International Journal of Transportation Science and Technology 6 (2017) 8698

Contents lists available at ScienceDirect

International Journal of Transportation

Science and Technology
journal homepage: www.elsevier.com/locate/ijtst

Evaluation of Real-Time Transit Information Systems: An

information demand and supply approach
Xavier J. Harmony, Vikash V. Gayah
Department of Civil and Environmental Engineering, The Pennsylvania State University, University Park, PA, USA

a r t i c l e i n f o a b s t r a c t

Article history: This study assesses current needs in the implementation of Real-Time Transit Information
Received 31 January 2017 Systems. Web surveys are used to better understand information supply and demand,
Received in revised form 23 May 2017 defined as the attitudes and experiences with real-time information of transit passengers
Accepted 26 May 2017
and agencies, respectively. The most valued types of information demanded were found
Available online 1 June 2017
to be related to vehicle location while the least valued information relates to vehicle char-
acteristics, like seating availability. Smartphone applications were found to be the pre-
Keywords:
ferred medium for receiving information followed by Internet/websites and dynamic
Public transportation
Real-time information systems
message signs. The surveys also revealed that demographic and socioeconomic status
Transit information supply and demand influence preferences for real-time information. The information supply survey found that
approximately 70 percent of surveyed agencies currently offer real-time information. The
largest constraint to providing or improving Real-Time Transit Information Systems
(RTTISs) was found to be funding, followed by staffing needs. A comparison between the
survey results found that the information currently being provided by transit agencies is
mostly in line with the information most valued by transit passengers. The few differences
that exist are generally because agencies do not provide information on the media pre-
ferred most by passengers. To address these differences, several suggestions are made to
improve the implementation of real-time information. This information can be used to bet-
ter develop and prioritize investment in real-time information systems.
2017 Tongji University and Tongji University Press. Publishing Services by Elsevier B.V.
This is an open access article under the CC BY-NC-ND license (http://creativecommons.org/
licenses/by-nc-nd/4.0/).

Introduction

Public transportation can provide a socially and environmentally beneficial alternative to private automobiles for the
mass transportation of people. One avenue to improve the transit experience is the deployment of Real-Time Transit Infor-
mation Systems (RTTISs). Current research on RTTISs primarily focuses on the benefits and advantages of providing this type
of information to transit users. First, having real-time information has been shown to significantly affect how users perceive
waiting times for transit service. Passengers are not only more willing to wait for transit, but they perceive their wait times
as being shorter and the service itself as more reliable (Transportation Research Board, 2003a; Watkins et al., 2011). Second,
access to real-time transit information has been found to make transit feel safer (Ferris et al., 2010; Gooze, 2013;
Transportation Research Board, 2003a). Third, these systems allow passengers to make more informed transportation

Peer review under responsibility of Tongji University and Tongji University Press.
Corresponding author.
E-mail address: gayah@engr.psu.edu (V.V. Gayah).

http://dx.doi.org/10.1016/j.ijtst.2017.05.003
2046-0430/ 2017 Tongji University and Tongji University Press. Publishing Services by Elsevier B.V.
This is an open access article under the CC BY-NC-ND license (http://creativecommons.org/licenses/by-nc-nd/4.0/).
 X.J. Harmony, V.V. Gayah / International Journal of Transportation Science and Technology 6 (2017) 8698 87

