation or holiday, and for whom using a bicycle as a
mode of transport during this time away is an integral part of their holiday or vacation. This
vacation may be independently organized or part of a commercial tour and may include the use
of transport support services and any type of formal and/or informal accommodation.”

2.2 The Travelling Desire for Bicycle Tourists

Ritchie (1998: p. 569) states that the academic study related to the demand of bicycle tourists
is still extremely inadequate. Much existing literature in this area is speculative in nature
which is just like that bicycle tourism managers commonly design bicycle route and product
on the base of their own feels, not the demands of bicycle tourists (Downward & Lumsdon,
2001; Ritchie, 1998). Chang et al. (2005) compare the difference in accessibility, activity
purpose and attraction between the two kinds of bicycle tourists who use national scenic
bikeways and use local bike lanes in Taiwan. They draw a conclusion that the main travel
motivations of bicycle tourists using national scenic bikeways are seeking tourist attractions
and sceneries along the roads, while others using local bike lanes want to relax and exercise.
From the perspective of transportation environment, the travelling desire of bicycle tourist
are commonly affected by travel time, bike lane or bike path, level of traffic, pavement or
riding surface quality, and presence of a bicycle facility on a bridge (Stinson & Bhat, 2003;
Chang, et al., 2004); bicycling equipment, basic riding skills, necessary training (Richard,
1994); safety, traffic volume, smooth pavement, facility characteristics, appealing scenery,
slow traffic, road width or sidewalk, few stops, few hills (Antonakos, 1993; Chang et al,
2004; Hopkinson & Wardman, 1996; Hyodo, Suzuki, & Takahashi, 2000; Schuett &
Holmes, 1996); number of establishments within one kilometre of home (Hanson & Huff,
1981); low traffic levels of regional settings, secure parking, climate, terrain (Efrat, 1981;
Chang et al., 2004; Schuett & Holmes, 1996); pavement quality, bicycle facility, traffic,
distance and travel time (Bovy & Bradley, 1985; Ortuzar, Iacobelli & Valeze, 2000;
Hopkinson & Wardman, 1996 ).

Furthermore, the Maine Department of Transportation (2001) also suggested

that the selective priorities of bicycle tourists are affected by cultural attractions,
attractive scenery, bike oriented services and accommodations, bicycle-friendly
roads or shared use paths, and features unique to the area.

2.3 The Travelling Experience of Bicycle Touring

The Maine Department of Transportation (2001) points that the travelling
preferences and tourism experiences of bicycle tourists depends on bicycling skill,
their age, organization form in the report on bicycle tourism in Maine in 2001.
Waerden, et al. (2004) pays attention to the influences of the transportation
characteristics on bicyclist experiences, including pavement, bicycle paths and
lanes, bus lanes, on-street parking facilities, and priority signs at crossings. In this
research, Waerden findings of the pavement of road and bicycle paths and lanes
along roads are emphasized.

Ritchie (1999) proposes that bicycle tourism supplier should offer four bicycle tourism
products to satisfy the demand of bicycle tourists during travelling, namely infrastructure
(such as accommodation), information (such as road networks and alternative routes,
signposting) and transport and bicycle tourism support services.

Gardner (1998) divides cycling life cycle into five stages, namely childhood, the break
from cycling, the return to cycling, returning only to lapse and return again and returning to
cycling, having been influenced by encouraging restarting, which will undoubtedly influence
tourism experience of bicycle tourists on different stages.

2.4 The Tourism Benefits of Bicycle Tourism

Krizek (2004) argued bicycle tourism can bring lots of benefits to tourist destinations
such as social transportation benefits, user transportation benefits, social benefits,
user safety benefits, user health benefits, and agency benefits.

After a series of interviews of over 500 leisure cyclists, non-cyclists and regularly
work cyclists, the Department of the Environment, Transport and the Regions in the
U.K (1998) argues that for most people the decision to use a bicycle purely for
leisure purposes is rational since bicycle tourists can enjoy many benefits including
health, fresh air, and being a social and relaxing pastime from the bicycle trip.
Matthew (2007) argues that bicycle tourism is a kind of tourism which is most
suited to been carried out in regional areas and from it local business sectors can
gain many economic benefits. Ritchie and Hall (1999) suggest that bicycle tourists
may be valuable to regional destinations as they are generally stay for a longer time
than other tourists and subsequently generate a higher total expenditure per trip.
The tourism consumptions of bicyclist are mainly used on accommodation, food and
beverages, groceries, cultural and natural attractions, transport and bicycle-related
products and services (EcoGIS Consultants, 2000).


3.1 Sample Selection

Lijiang city is located at west-north Yunnan, China, which is the joint between the
Tibet Plateau and the Yunnan-Guizhou Plateau. In 1997, the old town of Lijiang has
been inscripted as the World Cultural Heritage site by UNESCO with blending
diversified elements from several cultures over many centuries and retaining plenty
of historic townscapes with high quality and authenticity. Now tourism has become
the most valuable brand of Lijiang with 7.58 million visitors and 8.87 billion incomes
in 2009, contributing to 50% of GNP in Lijiang.

Now there are two universities in Lijiang, Tourism and Cultural College of
Yunnan University and Lijiang Normal Junior College. According to the issued city
planning, at least 6 new universities will be built before 2020 as one of important
parts of constructing the second universities town of Yunnan Province. At that time,
the number of university students will reach more than 100000 in Lijiang, who will
greatly stimulate bicycle excursion of Lijiang. Tourism and Cultural College of
Yunnan University (YUTCC) lies between the old town of Lijiang and Snow
Mountain of jade dragon, no more than 5 kilometre from the old town. YUTCC is a
newly-established independent college for undergraduate students, built in 2002.
Now in YUTCC, there are 7 departments involving tourism, art, economics and
management, language, computer and electronics, literature and communication,
and public and foundation, with more than 11000 students. In YUTCC, bicycle
association and tourism association are the main student organizations in charge of
bicycle tourism now. In this research, a network community opening for all YUTCC
students, “YUTCC Communication Community” (TCCC), has been chosen as a
research sample.
3.2 Sample Size
TCCC has 113 regular members coming from the 7 departments of YUTCC.
Because of all members join the community voluntarily, the overall population obeys
the normal distribution. Since the proportion of the sample size in the overall
population is more than 5%, the sample size can be determined by formula (1)
(McDaniel & Gates, 1998; Chiang, 2003; Guo, 1999):



n =

− 
E r
2 2
 
 −  + −

Where, n is the needed sample size; Z is the coefficient of confidence interval in a

certain confidence level; P is the standard deviation of overall population; E is the allowable
sampling error; r is the respondent rate.

In this research, according to the general statistic rule, a 95% confidence level
is used, namely, Z is chosen 1.96. The whole number of community members (N) is
113. Since there were no related researches before, it is not possible to estimate the
value of P. In order to improve the accuracy, the most pessimistic assumptions of
the overall standard deviation is used, namely P = 0.5, which make it possible that
the determined sample size is the biggest. Because of these students will only
receive the questionnaires after online approvals are gained, the sampling error,
namely E can be allowed for a bigger range. Therefore, this research determines E
at 15%.and r is determined at 90%. Then, the calculation result of the formula (1) is
33.62, so the sample size obtained is 34.

3.3 Research Design

Firstly, factors influencing bicycle touring are searched from the related literatures. Based on
that basis, three models respectively aimed to pre excursion, on excursion and after
excursion are constructed and used as the foundation for questionnaire design. The
questionnaire consists of four parts: basic information, factors evaluation of bicycle desire,
factors identification of effecting bicycle experience and benefits evaluation of bicycle
excursion. Secondly, 34 questionnaires are sent by email to these community members who
agree with online participation in investigating their attitudes on bicycle travelling desire,
bicycle excursion experience and bicycle excursion values. Finally, all data are analyzed to
specifically clarify the characteristics of university students bicycle excursion especially to
cultural heritage sites.
4 Analysis and Findings
Male students make up 62% of respondents and sophomores have the largest number
accounting for 35%, followed by freshman and senior students. As for bicycle excursion,
88% students show a strong enthusiasm. Only 6% of the students participated in bicycle
activities for the first time.

4.1 Pre Bicycling: Travelling Desire of Bicycling Excursion

The related literatures concern the bicycle desire form 3 aspects: destination quality,
transportation condition and bicyclist characteristics. The influencing factors are; scenery

quality ( x1), traffic condition and path quality ( x2), bicycle facilities and services ( x3),
distance and travel time ( x4), bicycle equipment ( x5), climate ( x6), safety ( x7), riding skills
and preliminary training ( x8), and cultural attractions ( x9). Based on that basis, a regression
equation (1) is constructed:

TD a = + 0 a1x + 1 a2x + 2 a3x + 3 a4x + 4 a5x + 5 a6x + 6 a7x + 7 a8x + 8 x9 (1)

Where, TD
is the abbreviation of travelling desire for bicycle excursion;

a0 , a1, a2, a3, a4, a5, a6, a7, a8 are the respective coefficients of these independent variables.

Regression method can be used to calculate the coefficients of these variables because
the Shapiro-Wilk statistic of TD is 0.4 >0.05, which means TD obeys normality distribution.
The model summary has shown two simulating models, namely 1 and 2. Because of the
adjustedR 2of model 2 (0.834) is bigger than the adjusted
R2of model 1 (0.797), model 2 is
chosen for further analysis ( Table 1 ).