decisions (Hickman and Wilson, 1995; Maclean & Dailey, 2002). All of these benefits have contributed to more favorable
views of transit (Ferris et al., 2010; Transportation Research Board, 2003a) and, consequently, have been linked to increases
in ridership and mode share (Brakewood et al., 2014a,b; Ferris et al., 2010; Tang and Thakuriah, 2012a,b). Instead of focusing
on the benefits of this information, this paper focuses on how the views of stakeholders influence preferences for, and the
implementation, of RTTISs.
A variety of stakeholders are involved in the development, dissemination, and operation of RTTISs. The key stakeholders
influence information demand (i.e., what type of information is desired) as well as the information supply (i.e., what type of
information is actually provided to transit users). For the purposes of this research, information demand refers to what tran-
sit customers want while information supply is defined as what the transit agencies provide. Unfortunately, information
demand and supply have generally been treated separated in the literature.
In terms of information demand, Caulfield and OMahony used a web survey to gather preferences in Dublin, Ireland
(Caulfield and OMahony, 2007), while Beul-Leusmann et al. used a passenger survey and interviews with transit employees
to obtain information on preferences and challenges with RTTISs in Osnabrck, Germany (Beul-Leusmann et al., 2013). The
former found most respondents used static sources of transit information. When dynamic information was sought, dynamic
message signs (DMS) were the most popular method for accessing the information. The study also found that real-time infor-
mation was more likely to be used as mode certainty decreased and trip complexity increased. The latter study found that
most people preferred to get their information from websites, a change from the Irish paper published a few years
earlier. Both papers identified criticisms with how information was provided; in each case passengers wanted to see an
improvement in how information was communicated. Both papers also found that RTTISs were not likely to increase rider-
ship, although the systems were viewed favorably. RTTISs were found to contribute more towards increasing the comfort
and satisfaction for passengers. There are several limitations with the current research on information demand. First, both
papers represent European perspectives which may be more favorable towards transit than U.S. perspectives (Buehler,
2005). Technology and information options have also changed significantly since both papers were published, the earliest
of which is nearly a decade old. Another limitation was the use of web surveys in an age where Internet use was not
ubiquitous.
On the supply side, various surveys of transit agencies have been performed regarding their implementation of RTTISs
(American Public Transportation Association, 2013, 2015; Transportation Research Board, 2011). The studies found that dif-
ferent agency demographics, like size and modal offering, affected the provision of information, and that an increasing num-
ber of agencies are providing information over time. The biggest reasons agencies do not provide information are funding
constraints or lack of technical availability (American Public Transportation Association, 2013; Transportation Research
Board, 2003b; Yoo et al., 2010) while the biggest reason for providing real-time information is to improve customer service
(Transportation Research Board, 2003b). Feedback on the systems suggests this has been working in favor of agencies
(Transportation Research Board, 2003b). Unfortunately, these studies were limited in the scope of information considered:
generally only vehicle arrival information was included while all other types of real-time information were neglected. This
does not provide a full picture of current real-time information practice.
Overall, both information demand and information supply have demonstrated that technology and information prefer-
ences and availability change over time. Consequently, there is a need to identify current preferences for and implementa-
tion of RTTISs. In addition to this, previous research exploring attitudes towards real-time information was found to
inadequately consider demographic and socioeconomic effects on RTTIS preferences. This is a limitation as other research
in information communication has demonstrated how these differences can affect how information is perceived and under-
stood (van der Meij and Gellevij, 2002; van Hees, 1996; Venkatesh et al., 2003). Other transit research also demonstrates that
different demographic and socioeconomic differences can affect transit preferences (Arabikhan et al., 2016). There are also
opportunities to expand on the types of information offered to transit customers. Finally, no research has compared attitudes
and preferences of both information demand and supply together. This paper will address all of these research opportunities.
In light of this, the goal of this research is to more closely examine what real-time information transit passengers want
and what real-time information agencies provide to identify ways to improve RTTISs. The following specific research objec-
tives were defined to achieve this goal:

1. Identify the types of real-time transit information (RTTI) users consider to be important/useful.
2. Identify the media and methods through which people would like to receive RTTI.
3. Clarify how people prefer to see RTTI presented.
4. Identify the kinds of RTTI that are currently being provided by transit agencies.
5. Summarize the methods transit agencies are utilizing to relay this information to passengers.
6. Identify the constraints/restrictions preventing transit agencies from communicating RTTI.
7. Develop strategies for prioritizing the implementation of RTTI that account for what transit users want and what agencies
can deliver.

The remainder of this paper describes the methodology used to gather and evaluate the information used to address each
of these objectives followed by key results and a comparison of information demand and supply. The paper concludes with a
discussion on potential strategies and applications that could use this information.
88 X.J. Harmony, V.V. Gayah / International Journal of Transportation Science and Technology 6 (2017) 8698

Methodology

Two surveys (one targeted to transit passengers and another targeted to transit agencies) were developed to gain a better
understanding of experiences and preferences with real-time transit information. The full detailed surveys and results can be
found in Harmony (2016). Web surveys were used for ease of distribution, cost effectiveness and quality control issues
(Schonlau et al., 2009; Transportation Research Board, 2006). Potential issues such as sample bias, coverage error, and non-
response error were accounted for by the order questions were asked, the inclusion of demographic questions, and data
cleaning. Note that the use of a web survey might skew the number of responses as those with more immediate Internet
access would be more likely to complete the survey. However, given the proliferation of internet access at the present, this
bias is less impactful than in the past.

Information Demand Survey

The Information Demand survey was targeted to transit passengers and included a total of 36 questions. The questions
were grouped into four categories:

Experiences and attitudes relating to real-time information systems.