Table 1: Model Summary of the equation (1

Model R R Square Adjusted R Square Std. Error of the

1 .896 a .803 .797 3.815
2 .918 b .844 .834 3.452
a.Predictors: (Constant), traffic condition and path quality
b.Predictors: (Constant), traffic condition and path quality, scenery quality

The regression equation has statistical significance as the outcome of ANOVA analysis
shows F in model 2 is 83 .620, P=0.000<0.01 ( Table 2 ).
Table 2: ANOVA a,b
analysis of the equation (1
Model Sum of Squares df Mean Square F Sig.
2 Regression 1993.258 2 996.629 83.620 .000 b
Residual 369.477 31 11.919
Total 2362.735 33
a.Predictors: (Constant), traffic condition and path quality, scenery quality
b.Dependent Variable: traveling desire of bicycle excursion
The outcome of coefficients shows significance of x2 (traffic condition and path quality) and

x1 (scenery quality) respectively at 0.000 and 0.008 ( Table 3 ), which are all less than 0.01,
meaning that these variables should be kept in the equation (1). Therefore, the regression

equation (1) is TD 6.309 2.842x 4.593x 2

= + +

Table 3: Coefficients a in the regression equation (1

Unstandardized Standardized
Model t Sig.
Coefficients Coefficients
B Std. Error Beta
2 (Constant) 28.501 4.518 6.309 .000
traffic condition and
.550 .120 .591 4.593 .000
path quality( x2 )
scenery quality( x1) .292 .103 .366 2.842 .008
a. Dependent Variable: traveling desire of bicycle excursion

Furthermore, the outcome of exclude variables shows, the significances of x3 , x4 , x5 , x6,

x7 , x8 , x9 are all more than 0.05, so it is not appropriate to preserve them in the equation
(Table 4 ).

Table 4: Excluded Variables a,b,c

in the regression equation (1
Model Beta t Sig. Partial Statistics
bicycle facilities and
2 -.051 b -.457 .651 -.083 .416
services( x3 )
distance and travel time( x4
.049 b .443 .661 .081 .424
bicycle equipment( x5) .015 b .153 .880 .028 .557
Climate( x6) -.080 b -.593 .558 -.108 .283
Safety( x7) -.003 b -.024 .981 -.004 .379
riding skills and
.027 b .287 .776 .052 .582
preliminary training( x8 )
cultural attractions( x9) .090 b .948 .351 .170 .564
a. Predictors in the Model: (Constant), traffic condition and path quality
b.Predictors in the Model: (Constant), traffic condition and path quality, scenery quality
c.Dependent Variable: traveling desire of bicycle excursion

4.2 On Bicycling: Tourism Experiences of Bicycling Excursion

The literatures about bicycling experiences specifically involve with 8 areas: bicyclist
characteristics, bicycling organization, bicycling experiences, transportation condition,
bicycling infrastructures, climate and weather condition, information identification of bicycle
touring and tourism services of bicycle touring. From the survey shows only 6% students in
YUTCC have once participated in bicycle activities, though 88% respondents are full of
enthusiasm to bicycle excursion. So in this stage, the questionnaires just investigate the
attitudes and concerns on the process of bicycling ( Table 5 ).

Table 5: The Influencing Factors of Tourism Experiences of Bicycling Excursion

Influencing type Influencing area Influencing factor
Inner influences Bicyclist characteristics Bicycling skill
Travelling mood
The quality of equipment
Bicycling organization Organization form
Professional level of organization
Familiarity with companions
Bicycling experiences Past experience
Bicycling expectation
Outer influences Transportation condition Traffic safety
Scenery along the road
Tree shade
Bicycling infrastructures Parking sites on road
Bicycle rent institutes
Bicycle maintenance
Climate and weather condition Atmospheric temperature
Shining of the sun
Information identification Signposting
Bicycle brochures
Tourism services Accommodations and camping
Drinking and rest sites along road
Timely traffic navigation

4.3 After Bicycling: Tourism Benefits of Bicycling Excursion

In the literatures related to tourism benefits of bicycling tourism, there are two research
angles: bicyclist benefits and supplier benefits (see Figure 1 ). The specific benefits are heath

and relax ( x10), bicycling safety ( x11), social contact ( x12), enriching experience ( x13),

social and cultural benefits ( x14), economic benefits ( x15), traffic benefits (x16)

environment benefits ( x17).

On the basis, equation (2) and equation (3) can be formed:

BB b 0
= +
b 2 x +
b3x + 12 b 4 x 13 (2)

Where, BB is the abbreviation of bicyclist’s benefits in bicycle excursion; b0 is the

constant in equation (2); b1 , b2, b3, b4 are the respective coefficients of these independent

SB c c
= +
0 1 x c2x
+ 15 c3x + 16 c4x 17 (3)

Where, SB is the abbreviation of supplier’s benefit of bicycle excursion; c0 is the

constant in equation (3); c1, c2, c3, c4 are the respective coefficients of these independent

Bicycling Social

Health and relax Enriching

Bicyclist's experience
Society Environme nt
and culture

Economi Traffi
Figure 1: Tourism Benefits of Bicycling Excursion

Similarly to the regression equation (1), Regression method can be used to calculate the

coefficients of x10 x 11 , x 12 , x13 in the equation (2), because the Shapiro-Wilk statistic of
BB is

respectively 0.190 > 0.05, which means BB obeys normality distribution.

From model summary, there are two simulating models for the equation (2), namely 1
and 2. Because the adjusted 2
Rof model 2 is 0.4 19, which is higher than 0.339, that of
model 1, so model 2 is chosen. The outcome of ANOVA in model 2 for equation (2) is
12.876, P=0.000<0.01, so the regression equation (2) has statistical significance. The

outcome of coefficients of regression equation (2) shows the significance of constant, x10 and

x13 are respectively 0.000, 0.000 and 0.027 which are all less than 0.05, while x11 and x12

are excluded because their significances are respective 0.626 and 0.296 > 0.05.
As shown in Table 6 , based on the calculation result, the regression equation (2) is,
BB 14.279 4.354x
= +
10 − 2.317x 13

Table 6: Coefficients a in the regression equation (2

Unstandardized Standardized
Model Coefficients Coefficients t Sig.
B Std. Error Beta
(Constant) 58.060 4.066 14.279 .000
2 heath and relax ( x1 ) 0
.359 .082 1.093 4.354 .000
enriching experience ( x1 -.295 .127 -.581 -2.317 .027
a. Dependent Variable: bicyclist's benefits

As for the equation (3), though the Shapiro-Wilk statistic of SB is 0.566 >0.05 obeying
normality distribution, no variables can enter into this equation in the regression analysis. It
is because students, they don’t care about the supplier’s benefits and to a large degree, they
have no many experiences and knowledge about supplier’s benefits.

The research analyzes the travelling characteristics of university students' bicycle excursion
to cultural heritage sites along the whole process of bicycle touring from preparation to
assessment. On the basis of literature review, the research finds that traffic condition and
path quality, scenery quality mainly affect university students’ travelling desire for bicycling
excursion. On the process of bicycle touring, tourism experiences are influenced by bicyclist
characteristics, bicycling organization, bicycling experiences, transportation condition,
bicycling infrastructures, climate and weather condition, information identification of bicycle
touring and tourism services of bicycle touring. After bicycle excursion, university students
commonly think that heath and relax, followed by enriching experiences, is the most benefits
from the touring.

Now the unparalleled advantages of bicycle tourism for protecting environment,

relaxing soul and body, etc. have make it become one of the fastest development and the
most potential sectors in China tourism even if it is just at an infancy stage. Since Lijiang is
one of the most famous tourism destinations featured with cultural heritages in China, the
bicycle excursion in Lijiang will provide a significance value for heritage tourism
development offer lots of inspirations and guidance for the development of bicycle tourism
in China.

Antonakos, C. L. (1993). Environmental and travel preferences of cyclists. The University of
Michigan, Michigan, USA.
Bovy, P. H. L., & Bradley, M. A. (1985). Route choice analyzed with stated-preference
approaches: Transportation research record.
Chang, H., & Chang, H. (2005). Comparison between the differences of recreational cyclists
in the national scenic bikeway and local bike lane. Journal of the Eastern Asia Society
for Transportation Studies, 6, 2178-2193.
Chang, H. L., & Chang, H. W. (2009). Exploring recreational cyclists' environmental
preferences and satisfaction: experimental study in Hsinchu technopolis. Environment
and Planning B: Planning and Design, 3 6(2), 319-335.
Chang, H. W., & Chang, H. L. (2003). A strategic study of bicycle tourism in Taiwan.
Journal of the Eastern Asia Society for Transportation Studies, 5, 1675-1685.
Chang, H. W., & H.N, H. (2004, September 11-13). Bicycle tourism development- a model
for sustainable development in Taiwan. Paper presented at the International Symposium
on City Planning, Sapporo, Japan.
Chiang, C. L. (2003). Statistical Methods of Analysis. Singapore: World Scientific
Publishing Co. Pte. Ltd.
Downward, P., & Lumsdon, L. (2001). The development of recreational cycle routes: an
evaluation of user needs. Managing Leisure, 6(1), 50-60.
EcoGIS Consultants. (2000). Submission on cycle tourism to the Tourism Strategy Group.
Gardner, G. (1998) Transport implications of leisure cycling. Transport Research
Laboratory, TRL Report 347. UK.
GUO, Z. (1999). Social Statistics Analysis Method. Beijing, China: Renmin University of
China Publishing Company.
Hanson, S., & Huff, J. O. (1982). Assessing day-to-day variability in complex travel
patterns. Collection and Use of Survey Data, 891, 18.
Hopkinson, M. (1996). Evaluating the demand for new cycle facilities. Transport Policy,
3(4), 241-249.
Hyodo, T., Suzuki, N., & Takahashi, K. (2000). Modelling of bicycle route and destination
choice behaviour for bicycle road network plan. Transportation Research Record:
Journal of the Transportation Research Board, 1705, 70-76.
Jabaudon, N. (2003). Cycling holidays in Europe. Travel & Tourism Analyst(6), 1-21.
Krizek, K.J. (2004) Estimating the economic benefits of bicycling and bicycle facilities.
TRB Annual Meeting CD-ROM
Lumsdon, L. (1996). Cycle tourism in Britain. Insights, March, 27-32. Available from the
English Tourist Board.
McDaniel, C. D., & Gates, R. H. (1998). Marketing research essentials. Kansas, USA:
International Thomson Publishing.
MDoT. (2001). Bicycle tourism in Maine economic impacts and marketing
recommendations. .
Ortuzar, J. D., Iacobelli, A., & Valeze, C. (2000). Estimating demand for a cycle-way
network. Transportation Research Part A: Policy and Practice, 34(5), 353-373.
Richard, A. L. (1994). The essential touring cyclist: the complete course for the bicycle
traveller: International Marine/Ragged Mountain Press.
Ritchie, B., & Hall, C. (1999). Bicycle tourism and regional development: a New Zealand
case study. Anatolia, 10(2), 89-112.
Ritchie, B., & Hall, C. (1999). Bicycle tourism and regional development: a New Zealand
case study. Anatolia, 10(2), 89-112.
Ritchie, B. W. (1998). Bicycle tourism in the South Island of New Zealand: planning and
management issues. Tourism management, 19(6), 567-582.
Report commissioned by the Cycling Promotion Fund of Australia and the Bicycle
Federation of Australia. Retrieved 2 December, 2005, from
Ritchie, B. W. (1999). Bicycle tourism in the South Island of New Zealand., University of
Otago, Dunedin, New Zealand.
Schuett, M. A., & Holmes, T. P. (1996). Using a collaborative approach to developing a
regional bicycle tourism plan. Journal of Hospitality & Leisure Marketing, 4(1), 83-95.
Simonsen, P., & Jorgenson, B. (1998). Cycle tourism: An economic and environmental
sustainable form of tourism? Unit of Tourism Research, Research Centre of Bornholm.
Retrieved 3 December, 2006, from
Stinson, M. A., & Bhat, C. R. (2003). Commuter bicyclist route choice: Analysis using a
stated preference survey. In Transportation Research Record: Journal of the
Transportation Research Board, 1828, 107-115.
Sustrans. (1999). Cycle Tourism Information Pack. Retrieved 16 November, 2005, from
van der Waerden, P., Borgers, A., & Timmermans, H. (2004). Cyclists' Perception and
Evaluation of Street Characteristics.


Ms Diana Mohamad 1
and Dr Matthew W. Rofe 2

1. PhD Candidate School of Natural and Built Environment The University of South

2. Senior Lecturer, School of Natural and Built Environments The University of

South Australia

ABSTRACT: This paper examines the potential contribution of telecommuting to alleviating

traffic congestion issues. It contends that telecommuting work arrangements should be of
great interest to transport planners specifically and to other planning bodies generally.
Increasingly, telecommuting is recognised as constituting a key strategy in the development
of a more ‘...effective transportation demand management strategy’ (Mokhtarian et. al,
undated). It is evident that the level of adoption and hence the impact of telecommuting
work practices differ for various employment sectors. Drawing upon a case study of a
Malaysia higher education institution in Kuala Lumpur, this paper analyses the potential of
telecommuting in reducing the need of physical travel by private motor vehicles. Whereas
previous research centred on either full or half-day telecommuting, this paper investigates
telecommuting over longer time-frames in order to accurately understand what motivates
individuals to adopt telecommuting. Combined with an examination on the barriers hindering
telecommuting engagement, this paper presents a balanced critique of telecommuting as a
realistic and achievable work option for some employment sectors.

Keywords : Academics, Telecommuting, Physical travel, Malaysia.


Potential policy measures striving towards a sustainable transportation system

has been persistent since the early 1970’s. To date, the pursuit of sustainable
transportation pattern remains critically important and the innovative solution(s)
is increasingly pressing. This would suggest that while important advances have
been realised, much more needs to be achieved. Pa´ez and Whalen (2010)
envisage sustainable transportation awareness as encouraging non-private
transportation practices and/or promoting active transport modes (walking and
cycling) (Zachary, 2007 pp 6). While these are undoubtedly important policy
directions, this paper proposes that significant attention should be given to the
potential for online working arrangement, hereafter referred to as telecommuting,
to contribute to the realization of more sustainable transportation system.
Telecommuting has gained recognition as a potentially effective transportation
demand management strategy whose implementation can contribute to reduced
and therefore smoother traffic flow. Simply stated, telecommuting contributes to
a reduction of on-road vehicle volume. Although widely discussed, there is no
clear, agreed definition for telecommuting and inappropriate usage in describing
various things has lead to the terminological distortion ( see Korte and Wynne,
1996). Korte (1988) argues there are three key facets to telecommuting: [1]
employee’s and employer’s workplace are geographically independent, [2] fully
or partially usage of information technology and [3] both electronic and
traditional (for example mail) means are used for communication. Consequently,
the core concept of telecommuting arrangement is to more flexible choices of
workplaces, not just restricted to the traditional environment.

To date, travel studies typology regarding telecommuting can suitably be

summarized into a number of categories for example urban forms’
characteristics (Zhang, 2005; Kerr, 2007), policy and planning decisions
(Zakaria, 2003) and analytical methods (McMillan, 2007; Røe, 2000). In addition,
travel patterns and travel behaviour are extensively researched fields of inquiry
(see Morikawa, 2001). A robust literature on telecommuting covering areas of
interest reveals promising findings on the penetration of telecommunications into
daily activity and the potential for substitute for physical travel (see Koenig,
1996; Mokhtarian and Bagley, 2000; Lyons and Kenyon, 2003; Walls and
Safirova, 2004; Morikawa and Dissanayake, 2008). The noteworthy episode is
growing and should be of priority to the transportation community (Pa´ez and
Scott, 2007). The telecommuting literature enjoys a long tradition dating back to
the 1960’s (Nilles, 1975; Nilles , 1976; Foster, 1977). Despite this,
conventional wisdom has proposed that travel is purely derived and it holds the
purpose of getting involved in geographically-separated activities (Mokhtarian
and Ory, 2005). Meanwhile, daily long-distance commute would both stimulates
and dissuades the engagement to telecommute. The latter part is supported by
Helmien and Risimaki (2007, pp 331) where people with long-distance home-
workplace travel prefer to find new accommodation closer to workplace, as a
result, travel is not omitted.

The intricacies inherent in this relationship can be divided into potential benefits
and drawbacks. At the most basic, these issues are express in the form of the
following questions; [1] is the daily commute considered stressful (Lucas and
Heady, 2002) and directly deteriorates travel value or [2] does travel play a
significant role as a ‘transitional/ intermediary’ medium for personal activities
(Mokhtarian and Salomon, 2001)? It is of great importance to critically engage
with the complexities embodies within the abovementioned questions and
changes in circumstances that might take place (Mohamad and Bachok, 2009).
Previous studies have verified how certain parameters (for example spatial and
land use planning, technology, economic forces) significantly instigate travel
(see Hamed and Olawayyah, 1999; Black, 2001; Timmermans, 2003 and Syiftan
et. a! , 2003). These parameters impact upon individuals and indeed different
employment sectors in complex ways. To explore these complex issues and
their implications, this paper reports on a detailed case study exploring the
opportunities for telecommuting arrangements to be employed by University
academics and the potential for this to contribute to a more sustainable
transportation system. In doing so, the findings of this research contribute to the
telecommuting literature in two ways [1] it investigates the impact of information
technology on academics trip generation and explores daily trip patterns and [2]
identifies the differences in perceived telecommuting by gender and occupation.


This study exercises both quantitative and qualitative research techniques. The
authors developed an outline of the research framework where it looks at the
concept(s) of the selected work arrangement either theoretically, empirically or at
times both; allows a better understanding on the degree of substitution.
Academics (N =42, female/male = 50% / 50%) from the International Islamic
University Malaysia were selected as the targeted group. Data was collected over
a four-month timeframe using a two-way approach: traditional travel diary survey
(Walls 2006, cited in Haddad et. a! , 2009; Popuri and Bhat, 2003) and personal
interview. Pilot study on the questionnaire survey came back positive with
average of 15 minutes filling in travel diary on daily basis. Data were analysed
using Correlation (bivariate, partial) and Regression (linear, multinomial logistic)
to find the relationship between variables and the causality of the relationship in
terms of direction and degree of influence, if possible. Below are the five
procedural steps to sample selection:
Convenience Sampling: Judgment Sampling:
To select the institution To choose departments/faculties

Stratified Sampling: Simple Random Sampling:

To split first into telecommuter and To divide the academics into
non-telecommuter groups groups of gender

Proportional Quota Sampling:

To decide the number of respondents from each group to be approached

Chart 1: Flow chart of the process of sampling the respondents


3.1 Definition of telecommuter and hypothesis to be tested

The telecommuting term was first coined by Nilles (1975) and since
then, Korte’s (1988) three facets were applied where necessary to
describe telecommuting which comes in scores of forms and
multi-disciplinary operated. Improvised these, limited characteristics
(appropriate worksite, engaged occasionally, full-day arrangement
and applicable computer usage) are chosen for definition purpose. In
this study, telecommuters are labelled as ‘Academics who work at a
remote location or at home on particular working days who either
using ICT or not using ICT to support their productivity, and are
also associated with a traditional workplace at other times and
especially for holding face to face meetings’.

Study focuses to accomplish a better comprehension on

travel, telecommuting and their connection within a higher
educational context; and to examine the relationship, if any. By
understanding academics travel pattern, pervasiveness of habits
and travel demand; study looks for the degree of transportation
substitution by telecommuting. In this context, a reduction in physical
travel and maybe, a decrease in an institution’s space area. Thus,
the hypothesis to be tested is ‘telecommuting can lead to a
reduction in the vehicle kilometres travelled (VKT) by academic staff
working in selected Malaysia university and can make a
3.2 Results

contribution, therefore, reducing the traffic congestion and an

attempt (see also Shannon et. al , 2006) for smaller numbers of
parking spaces’.
This study focuses on questionnaire relating to the one week travel-diary.
The travel-diary survey was developed based on literature (e.g Reuscher et.
al, 2001; Greaves, 2006; Stopher , undated). Respondents were asked
to provide information on time spent and distance travelled for one-way
home-workplace travel, in addition to record the travels on hourly based with
choices for location, activity and transport mode are provided. ‘Other’ choices
could be specified by respondents; which were then included in the list where
possible. The results are as follows:

a. One way origin-destination journey (trip)

Respondents were asked to fill in personal information on administration

position, one-way home-workplace commute distance (dist), one-way
home-workplace commute time (time), travel cost (cost1), travel time
(time1), usual mode (mode) and vehicle ownership (car,motor ). Study
looks at these predictive variables (excluding the administration position)
to find possible trip relationship(s), if any.