Information display preferences.

General transportation questions; and,

Demographics.

Convenience sampling methods were used to distribute the survey. This included encouraging transit agencies to share
the survey with their customers and sharing the survey through a variety of channels that would be seen by transit passen-
gers (professional transit and transportation organizations, transportation blogs, and social media). Although this method
has limitations with quantitative analysis, it can be useful to identify large-scale trends (Fricker and Schonlau, 2002) and
comparison with responses from transit agencies, which was the primary purpose of the survey. It was also prohibitively
expensive to obtain a large number of responses by other means. The survey specifically focused on transit users in the Uni-
ted States for comparison with US transit agency responses from the informational supply survey. Because of the size of the
survey, the number of responses may have been limited due to survey fatigue.

Information Supply survey

In the Information Supply survey, four types of questions were used to understand what transit agencies currently do and
how they make decisions regarding RTTISs:

Identification of information provided.

Identification of issues restricting agencies from expanding on their RTTISs.

Questions on how agencies develop their RTTISs; and,

Agency characteristics.

Two methods of convenience sampling were used to distribute the information supply survey: cold-calling agency per-
sonnel and leveraging professional connections. Most of the agencies contacted for the purposes of completing the survey
were contacted using publicly available information. The response rate for this approach demonstrated this was fairly effec-
tive. Approximately 20 percent of agencies responded positively to the survey request. The second method was more effec-
tive with near complete compliance.

Analysis procedure

Once the survey responses were collected, the responses from each survey were cleaned and then analyzed. The cleaning
process involved identifying erroneous responses that may potentially influence the results and either correcting them or
removing them. As both surveys had several open-ended questions, the data cleaning process also involved coding the
responses from open-ended questions to enable the identification of trends or patterns.
The data results were analyzed using two main methods: single variable frequency (univariate) analysis and cross tabu-
lation. These methods were chosen due to the large amount of categorical data gathered from the survey responses. Due to
the sampling method used to gather survey results, the univariate analysis was completed using only descriptive methods
like frequency distributions. This analysis was performed for every survey question. Cross tabulation involved comparing
responses to questions across unique groups of respondents to identify trends or patterns that differed between these
groups. This analysis was useful for determining what impact, if any, different respondent characteristics had on responses.
When these analyses were completed, statistical tests were performed to determine if any of the differences identified
 X.J. Harmony, V.V. Gayah / International Journal of Transportation Science and Technology 6 (2017) 8698 89

between groups were statistically significant. Statistical tests included chi-squared tests and binomial distribution tests of
proportions, depending on the number of responses for a variable.
Due to the large number of variables in the survey, not all possible cross tabulations were performed. Specific relation-
ships considered for cross tabulation were identified in three ways, including: relationships identified in the literature
review; intuitive relationships; and, relationships based on other results in the analysis. Three-way cross tabulation was con-
sidered for the passenger survey, however, the number of survey responses were not large enough to provide valid results.
An attempt at three-way cross comparisons found that, in almost all cases, samples sizes were too small to be statistically
significant.

Results

Information Demand results

Approximately 400 people responded to at least one question on the transit user survey while 332 people completed the
entire survey. The differences between these numbers could be attributed to interest in the subject matter as well as the
length of the survey. This section describes some of the respondent characteristics as well as a description of respondent
experiences and attitudes with, and preferences for, real-time information.
Both demographic and transportation characteristics of survey respondents were recorded to enable a better understand-
ing of how these differences relate to information preferences. The survey respondents were approximately 60 percent male,
well-educated (82 percent had a bachelor degree or higher), and were well represented across all income and age groups.
Over 62 percent of respondents identify as coming from the United States, a significant proportion of responses. The second
largest geographic group, comprising of 18 percent of respondents, did not identify their location. Consequently, the number
of representatives from the United States is likely to be higher. Compared to national passenger characteristics (American
Public Transportation Association, 2007), respondents were generally more male, more educated, and had higher incomes
but had a similar age distribution.
The user survey showed that both young people and people from low-income groups (<$25,000/year) were more captive
to transit: they were more likely to use transit and are less likely to have access to a car. When low-income people use tran-
sit, they were found to be more likely to use a bus. As Fig. 1 shows, those that use buses as their primary mode choice were
more likely to indicate they would benefit from improvements to RTTISs. The figure also shows there is a difference between
bus and train riders, which is likely because trains are more reliable than buses (due to dedicated space).
Understanding the experiences passengers have with different types of real-time information can provide a snapshot of
what types of information are currently available. This can be useful for identifying gaps in the provision of information. Sur-
vey respondents were asked if they were familiar with each of the various types of real-time information included in this
survey; see Fig. 2. As the figure shows, the most popular types of information experienced include information about the
location of the vehicle while the least experienced information relates to vehicle attributes, like vehicle type or seating avail-
ability. Other types of information that have not been experienced significantly include information about transit facilities,
like parking availability or elevator operational status.
Prior experiences with real-time information are likely to affect user attitudes and perception of RTTISs; e.g., previous use
of RTTISs can help people better determine the value of this information to themselves. For the purposes of this research,
attitudes and perceptions towards real-time information were measured by how important respondents found different