Table 1: Pearson Correlation for One Way Origin-Destination Journey Variables.

time dist mode car motor cost1 time 1
time Pearson 1 .857** .537** .503** -.014 .133 .044
Sig. .000 .000 .001 .932 .401 .781

dist Pearson .857** 1 .340* .320* .071 -.005 -.134

Sig. .000 .028 .039 .656 .976 .397

mode Pearson .537** .340* 1 .966** .050 .177 .200

Sig. .000 .028 .000 .753 .262 .204

car Pearson .503** .320* .966** 1 .074 .127 .136

Sig. .001 .039 .000 .640 .424 .391

motor Pearson -.014 .071 .050 .074 1 -.039 -.055

Sig. .932 .656 .753 .640 .804 .728

cost1 Pearson .133 -.005 .177 .127 -.039 1 .546**

Sig. .401 .976 .262 .424 .804 .000

time1 Pearson .044 -.134 .200 .136 -.055 .546** 1

Sig. .781 .397 .204 .391 .728 .000

Source : Adapted from SPSS Pearson Correlation

Pearson = Pearson Correlation where N = 42
Usual mode = car, public transport and motor
** . Correlation is significant at the .01 level (2-
tailed) *. Correlation is significant at the .05 level (2-
3.2 Results

It is important to emphasize that data on commute time is the average
time spent on travel from home to workplace where the estimation
includes the traffic congestion issue. Table 1 (above) presents the trip
relationship in three ways: [1] there are significant relationships between
time, dist, mode and car. [2] Time 1 and cost1 are exclusively significant.
[3] All variables have a significant relationship except for motor . Table
accepted null hypotheses [time and dist] are statistically significant
however, the causality direction is unknown. Nevertheless, Linear
Regression which provides the predictive power for [time and dist] shows
that time will increase by .950 minutes for every additional kilometre. In
Partial Correlation, time and dist (cost1 *) correlation is slightly higher
when the third variable* is held constant, explaining the suppressor
variable effect. Results for [time and dist] indirectly clarifying the traffic
problems face by the respondents.

Table 2: Multinomial Logistic Regression for Cost1*Time

Cost 1 ª B Std. Wald df Sig.
New Intercept -.388 .820 .224 1 .636
accommodation time -.140 .177 .623 1 .430
Telecommuting Intercept -.569 .826 .474 1 .491
time -.095 .171 .304 1 .581

Source: Adapted from SPSS Multinomial Regression Table

ª. The referent category is: adapt

Table 3: Multinomial Logistic Regression for Cost1*Dist

Cost1ª B Std. Wald df Sig.
New Intercept -.954 .763 1.561 1 .211
accommodation dist -.004 .091 .002 1 .967
Telecommuting Intercept -1.170 .781 2.244 1 .134
dist .025 .89 .081 1 .776
Source: Adapted from SPSS Multinomial Regression Table
ª. The referent category is: adapt

Respondents were provided with four options (moves to a new

accommodation closer to workplace, engage in telecommuting, a change
in career and adapt to the increase in travel time and/or travel cost) as
this paper seeks to study the impacts of 50% increase in travel time and
travel cost on the respondents travel choice. Interestingly, both Table 2
and Table 3 (above) rejected the null hypothesis of ‘50% of increase in
travel time and travel cost would result in preferences in choices
provided’. This translates directly that time and distance is not the factor
variables for telecommuting. Provided that the equations are significant,
tables would present that increases in cost1 do not influence the
respondents to choose new accommodation and telecommuting
arrangement over adapt. This may be caused by third variables for
example income and habit pervasiveness (Mohamad and Amin, 2007).
Table 4 (below) demonstrates an interesting impact of time 1 on mode. A
unit increase in mode gives the same weight of respondents preferring
new accommodation and telecommuting to adapt. In conclusion, mode
acts as the third variable to dist and time in influencing travel and
predicting telecommuting preferences.

Table 4: Multinomial Logistic Regression for Time1*Mode

Time1ª B Std. Wald df Sig.
New Intercept 8.266 .421 386.564 1 .000
accommodation mode -9.151 .000 . 1 .
Telecommuting Intercept 8.772 .360 593.942 1 .000
mode -9.151 .000 . 1 .
Source: Adapted from SPSS Multinomial Regression Table
ª. The referent category is: adapt

b. Travel-diary with specific reference to Friday: 7am, 12pm, 1pm and


Study aims to establish two relationships concerning travel and activities

on Friday: [1] relationship between particular time in a day with travel
and [2] relationship between gender and travel. In this direction,
variables studied include internet usage (purpose ), gender , location
(loc), activity (act), travel mode (mode), telecommuter (tc) telecommuting
days (days) and telecommuting pattern (full, half ). Gender shows no
relationship with tc and purpose, respectively. This means that gender
(Mokhtarian et. al, 1998 pp 1132) is not a predictive variable in
influencing telecommuting arrangement and type of internet usage. In
addition, telecommuters are found to be occupied with fieldwork-related
activities more than getting engaged in computer-based work-related
activities on telecommuting day.

Table 5 below explains the findings in two ways: [1] a strong relationship
between grouped purposes [purpose 1*2] and [purpose3*4]. [2]
Tendency of continuity of usage may cause a slightly higher correlation
of purpose2*purpose3 than of purpose1*purpose2. The causality of
direction however, is unpredictable. It is of interest to note that gender
and purpose4 has no significant relationship, which brings this paper to
question: What might be the reason behind this when online
entertainment is perceived as an advantage of information technology?
Is it because the respondents have allocated the time spent on computer
at the working place as time solely for work-related purpose activities?
Or maybe because of the restrictions impose by the university on
internet usage influenced the internet usage?

Table 5: Bivariate Correlation for Gender*Purpose

gender Purpose 1 Purpose2 Purpose3 Purpose4
gender Pearson 1 -.174 .202 -.112 .000
Sig. .270 .199 .481 1.000
Purpose1 Pearson -.174 1 .369* .389* .312*
Sig. .270 .016 .011 .044
Purpose2 Pearson .202 .369* 1 .198 .224
Sig. .199 .016 .210 .155
Purpose3 Pearson -.112 .389* .198 1 .760**
Sig. .481 .011 .210 .000
Purpose4 Pearson .000 .312* .224 .760** 1
Sig. 1.000 .044 .155 .000
Source: Adapted from SPSS Bivariate Correlation Table
N = 42
Purpose1 = class-related work
Purpose2 = non class-related work
Purpose3 = travel and trip planning
Purpose4 = entertainment
**. Correlation is significant at the .01 level (2-tailed)
*. Correlation is significant at the .05 level (2-tailed)

The paper chooses Friday as it consists of all three working patterns

examined (normal, full, half ). Difference in gender’s activity is the
deciding factor as to why these particular hours are selected. Figure 1
(below) shows the Matrix Scatterplot of loc, act and mode of Friday on
the designated hours. It identifies two similarities in the relationship in
terms of characteristic: (1) [mode and loc] and [mode and act] are
inversely related with negative relationship in all scatterplots except for
2pm. (2) loc and act produce both negative (7am) and positive
relationships (12pm and 1pm). Logically, each loc should be pairing up
with one mode since mode is carefully characterized (private vehicle,
public transport, active modes, other). What causes the negative
relationship? One possible explanation: model does not differentiate
between the former, present and future mode used for every activity and
location. Thus, mode is counted as 1 for the same transport mode used
when attending scores of activities at different locations. Meanwhile,
multiple activities done in one place describes the negative relationship
of mode*act and of loc*act at 7am. Correlation results (R2= -.489 for both
bivariate and partial) support the scatterplots findings.

Figure 1: Matrix Scatterplot of Loc, Act and Mode of Friday on Selected Hours

Study found that unless gender is held constant (loc*mode are

statistically significant), a relationship does not exist between gender,
loc, act and mode . All analysis methods employed rejected null
hypothesis ‘gender and act, loc are statistically significant since the male
Muslim lecturers’ act and loc are held constant on Friday between 1 pm
until 3pm as the lecturers are attending a religion event’. Since sample
recorded only one non-Muslim male lecturer, study is positive that
results are influenced by other third variables. Within this paper
knowledge, the following are not the third variables: data entry error,
weather, respondent’s class timetable, Muslim male respondent’s
activity, working hour, holidays and festivals. Question: What might the
third variable(s) be? Table 6 (below) documented the designated hours’
correlation where correlations for 2pm are not statistically significant and
the R2 values are very small when compared to the other three. Is there
any explanation on this, from the pervasiveness of habits’ point of view,
perhaps? What about from the perspective of the controlled variables for
example the choices given to each variable are considered insufficient
and not representative? What if the population and the sample itself
cause the difference?

Table 6: Bivariate Correlation for the Designated Hours

7am 12pm 1pm 2pm
Pearson -.489 -.511 -.750 .005
Sig. .002 .001 .000 .976

Kendall’s -.601 -.504 -.664 .015

Sig. .000 .000 .000 .920

Spearman -.681 -.559 -.752 .017

Sig. .000 .000 .000 .912

Source: Adapted from SPSS Bivariate Correlation Table


This paper has identified several predictive variables namely: gender

(Mokhtarian et. al, undated pp 707), internet usage, commute distance,
commute time, increases in travel time and travel cost are of insignificant
factors in encouraging telecommuting work arrangement, with specific reference
to University academics. It is interesting to note that telecommuter and the
internet usage variables are not statistically significant. Study found that
telecommuters were occupied with fieldwork-related activities on telecommuting
day. This translates into physical travels were made on telecommuting day, thus
defeat the telecommuting purpose of reducing the physical travel need. In
addition, the small R2 value of usual mode (type of transport used for home-
workplace travel) and private vehicle (car) variables expresses that the third
variable(s) is either stimulates vehicle ownership or acts as a suppressor
variable(s) to promoting telecommuting. Moreover, respondents’ perceived and
experienced of potentials and drawbacks of telecommuting seem not having the
propensity impacts on respondents in upgrading their existing telecommuting
plan. Study has confirmed that location and transport modes are statistically
significant only when the third variable gender is held constant.
Set aside the fact that this paper only focuses on particular hours within one
working day. Findings argue that moving to a new accommodation nearer to the
workplace is more likely to be preferred over getting engaged in telecommuting
(see Helmien and Ristimaki, 2007) when travel time and travel cost are to
increase by 50%. According to this, results obtained have failed to test the study
hypothesis that telecommuting can influence travel substitution. On the other
hand, study documented that usual transportation mode acts as a third variable
in stimulating the awareness on telecommuting. This shows that transportation
is a predictive variable in telecommuting research, to a certain extent.
Interestingly, data showed that the only difference between administration
academic and non-administration academic: the first group’s travel-diary starts
earlier than the latter group’s. Travel for these two groups, in general, is of no
different (this refers to the number of travel made). Question remains: [1] what
are the other third variable(s) that control travel characteristic(s) and at the
same time persuade changes in the degree of telecommuting awareness and
[2] provided that other variables are held constant (income, habit
pervasiveness, professional obligation), what is the possibility of telecommuting
implementation in a university setting?


Black, WR (2001). An Unpopular Essay on Transportation. Journal of Transport

Geography 9 (2001) 1-11 . Elsevier Science Ltd.
Dholakia, N; Mundorf, N; Dholakia, R & Xiao, JJ (June 2004). Exploring ways of
Influencing Transport Behaviours by Using Telecommunications Technologies.
University of Rhode Island.
Foster, R (1977). Is Telecommuting Economically Viable? Future
Greaves,S (2006). Simulating Household Travel Survey Data. Institute of Transport
Haddad, H; Lyons, G & Chatterjee, K (2009). An Examination of Determinants
Influencing The Desire For and Frequency of Part-day and Whole-day
Homeworking. Journal of Transport Geography, vol 17, pp 124-133
Hamed, M.M and Olaywah, H (1999). Travel-Related Decisions by Bus, Servis Taxi
And Private Car Commuters In The City Of Amman, Jordan. Jordan University
of Science and Technology.
Helminen, V and Ristimaki, M (2007). Relationships between commuting distance,
frequency and telework in Finland. Journal of Transport Geography 15 : pp. 331-
Kerr, J; Frank, L; Sallis; JF & Chapman (2007). Urban Form Correlates of
Pedestrians Travel in Youth: Differences By Gender, Race-Ethnicity and
Household Attributes. Transportation Research Part D (2007) , pp 177-182
Koenig,B; Henderson,D and Mokhtarian,PL (1996). The Travel and Emissions
Impacts of Telecommuting for the State of Californis Telecommuting Pilot
Project . Transportation Research Part C: Emerging Technologies, vo 4, issue 1,
pp 13-32
Korte, WB and Wynne, R (1996). Telework : Penetration, Potential and Practice in
Europe. IOS Press Ohmsha .
Korte, WB (1988). Telework : Potential, Inception, Operation and Likely Future
Situation. In: Korte et al. (eds.). Telework :present situation and future
development of a new for of work organization. North Holland, Elsevier.
Lucas, J & Heady, R (2002). Flextime Commuter and Their Driver Stress, Feelings
of Time Urgency and Commute Dissatisfaction. Journal Business and
Psychology 16 , pp 565-571.
Lyons G and Kenyon S (2003). Social Participation, Personal Travel and Internet
Use. Proc 10th International Conference on Travel Behaviour Research,
Lucerne, August 2003.
McMillan, TE (2007). The Relative Influence of Urban Form on a Child’s Travel
Mode To School. Transportation Research Part A 41 (2007), pp 69-79
Mohamad, D and Bachok, S (2009). Literature Reviews Comparison Essay On
Travel Focusing On Travel Need, Travel Behaviour And Travel Pattern.
Proceeding of the 13th Conference of the Road Engineering Association of Asia
and Australasia (REAAA): ‘Future Road - Safer, Greener & Smarter' . Incheon,
South Korea.
Mohamad, D and Rofe, M (2009). A Review of Research on Telecommuting,
Telecommunications and Travel. Proceeding of the 13th Conference of the
Road Engineering Association of Asia and Australasia (REAAA): ‘Future Road -
Safer, Greener & Smarter' . Incheon, South Korea.
Mohamad, J and Amin, T. Kiggundu (2007). The Rise Of The Private Car In Kuala
Lumpur, Malaysia- Assessing the Policy Options. IATSS Research Vol. 31 No 1
Mokhtarian, PL & Bagley, MN (2000). Modelling Employees’ Perceptions and
Proportional Preferences of Work Locations: The Regular Workplace and
Telecommuting Alternatives. Transportation Research Part A 34 (2000) , pp
Mokhtarian, PL; Bagley, MN, Hulse, L & Salomon, I (undated). The Influence of
Gender and Occupation on Individual Perceptions of Telecommuting.
Proceedings from the Second National Conference . Women’s Travel Issues.
Mokhtarian, PL; Bagley, MN & Salomon, I (1998). The Impacts of Gender,
Occupations and Presence of Children on Telecommuting Motivations and
Constraints. Journal of The American Society for Information Science , 49 (12),
pp 1115-1134.
Mokhtarian, PL & Salomon, I (2001). How Derived Is the Demand for Travel? Some
Conceptual and Measurement Considerations. Transportation Research Part
A – Policy and Practice 43 (1) , pp 26-43.
Mokhtarian, PL & Ory, DT (2005). When is Getting There Half the Fun? Modelling
the Liking for Travel. Transportation Research 39A (2-3) , pp 97-1 24. University
of California.
Morikawa, T (2001). Travel Behavior Analysis and its Implication to Urban Transport
Planning for Asian Countries : Case Studies of Bangkok, Kuala Lumpur, Manila
and Nagoya. ICRA Project Report .
Morikawa, T & Dissanayake, D (2008). Impact Assessment of Satellite Centre-
based Telecommuting on Travel and Air Quality in Developing Countries by
Exploring The Link Between Travel Behaviour and Urban Form. Transportation
Research Part A 42 (2008), pp 883-894
Nilles, JM (1975). Telecommunications and Organizational Decentralization. IEEE
Transactions On Communications 23, pp 1142-1147
N I l l e s , J M ; C a r l s o n J r , F R ; G r a y , P & H a n n e m a n , G J ( 1 9 7 6 ) . The
Telecommunications-Transportation Trade-Off: Options for Tomorrow . New
York and Chichester: John Wiley.
Pa´ez, A & Scott, MD (2007). Social Influence on Travel Behaviour: A Simulation
Example of the Decision To Telecommute. Environment and Planning A 2007 ,
pp647-665. McMaster University
Pa´ez, A & Whalen, K (2010). Enjoyment of Commute: A Comparison of Different
Transportation Modes. Transportation Research Part A 44 , pp 537-549.
McMaster University, Canada.
Popuri, YD & Bhat, RB (2003). On Modelling the Choice and Frequency of Home-
based Telecommuting. TRB 2003: For Presentation & Publication . TRB 2003
Annual Meeting CD-ROM
Reuscher,TR; Schmoyer,RL Jr and Hu,PS (2001). Transferability of Nationwide
Personal Transportation Survey Data to Regional and Local Scales. Center for
Transportation Analysis
Røe, PG (2000). Qualitative Research on Intra-Urban Travel: An Alternative
Approach. Journal of Transport Geography 8 (2000), pp 99-106
Shannon, T; Giles-Corti, B; Pikora, T; Bulsara, M; Shilton, T & Bull, F (2006). Active
Commuting in a University Setting: Accessing Commuting Habits and Potential
for Modal Change. Transport Policy 13 (2006), pp 240-253. University of
Western Australia.
Stopher,PR; Greaves,PS and Min Xu (undated). Using Nationwide Household
Travel Survey For Simulating Metropolitan Area Household Travel Data.
Institute of Transport Studies, UNSW, Australia
Syiftan, Y; Kaplan, S & Hakkert, S (2003). Scenario Building As A Tool For Planning
A Sustainable Transportation System. Transportation Research Part D 8
(2003) , pp 323-342. Transportation Research Institute, Israel.
Timmermans et al (2003). Spatial Context and The Complexity of Daily Travel
Patterns: An International Comparison. Journal of Transport Geography 11:
Walls, M and Safirova, E (2004). A review of the Literature on Telecommuting and
Its Implications for Vehicle Travel and Emissions. Resources for the Future.
Discussion paper 04-44.
Zakaria, Z (2003). The Institutional Framework for Urban Transportation and Land
Use Planning and Management in the Globalizing Kuala Lumpur Region. Kuala
Zhang, M (2005). Exploring The Relationship Between Urban Form and Nonwork
Travel Through Time Use Analysis. Landscape and Urban Planning 73 (2005) ,
pp 244-261
B. Yousefi 1, Seyed Mostafa Mirhassani 2 ,M.Soltani 3 ,M.J. Rastergar Fatemi 4,and
H.Hakim 5
1, 2, 5 Shahrood University of Technology, Shahrood, IRAN
Islamic Azad University, Ahvaz Branch, Ahvaz, IRAN
Islamic Azad University,Saveh Branch,Saveh, IRAN

ABSTRACT Recently, intelligent transportation systems have been the most attractive topics
among scientific researches. One of the bases of such systems is generating digital maps,
abundantly utilized in GIS, which is why many scientific decisions need this kind of
information. In this case, digital characteristics of roads will be derived from image such as
coordination of road centerline, the solid objects at the road sides and the bridges. In this
paper our aim is to detect roads for providing digital maps in remote sensing satellite images
from urban areas. For this purpose, firstly, based on road appearance, a road hypothesis is
developed as a pre-classification step. Afterwards, based on structural labeling method and
according to some of the image object characteristics, few unknown road parts were
detected and classified to freeway along with one or two lane road. Experiments indicate
promising results which demonstrate efficiency of the proposed road extraction algorithm
from satellite remote sensing images.