Fig. 1. RTTIS impact on transit experience by mode choice (N = 330).

90 X.J. Harmony, V.V. Gayah / International Journal of Transportation Science and Technology 6 (2017) 8698

Fig. 2. Types of information respondents are familiar with (N = 395).

information as well as the willingness to pay for access to information. Willingness to pay was used as a surrogate to better
understand how people value information. Fig. 3 shows how important different types of real-time information are per-
ceived in the eyes of the survey respondents. As the figure shows, the types of information considered the most important
is generally similar to the types of information that have been experienced the most. This indicates that most of the real-time
information being implemented in RTTISs is either information people want or that people consider information more
important if they have already had experience with it. According to the responses, the least important information includes
information about vehicle attributes as well as facility characteristics. Although neither of these types of information were
found to be important, they could be potentially valuable information to those with special needs, specifically those with
disabilities or are mobility-impaired.
With respect to willingness to pay for information, over half of the respondents were found to not be willing to pay for
real-time information. This is true for all income groups. Although there are small differences in willingness to pay, none of
the differences were found to be statistically significant. Respondents earning over $100,000 a year are just as unlikely to
want to pay for information as respondents earning less than $25,000 a year. Those willing to pay would prefer to pay a
one-time fee to access RTTISs as opposed to a regular subscription. However, the preferred payment type differs based on
the sociodemographic distribution of a population. For example, an analysis of gender found a somewhat statistically signif-
icant relationship for payment preferences between men and women (Df 4; X 2 9:21; p < 0:10). Respondents who iden-
tified as male were almost twice as likely to accept an increase in fares in return for real-time transit information when
compared to respondents who identified as female.
Passenger experiences and attitudes with real-time information could directly influence information preferences. There
are two primary characteristics regarding the provision of real-time information: information placement and information
type (Hickman and Wilson, 1995). Fig. 4 provides the preferred mediums for different types of information, as identified
by respondents. As shown by the figure, there are slight variations in the choice of medium for each type of information.
However, there are some overall preferences observed across all information types. The figure shows that smartphone appli-
cations are the preferred medium for receiving information with the Internet/websites and DMS following in second and
third place, respectively. The least preferred medium was the use of a phone call system. In some cases, respondents indi-
cated that they were not interested in receiving information at all. The information types with these preferences appear to
correlate with the information of low importance identified in Fig. 3.
There are a variety of respondent characteristics that had statistically significant impacts on the way information is pre-
sented. Key characteristics include age, gender, income, and primary transit mode usage. First, older respondents (defined as
 X.J. Harmony, V.V. Gayah / International Journal of Transportation Science and Technology 6 (2017) 8698 91

Fig. 3. Importance of different types of information (N = 399).

55+ in this analysis) have a greater preference for information from phone calls relative to younger people (defined as age
1824) and a lower preference for smartphone applications and the Internet/websites. This seems reasonable as younger
people are more likely to be comfortable using new technologies. Older respondents were also much less likely to want
information about seating availability or the transit vehicle type or capacity. This could be because they either do not see
the benefit of the information or they do not want informational overload.
There were fewer differences found between men and women. Women were found to generally have a greater preference
for text messaging. Women also preferred DMS more than men for the real-time location of vehicles, while men prefer DMS
for information about the next transit service.
Although income did not play a role in the willingness to pay for information, statistically significant relationships were
found in preferences for the provision of real-time information. For example, relative to lower income respondents (defined
here as less than $25,000 per year), higher income respondents (defined as over $100,000 per year) preferred information
displayed on DMS. Lower income respondents were also found to desire seating information less than higher income
respondents.
Differences between transit modes is the last comparison considered here. Comparisons were attempted between
respondents who primarily used transit and non-transit modes as well as between respondents who primarily used cars
and those who used transit, but none of the differences were statistically significant. However, there were significant differ-
ences between those who use a bus and those who use trains. Compared to bus riders, train riders generally prefer Internet/
website information and smartphones for their information. The most significant differences involve preferences for what
information should not be shown at all. Bus riders have a greater preference for seating availability information while train
riders have a greater preference for information about the operational status of elevators and escalators. It makes sense that
bus riders prefer seating information more than train riders as there is generally more space allocated for seating on buses
than trains, buses are generally more uncomfortable than trains so seating is more valuable, and crowding seems to be a
greater issue on buses than on trains. It also makes sense that the operational status of elevators and escalators is greater
for trains as train stations are much more likely to have these systems than bus stops.
92 X.J. Harmony, V.V. Gayah / International Journal of Transportation Science and Technology 6 (2017) 8698