Keywords: Road Extraction, GIS, Intelligent Transportation System

I n the recent decade, remote sensing imagery makes the monitoring of the
earth's surface and atmosphere possible in various scales. As the technology of
the imagery sensors improves, the remote sensing images with higher quality
become available. Scientists manage to collect constructive information from the
satellite images. In this way, classification of remote sensing images in urban
area obtains a lot of information which are useful for emergency application,
traffic surveillance, earth survey, map updating and GIS, planning, emergency
response and management, and homeland security applications. Thus,
automated and semi automated methods for the classification of roads,
buildings, and other land cover types in the urban areas attract many research's

One of the most important applications of this technique is map updating. In

particular, in some regions that their map is not prepared, using of such
automatic methods, effectively help civil engineers to make important decisions,
immediately. Providing the digital maps, give opportunity to the engineers to
evaluate several alternatives for new roads, because these kinds of maps can
be imported to navigation instruments such as GPS (Global Positioning System)
for finding actual position on the earth.
Classification of man-made objects is accomplished using pixel-based or object-
based methods. In Pixel-based methods (A. J. Tatem et al., 2001; M. Pesaresi,
1999) n-dimensional vectors from the gray level data of each part of input image
is made afterward, they are compared to a reference vector, trained using a
remote sensing image database. Whilst in the object based approaches, groups
of pixels instead of each pixel are considered to recognize the image objects.
Consequently, neighborhood relationships and shape characteristics are
significant for classification of such images.

By increasing the resolution of images, the accuracy of pixel-based methods for

classification of multi-spectral remote sensing imagery such as minimum
distance from means and maximum likelihood (J. R. Jenson, 1996; C. H. Davis
and X. Wang, 2002) declines. Furthermore, different classes have similar
spectral. As a result, classification of such classes encounters error. Meanwhile,
Fuzzy based methods for classification provides better solution for such
problems, because they attribute fuzzy membership class to pixels. In (Mathieu
Fauvel et al., 2006) A fuzzy based classifier was introduced which has shown its
superiority over simple ANN classifier. Fusions of fuzzy approaches have been
utilized in (Mathieu Fauvel et al., 2005) to improve classification accuracy. In(A.
K. Shackelford et al.,2002), based on spectral similarity of many urban lands
cover types and spatial information such as texture and context an accurate
classification map from input image has been obtained. Then, a fuzzy classifier
has been utilized for the classification of urban area. An object-based algorithm
for the classification of dense urban areas from multi-spectral IKONOS images is
introduced (C. Steger, 1998) in which a cascade combination of a fuzzy pixel-
based classifier and a fuzzy object-based have been exploited. The fuzzy pixel-
based classifier has extracted the spectral content of the scene while the spatial
context information has been analyzed, fuzzy object-based classifier. Employing
support vector machine (SVM) for classification of urban area in satellite images
has been presented in (L. Bruzzone, 2004). Firstly the hierarchical relationships
between the pixels and the adaptive regions to which they are associated at
different levels are considered to make the feature vectors. Then, these feature
vectors have been fed to SVM classifiers.

Segmentation techniques, proposed in (M. Pesaresi et al., 2001; R. P. Kressler

et al., 2002), have been applied to remote sensing imagery for classification. In
(M. Pesaresi et al., 2001), some of morphological operations have been utilized
for segmentation. In (B.Yousefi et al., 2007) an algorithm, which applied
laplacian as feature, has been introduced. First Laplacian of satellite image is
obtained. Afterward, it is fed to a special Bayesian classifier for building
extraction. Then urban areas, roads and streets are extracted using structural
features. In (H.T.shandiz et al., 2008) the unsharp masking technique as
preprocessing step is applied to improve the local image. Then the building
extractor which was mentioned in (B.Yousefi et al., 2007) detected the buildings
with an improvement of accuracy. In the proposed approach Gabor filter bank as
feature is utilized. Before that some of morphological operations (closing and
filling) is applied for the extraction of large buildings .then the buildings are
eliminated from original image. Finally, feature based classifier removes the
open areas. After that, opening with multi-sized and multi direction structure
elements is applied to Gabor filtered image so streets and highways are
extracted in favorable sizes. To obtain road centerlines, thinning operation is
employed. Then, the contour of each road is traced and smoothed using a
Gaussian filter.

The rest of this paper organized as follow. In section 2, a review of Laplacian as

criterion which indicates the variation of image intensity introduced. The
techniques for identifying building, streets, highways, shadows and open areas
are described in section 3. Experimental results are presented in section 4,
followed by conclusion in section 5.

m. Gabor Filter

2-D Gabor filters are effectively utilized in many computer vision applications.
They are one of the most appropriate tools for image analysis approaches such
as texture boundary detection, texture image segmentation/ discrimination and
texture classification/recognition(I. Fogel et al., 1989; A.K. Jain et al.,1991).
Recently for textile flaw detection (A. Bodnarova et al., 1999) 2-D Gabor filters
have been applied as an efficient step for feature extraction.

In varieties of applications such as texture analysis, classification, segmentation,

etc, which distribution of edges gives valuable information, employing edge
detection for better classification is necessary. Gabor filter is used to improve
desirable frequency com ponent s of im age. Three ways hav e been
recommended for designing of 2-D Gabor filters: filter banks, tuned matched
filters and individual filter-design. For covering the frequency plane the filter-
bank approach (A.K. Jain et al., 1991), requires a large set of filters with
predetermined parameters. Although such a large bank of filters may aid
segmentation, it can dramatically affect the quality of recognition during the
classification (I. Couloigner et al., 2000). Also the individual filter-design
approach involves selection of the appropriate filter parameters. The choice of
these parameters is crucial for texture processing tasks. For instance, texture
boundaries can only be detected if the filter parameters are chosen suitably
(T.P. Weldon et al., 1996). In both, tuned matched filters and individual filter-
design methods, “high-dimensionality” is reduced.
Figure 1 depicts spatial conception of Gabor filter applied to the urban area
remote sensing image. 2D Gabor filter which is an example of wavelet filters,
achieve joint localization and resolution in spatial space and frequency domains.
I(x,y) Presents the intensity of input image located in raw and column.
xth yth
Here, the formula of a complex Gabor function in space domain is given:
g(x,y) s(x,y)
= ù
(x,y) (1)
Where is a complex sinusoidal, known as the carrier, and is a 2-D
S(x,y) ù
Gaussian-shaped function, known as envelop. The complex sinusoidal is
defined as follows,
s(x,y) exp( j (2 ( u 0 x v 0 y ) P ))
= π + +

Where and (2)

are the spatial frequency and the phase of sinusoidal,
(u 0,v 0) P
correspondingly. We can imagine this sinusoidal as two separate real functions,
conveniently allocated in real and imaginary part of a function.

Figure 1. The figure shows Gabor filter

The define the spatial frequencies of the sinusoidal in Cartesian
u0 , v 0
coordinates. This spatial frequency can also be expressed in polar coordinates
F 0 ù0
by as magnitude and as direction:
1 0

F u v 2 v  

0 = 2
0 0
+ , ù 0
,  
, (3)
u F cos 0 = 0 ù 0 v 0 = F sin 0 ù 0
u  0 

The representation of complex sinusoidal is:

sxy jFx 0y 0P
( , ) exp( (2 cos
sin (
)) (4) ()
π 0 ù ù

The Gaussian envelop is:

ω r
(,) exp( ( ( ) ( ) ) )

As following in equation we have:

g x y K ( ( a ( x x ) b ( y y ) ) ( j u x v y
( ( ) P ) )

(,) exp exp 2

= − − + − + +
π π
2 2 2 2
0r r0 0 0
K −
( ( a ( x x
− +
) b ( y y

) ) j F x ( ( (
y +
) P )

exp π
exp 2 cos π ω
sin ω

2 2 2 2
0r r 0 0 0 0

Ig (x,y) I(x,y)*g8 (x,y) =

In this paper we have K 1, a,b 15, P 1, x 0 ,y 0 are the half of the height and = = =

width of , in that order. Gabor filter is constructed in 8 directions (from 0 to

360 degree, 45 degree per step). Using 8 direction of Gabor filter and convolving
to the original image, we can present edges in different direction in the urban
image. Several details are indicated in the image results. Urban remote sensing
images, analyzed in this way, are more efficient and more reliable than other
high pass filtered images.

n. Morphological Operation
Using Morphological operations, the objects can be extracted easily and
effortlessly. The precise nature of the expanding or shrinking is determined by a
kernel provided by the operator. Morphological operations have been used for
extraction of image objects and spatial form of image components modification.
We defined an image as an (amplitude) function of two, real (coordinate)
variables or two. An alternative definition of an image can be based
Ig ( x , y )
on the notion that an image consists of a continuous or discrete coordinate set.
In a sense, the set corresponds to the points or pixels that belong to the objects
in the image.
The common effect of dilation is to take each pixel of intensity in source image
and expand it into the shape of the kernel. The contribution of the source pixel to
the kernel-shaped region depends on two parameters: the brightness of the
source pixels (pixels intensity contribute supplementary) and the values of the
kernel pixels (pixels are dark, relative to the centre of the kernel; contribute more
to their locations in the kernel-shaped regions than pixels are intense). But
morphological operations, opening and closing, are not performed directly. Also,
the opening operation can be accomplished by erosion operation followed by
dilation operation. The close operation is a dilation followed by erosion. Opening
and closing have been used because these operations help to illustrate the
structural properties of different objects. By filling operation all of the closed
clusters using mentioned morphological operations are filled. Moreover, some
t echniques have been utilized for elimination of non-r oad object s.

o.Structure Element (SE)

Pattern of the extracted objects was represented in last section; the mentioned
method obtains empirical knowledge from components of urban areas remote
sensing images and understanding about location and characteristics of urban
objects. In this step of our approach, extraction of roads and their size
classification and another parts of urban areas components. Here, first we
extract the roads from the original image. We initialize the rectangular structure
element for searching the roads, and then we slip it along image. Then we resize
and rotate it to find roads. is road's cluster which is collected in this set.
Ig x , y
( )

p.Evaluating The Road Centerlines

Evaluating the road centrelines, referred to as road tracing in literature, is an
important step toward generating a digital map. For this purpose, in this
approach, a thinning operator is applied to the detected road. Afterward, for
declining abrupt changes of the centreline location, its contour is obtained and is
smoothed by applying a Gaussian filter. After obtaining the location of each pixel
in the road centreline, we can precisely generate a map in which by clicking in
each part of the roads, its precise coordination is clear. Consequently, the digital
map can be used for navigation instruments such as GPS.
In this section, we demonstrate the application of the proposed common
methods to improve intelligent transportation systems. The discussed method
applied to very high resolution satellite images from Reykjavik, Iceland. The
most important classifications were considered in each case, namely large
buildings, houses (small buildings), shadows, roads, and open areas. Each
image consists of urban area components including buildings, roads and open
areas but we classified it as roads and non-roads classes because of focusing
on roads as transportation system. The primary experimental result illustrates a
significant improvement in extraction of roads. The SE size in morphological
operation filtering is equal to a 3-by-3 square matrix form and the linear shape
structure element with various directions by 30 pixels length.

a. Pre-classification

The images used to test in the proposed approach have 1024 × 716 pixel. As we
know about the results of applying Gabor filter in the input image and the ability
of the discrimination; some urban area images are used to benchmark the
approach. The methods which haven’t being pre-processed, preprocessor like
Gabor filter part, do not enhance the image contrast. Thus, the discriminator
shows error because some image components have similar level of intensity to
the level of background.

The Gabor filtering gives us an opportunity to focus in the direction of defining

edges. As it’s mentioned earlier, we have a trade between sides and time
consumption in our algorithm. In fact, if we consider lot of sides to apply Gabor
filter, the algorithm will be more time consuming than main and secondary
directions. Appropriate levels of discriminated frequency and amount of edges,
is obtained by using the training map. It was also considered that, the small
buildings have high-frequency components and they have a grained texture on
their edges. While building roofs have smooth texture in satellite images.
Figure 2. The figure presents some results of mentioned approach before relaxation

b. Structural Labeling

Structural labeling method can decrease the rate of false positive results and
other errors and the way of using it make this paper novel and interesting.
Structural operates on objects, roads and highways, with labels attached to

The haziness in detector responses is resolved by using structural labeling

instead of pixels labeling which used in (Steven W. Zucker et al., 1977). The
reference structure obtained from training map by representation as a
mathematical graph. For using structural labeling, the training map has been
made. The training map obtained from our data has the properties of being
quintessence in urban region. This technique classifies labels to objects.

In addition to the foregoing, a neighbourhood relation is specified on the set of

objects. The importance of implementing algorithm was to obtain worthwhile
probabilistic function. We must organize some rules regarding, how the
components of urban area images set to their own places. By applying rules, we
can analyze the entire structure of image. The rules are defined as follow:

1.Beside buildings, there are always open areas and roads.

0.The open areas beside the roads do not interrupt as it has special direction.
2.Open areas have nonspecific and various shapes as compared to the roads.
4. The roads are always oblong in shape and continuous.
These rules are obtained by analyzing the urban area images. On the other
hand, it does not mean that we are going to classify all of the urban image
components. We just check the accuracy of components from roads side view.
Then, we arrange a set of Graph models from structure of urban areas and
component shapes as it is required for implementing the labeling.

The applied structural labeling in pixels mode is understood. In addition to the

system, it will be time consuming if the pixel is utilized for relaxation labeling.
Because urban areas images are very large in size and we should organize
each note of graph to each pixel.

Also the number of notes in graph makes the analysis very lengthy and time
consuming. In fact, we have solved this problem by using structures instead of
pixels as it reduces the time span for algorithm. As considered earlier, the
proposed approach used was based on structural, contextual, and statistical
information of urban images. It means that the purposed approach considered in
aspect of structural by using morphological operation, and using the contextual
information of urban image in a similar manner by using structural labeling.
Figure 3. The figure presents some results of mentioned approach after applying relaxation

Moreover, labeling the structural information of image component helps in

making the training map and structural graph model. The figure 3 roughly shows
the comparison between mentioned technique and previous results.


In this paper a new method for extraction of road centerlines in remote sensing
images to aid intelligent transportation presented. The mentioned system can
generate digital map of roads, and abundantly utilized in GIS. Digital
characteristics of roads will be derived from image such as coordination of road
centerline, the solid objects in the road sides. The opening with multi-sized and
multi direction structure elements is applied to Gabor filtered image. So streets
and highways are extracted in favorable sizes. This part developed as a pre-
classification step. Afterward, structural relaxation labeling method and
according to the some of the image object characteristics, some of the unknown
road parts detected and classified to free-way and one or two lane road. For
evaluating road centerlines morphological operation is applied to the extracted
roads and highways.

Finally for providing smooth centerlines a Gaussian filter was applied to the
contours. As it mentioned before, the novelty of this approach is because of the
way of using structural labeling in structural form. Experiments indicated
promising results in evaluating road centerlines as an effective step toward
generation of digital maps from remote sensed images.


A. J. Tatem, H. G. Lewis, P. M. Atkinson, and M. S. Nixon (2001) .Super-resolution

mapping of urban scenes from IKONOS imagery using a Hopfield neural
network,” in Proc. IGARSS, vol. 7, pp. 3203–3205.

A. K. Shackelford and C. H. Davis (2002) .A fuzzy classification approach for high-

resolution multispectral data over urban areas. IGARSS, vol. 3, Toronto, ON,
Canada, June 24–28, pp. 1621–1623

A. Bodnarova, M. Bennamoun, and S.J. Latham (Sept. 1999).Flaw detection of

jacquard fabrics using Gabor filters. In SPIE Znt. Symp., Boston, MA, USA

A. K. Jain and F. Farrokhnia (1991).Unsupervised texture segmentation using Gabor

filters. Pattern Recognition, 24(12):1 167-1186

B.Yousefi, S.M.Mirhassani, and H. Marvi (2007) .Classification of remote sensing

images from urban areas using Laplacian image and Bayesian theory.
Proceedings of SPIE –Vol 6718 Optomechatronic Computer-Vision Systems II
doi:10.1 117/12.754563, 6718 67180F, pp.1-9.

C. Steger (1998) .An unbiased detector of curvilinear structures. IEEE Trans.

Pattern Analysis. Machine Intelligent. vol. 20, pp. 113–125

C. H. Davis and X. Wang (2002) .Urban land cover classification from high
resolution multi-spectral IKONOS imagery. in Proc. IGARSS, vol. 2, pp. 1204–
1206, Toronto, ON, Canada

Fauvel, Jocelyn Chanussot, and Jón Atli Benediktsson, (2006) "Decision Fusion for
the Classification of Urban Remote Sensing Images" IEEE Transactions On
Geoscience And Remote Sensing, vol. 44, no. 10

F. Dell’Acqua and P. Gamba (Oct. 2001) .Detection of urban structures in SAR

images by robust fuzzy clustering algorithms: The example of street tracking.
IEEE Trans. Geosci. Remote Sensing, vol. 39, pp. 2287–2297.

H.T.shandiz, B.Yousefi and S.M.Mirhassani (2008) .Hierarchical Method for Building

Extraction in Urban Area's Images using UnSharp Masking [USM] and Bayesian
classifier, 15th International Conference on Systems, Signals and Image
Processing IWSSIP, Slovakia

I. Fogel and D. Sagi (1989). Gabor filters as texture discriminator. Biological

Cybernetics, 61:103-113

I. Couloigner and T. Ranchin (July 2000) .Mapping of urban areas: A multiresolution

modeling approach for semi-automatic extraction of streets,” Photogramm. Eng.
Remote Sens., vol. 66, no. 7, pp. 867–874.
J. R. Jenson (1996) .Introductory Digital Image Processing: A Remote Sensing
Perspective. 2nd ed. Upper Saddle River, NJ: Prentice-Hall

L. Bruzzone, L. Carlin and F. Melgani .A Multilevel Hierarchical Approach to

Classification of High Spatial Resolution Images with Support Vector Machines.

M. Pesaresi (1999) .Textural classification of very high-resolution satellite imagery:

Empirical estimation of the interaction between window size and detection
accuracy in urban environment. in Proc. ICIP, vol. 1, pp. 114–118.

Mathieu Fauvel, Jocelyn Chanussot and Jon Atli Benediktsson (2006)

.Classification of Remote Sensing Images From Urban Areas Using a Fuzzy
Possibilistic Model. IEEE Geoscience And Remote Sensing Letters, vol. 3, no. 1
Mathieu Fauvel, Jocelyn Chanussot and Jon Atli Benediktsson (2005) Fusion of
Methods for the Classification of Remote Sensing Images from Urban Areas.
IEEE. 0-7803-9050-4/05/.
M. Pesaresi and J. A. Benediktsson (2001) .A new approach for the morphological
segmentation of high-resolution satellite imagery. IEEE Trans. Geosci. Remote
Sensing, vol. 39, pp. 309–320

R. P. Kressler, T. B. Bauer, and K. T. Steinnocher(2002) .Object-oriented perparcel

land use classification of very high resolution images. IEEE/ISPRS Joint
Workshop on Remote Sensing and Data Fusion over Urban Areas, Nov. 2002,
pp. 164–167.

Steven W. Zucker, Robert A. Hummel (1977).Azriel Rosenfeld ().An Application of

Relaxation Labeling to Line and Curve Enhancement. IEEE Transactions on
Computers, vol. c-26, no. 4 p 394-4-3
T.P. Weldon, W.E. Higgins, and D.F. Dunn. Efficient Gabor filter desing for texture
segmenation. Pattern Recognition, 29(1 2):2005-201 5, 1996.

Raziah Noor Bt Razali 1, Mohamad Iskandar bin Illyas Tan 2 and Mohammad
Ishak Desa 3
Halal Informatics Research Lab (HOLLISTICS), Faculty of Computer Science and
Information System, Universiti Teknologi Malaysia, 81310 Skudai, Johor, Malaysia.
Halal Informatics Research Lab (HOLLISTICS), Faculty of Computer Science and
Information System, Universiti Teknologi Malaysia, 81310 Skudai, Johor, Malaysia
Department of Modelling and Industrial Computing , Universiti Teknologi Malaysia,
81310 Skudai, Johor, Malaysia, iskandar@fsksm .utm .my,

ABSTRACT: This paper presents a review on existing information communication

technology (ICT) applied towards transportation operation in logistic and also discusses the
possibility of ICT adoption for transportation in conducting Halal logistic. It discusses the
issues on Halal control in transportation operations in logistic as Muslims consumer requires
the integrity of Halal products from farm to fork . Issues such as sharing container, lack of
visibility across an entire supply chain includes poor containers identification, segregating
allocation space between Halal and non Halal product in same container (for contamination
avoidance) increased the risk toward Halal integrity being compromised. Besides, lack of
study on Halal transportation in logistic has hampered the progress of research in this area.
The objectives of this study are to (1) provide an understanding of Halal control in
transportation system in Malaysia and (2) highlight issues on the Halal transportation
process (3) discuss the possibility of ICT for transportation in conducting Halal supply chain.
Thus this research critically examines how IT can be implemented towards ensuring an
efficient Halal transportation to gain the integrity of Halal goods being delivered while
transforming Halal Industry in Malaysia to be one step advance in today’s digital economy. In
this study, semi-structured interview and document review are conducted. The interview
involved experts in Halal transportation process also Halal certified company in transporting
Halal product. The document such as Halal manual procedures, Malaysian Halal Logistic
Standard MS 2400: 2010, magazines, and newspaper articles, journals and so forth are the
prime sources in this study.

Keywords: Information Communication Technology (I CT), Halal Transportation, Logistics,

Halal Control
Malaysia has been aggressively seeking to have the upper hand in the
booming and lucrative Halal industry by becoming a global centre for the
manufacture and export of halal product. (Hassan and Mohd Salleh Darus 2008).
Minister in the Prime Minister's Department Tan Sri Nor Mohamed Yakcop said in
The National Toyyiban Halal System Consultative Forum 2010, the local Halal food
industry was worth RM45 billion while the global Halal market was valued at about
RM2 trillion.

A supply chain activity such as transportation, storage and packaging plays

an important role in informing and ensuring products are Halal and Shariah-
compliant. Nowadays, technology is moving forward rapidly and government
committed to ensuring that Information communication technology (ICT) is used to
further enhance the logistic process. The advancement of ICT is seen to be an
important element to assist facilitates Halal logistics industry as Malaysia is
struggling to be the global Halal Hub of Halal products and services to the Muslims
countries all over the world.


Halal transportation is a fascinating new area, where many innovations can

be expected over the coming years. Halal transportation embeds sustainability and
integrity into the supply chain, which will receive global acceptance (Tieman 2009).
The products and services can have a high value if they are in the possession of the
consumer when (time) and where (place) they wish to consume them. It is
transportation that adds place value to products, whereas inventory adds time value.
But in a case of Halal Transportation, Halal product not only must reach consumer in
the right time and at the right place, the halal product integrity must also be
maintained. Zuhra Junaida Mohamad Husny (2010).

Government sees Halal transportation as the next big thing to be implemented. It

is very important to know whether product Halal integrity can be maintained while it
being delivered. The separation of Halal products with non-halal product through
transporting goods to consumer is one of the most critical components of an intact
halal food chain process, and measures need to be put in place throughout to avoid
such contamination from occurring. Zuhra Junaida Mohamad Husny (2010). This is
where ICT can play an important role to verify the Halal integrity of product to the
point of consumption.
Recently, many logistics provides try to improve their operation efficiency by
continuous implementation of information or automation technologies according to
their business characteristics (Mason-Jones 1999). As for container trucking
industry, a comprehensive container management enables manufacturers,
distributors, and third-party logistics providers to achieve complete visibility and
control of the movement of inbound containers and inventory in order to maintained
Halal status of product or goods while it being delivers. Many companies are using
manual, paper-based methods to manage the free time of containers, poor visibility
into what inventory is in which containers, where the container is in transit increased
the risk toward Halal integrity being compromised.
Using ICT to manage distribution and supply chains can increase efficiency and
predictability and reduce waste in value chains and have positive impacts on all
market actors, Enterprise (2009). Since this study is discussed about the Halal
transportation process, the scope will be focus more on the ICT applications for
distribution and supply chain management.

ICT applications presented in this paper are divided into the following
categories: 1) applications that assist in the management of supplier networks, 2)
applications that facilitate traceability, and 3) applications that assist input supply
companies to manage their distribution networks. The table 1 shows the definition of
these three ICT’s and its application.

Table 1: ICT application for distribution and supply chain management

ICT Description Application

- Applications address record keeping,
Database software:
Management monitoring field agent activities, Oracle Enterprise Resource
Supplier procurement operations, credit and Planning (ERP), Microsoft SQL
Network payment tasks, input distribution, Server, Access; Managing data
measuring productivity, and in many forms, from basic lists
forecasting, Enterprise (2009) e.g. customer contacts through
- E.g: Large buyers often use ICT to complex material (e.g.
applications to manage their producer catalogue. Patterson (2003)
supply networks.
Barcode, Radio Frequency
Traceability - Traceability refers to the recording of Identification(RFID), Mobile
movements of products along the Technology(SMS), Zigbee
food chain from production to Signal, Ru Bee Signal,
consumption (i.e., tracing products Geographical Information
back to their source). System(GIS), Global
Positioning System
(GPS).Container Depot
Management Support System
(CDMSS), Giannopoulos
Management These applications include systems that Feed Management Systems
Distribution process seed orders and invoice (FMS), Microsoft-based
Network products electronically, control inventory distribution management
and costs, communicate with clients, and solutions, GPS,GIS,
identify new markets. Transportation management
system (TMS) (through wireless
- E.g: Input supply companies selling
seed, fertilizer, and animal feed
frequently use ICT to help manage
their inventory and rural distribution
networks, Enterprise (2009).

3.1 Research aims and objectives
The aim of this study is to demonstrate an understanding of Halal
transportation process in Malaysia using a value chain analysis perspective. This
study will focus on process, Halal logistic standard, data and information flows
involved in the Halal transportation operation. Besides, this study also discusses the
possibility of ICT for transportation in conducting Halal supply chain.

3.2 Research design

In this study, semi-structured interview and document review are conducted.
The interview involved experts in Halal process also Halal certified company that
transport Halal product. Kontena Nasional Berhad (KNB) together with Halal
authorities in Malaysia is taken as a case study. KNB is Malaysia’s premier logistics
company with 40 years of experience in logistics operations. KNB also Malaysia’s
first JAKIM Halal-certified logistics provider who offering a full range of Halal
logistics services including Halal transportation, Halal warehousing and Samak
services for containers KNB (2010). In meeting the objective of this study, the
summary of the research design is shown in figure 1. The research design was
based on adapted work of NurulHuda (2009).

Phase one : Review the literature of Halal transportation issues and ICT usage in
transport and logistics services.

In this stage, the Halal manual procedures, Malaysian Halal Logistic Standard
MS 2400: 2010, magazines, and newspaper articles, journals, brochures and slides
presentation were reviewed in order to gain the understanding of the Halal
transportation process in Malaysia. The related authority websites such as
MALAYSIAN STANDARD, JAKIM, HDC and Yayasan Ekonomi Sejagat (YES) were
also reviewed. The details are simplified in Table 2.

Figure 1: Research design

Phase two : Conduct interviews with the experts in Halal transportation process.
The semi-structured interviewed was conducted based on one-to-one basis.
Interviews had been conducted in three ways: 1) Unstructured, open-ended
interview, audiotape the interview and transcript it; 2) Unstructured, open-ended
interview using email and; and telephone interview as follow up interview to clarify
answer that was not comprehensible.

Phase Three : Process Mapping : Mapping the Halal transportation process, the
Halal standard procedural rules involved and the typical record for every process. In
the last field, the researcher put the information element to see the possibility of ICT
applications that can be imply towards transportations activities. The data were
collected during the interviewed.

Phase Four : Develop overall Halal Transportation Technologies framework.

The tabulated data that have been collect in phase three is then being used to
develop the theoretical framework of Halal transportation technologies. The
framework was developed using Porter’s value chain analysis whereby this method
is easy to understand and helped identify a firm's core business process by identify
the primary and secondary activities. The Halal transportation technologies
framework was developed based on the four main elements which are (1) the
Table 2 : A review of Halal Transportation Issues in Malaysia (data collection)

Construct Descriptions Authors/websites

Halal Transportation MS2400:2010 Halalan-Toyyiban Assurance Pipeline - Management system requirements for transportation of (Standard 2010)
Standard goods and/or cargo chain services
MS1500:2009, MS1500:2004, HACCP and Good Manufacturing Practice (GMP)
Zuhra Junaida Mohamad
Halal Transportation’s The Needs Of Halal Transportation Control In Malaysia: A Multiple Case Study Approach Husny(201 0) (Tieman 2009)
Research : Niche Area Halal Transportation. Part 2/3: The Building Blocks of Halal Transportation System. (Tieman 2007) (Tieman 2007)
NurulHuda Noordin et al. (2009)
Halal Logistics Orchestrating Model.
The Future of Halal Logistics Solutions
Value chain of Halal Certification System: A Case of the Malaysia Halal Industry

Halal Transportations Sharing container (or segregation), managing returnable container and history of immediate suppliers, Time Zuhra Junaida Mohamad Husny
Issues sensitivity, Warehousing and inventory control , Process of HACCP at receiving inbound gate and goods received (2010)
at the inbound gate should be screened to ascertain its Haram and Hazards status.

Containerized and conventional halal transportation,

Halal Transportation Halal distribution, halal shipping, returns and recalls, Halal freighting for sea and air cargo (KNB 2010)
Services Samak service for containers,
Customs facilities and other halal value-added services.
Consultation of Halal applications to relevant auhorities for consumer food and product manufactures,
destination halal hub/port and customs services

Halal certified company / Kontena Nasional Berhad, TUV Rheinland Malaysia Sdn Bhd, Cold Chain Network (M) Sdn Bhd, Dagang Net Halal (Focus 2010)
players. Technologies Sdn Bhd and ANFACO-CECOPESCA Malaysia Sdn Bhd

Halal Certification Tools/Technologies (Yaakob and M. 2005)

Information technology Halal Gadget