Fig. 4. Preferred mediums for different information types (N = 390).

Information supply results

A total of 58 transit agencies from across the U.S. contributed to the information supply survey results with 24 completing
the entire survey. These transit agencies varied in both size and the types of services offered. Of the agencies surveyed, 62.5
percent were small (less than 75 fixed-route buses operating in a peak hour), five percent were moderately sized (between
75 and 100 fixed-route buses operating in a peak hour), and 32.5 percent were large (more than 100 fixed-route buses oper-
ating in a peak hour). The agencies who were responded represented a variety of geographic locations with approximately 22
percent from rural areas, 43.9 percent from small urban areas, 24.4 percent from large urban areas, and 9.8 percent from
extra-large urban areas. Most agencies who responded to the survey stated they offered bus and paratransit services
(90.2 and 68.3 percent, respectively) while approximately 30 percent of agencies stated they provided one or more types
of rail. Other types of services offered by responding agencies include taxi, vanpool, and ferry services. As Fig. 5 shows, transit
mode varies by geographic location. While rail is predominantly offered in large or extra-large urban areas, bus and para-
transit services have a similar distribution across all four geographic areas.
Of the responding agencies, approximately 69 percent offered some form of real-time information. However, real-time
information provision was found to depend on several different characteristics like agency size, geographic location, and
the type of services provided. While all responding moderate and large agencies stated they provided real-time information,
only 44 percent of small agencies said the same. In addition to this, a trend between geographic areas and the provision of
real-time information was identified: regions that were more urban were more likely to offer real-time information.
Although this relationship appears reasonable, this is opposite of what might be the most beneficial for transit customers.
For example, vehicle headways are generally larger in rural areas so deviations from schedule or disruptions to a service
could be more impactful (Papangelis et al., 2013).
Fig. 6 shows the types of information offered by responding agencies. When information is not provided, agencies were
able to respond with Not Applicable. As the figure shows, the information types most often provided by transit agencies
include information about the vehicle locations, route disruptions, and emergencies. Information that were not provided
well includes seating information, information about the type of transit vehicle being used, and information about parking
availability. The mediums used for providing the information range between agencies and information types. Overall, it is
clear information is primarily provided through websites followed by smartphone applications and DMS. It is also evident
 X.J. Harmony, V.V. Gayah / International Journal of Transportation Science and Technology 6 (2017) 8698 93

Fig. 5. Transit mode by geographic area (N = 41).

that text messaging and phone calls are popular methods for providing various types of information, such as arrival and
departure times for the next transit vehicle or information about service disruptions.
To gain insight into how transit agencies have been prioritizing the implementation of real-time information, each agency
was asked how long they have provided the different types of information. The responses indicated most types of informa-
tion have been implemented by agencies for five years or more. Parking availability is the only form of information that has
not been provided for at least five years. This information suggests that the ability to provide most types of information has
existed for many years. Consequently, there may be other reasons why not all of the information types identified are found in
a wider range of RTTISs.
The information supply survey also included questions about how RTTISs are currently funded, as funding is a source of
concern for many transit agencies. Specifically, agencies were asked about how the installation and maintenance of RTTISs
were funded. There was significant variability in how the implementation of RTTISs was funded. About 30 percent of
responding agencies appear to use their existing budgets while others use a combination of funds ranging from local sales
taxes to investment from the Federal Transit Administration. However, the funding of RTTIS maintenance was more consis-
tent with over 50 percent of agencies responding stating that maintenance is funded through their operating budget. Con-
sequently, the addition of a RTTIS results in an additional, on-going expense that has to be accounted for without a guarantee
in increased revenue. Furthermore, all responding agencies stated costs were not directly passed onto the customer. How-
ever, a few agencies did indicate that costs were indirectly passed on through changing fare recovery or altering the choice of
investments for the agency.
The agency survey also asked agency representatives about how RTTISs are developed. The first set of questions asked
representatives to rank several different factors on their influence in the development of RTTISs. The results revealed the fol-
lowing general ranking among agencies that participated:

1. Cost of system
2. Customer demand
3. Technological availability
4. Service reliability (tied with Service frequency)
5. Differently-abled user requirements
6. Other factors

As expected, the most important factor when developing RTTISs is cost, closely followed by customer demand. These
rankings also indicate that differently-abled users are not very influential in the development of many RTTISs. This is an
94 X.J. Harmony, V.V. Gayah / International Journal of Transportation Science and Technology 6 (2017) 8698

Fig. 6. What real-time information agencies provide and how they provided the information (N = 29).
 X.J. Harmony, V.V. Gayah / International Journal of Transportation Science and Technology 6 (2017) 8698 95

important point as the literature indicates differently-abled users have a lot to benefit from a RTTIS and use these systems to
be more independent (Steinfeld et al., 2011). The Other category primarily emphasized passenger demand and technology.
However, equipment status was also mentioned, suggesting previous experience with system failure or inaccuracy can influ-
ence how real-time information is prioritized.
The survey also asked questions on what factors limited an agencys ability to implement or expand RTTISs. The results
found the most common limiting factor was funding, which confirmed what was found in the APTA survey (American Public
Transportation Association, 2013). Staff resources have not been explicitly identified in previous literature but were found to
be the second most popular restriction for improving or implementing RTTISs. The results also showed almost nine percent
of systems are currently being installed or updated, a number similar to what was found in the research by APTA.
Agencies responding to the information supply survey were also asked questions about customer involvement in the
development of RTTISs and opportunities for customers to provide feedback on said systems. Most (61.5 percent) agencies
stated customers were not involved in the initial development of RTTISs. For those that incorporated customer involvement,
a variety of methods were used, including in-depth methods like public meetings and focus groups to less involved methods
like asking for feedback once the system has already been implemented. Approximately 84.6 percent of agencies stated they
provided feedback opportunities for customers after system installation. Feedback consisted of both active and passive
methods, including: website forms designed for feedback, email, surveys, and social media. No responding agencies stated
that they provided opportunities for feedback through any part of the RTTIS explicitly. However, it is possible that some of
the responding agencies who provide feedback through their website also provide real-time information through the same
website.
Finally, transit agencies were asked about what issues and concerns they had with real-time transit information. The
information supply survey provided an open-ended question to allow responding transit agencies to report what their big-
gest concerns with RTTISs were. The responses were coded by themes and are summarized. The most important issue was
with accuracy and reliability of the information being provided. In general, information accuracy is an important concern
from a customer point of view (Hickman and Wilson, 1995; Transportation Research Board, 2003a), but the survey results
show it is also very important for agencies. The second largest concern agencies have is with the cost of real-time systems, an
issue that has been previously mentioned. The cost of the system is also directly related to system maintenance, recorded as
the third greatest concern by agencies. An important concern that was noted in the agency responses was the question of
whether RTTISs provide value. Evidently, there may be a need to better demonstrate what benefits are generated by RTTISs
so that agencies can better understand the impact and the return on investment of providing real-time information.
Although there are many issues and concerns that have been identified by transit agencies, almost 20 percent of agencies
stated they have not experienced any issues or concerns with RTTISs.

Discussion and applications

The surveys created and used in this research examined both demand and supply for RTTISs. Both expected and unex-
pected trends and relationships have been observed in these data sets. When the results from the two data sets are com-
pared, several similarities and differences can be identified. Table 1 summarizes this comparison. As the results of the
surveys indicate, transit agencies appear to be doing an adequate job of providing all types of real-time information. Transit
agencies are doing a particularly good job with information relating to the location of a transit vehicle or arrival/departure
times of a vehicle.
However, there are several opportunities for improvement; specifically with respect to providing more information in
cases where it is not well-provided and providing information in different ways. For example, not many transit agencies pro-
vide seating information; however, the information demand survey indicated bus riders are interested in this information.
This results in a disconnect between what is desired and what is provided. For agencies that use Automated Passenger Coun-
ter (APC) systems on their vehicles, this information can be easily incorporated into RTTISs to alleviate this gap. For those
that do not, cheaper (but less reliable) alternatives might be a way to obtain this information, such as crowd sourcing. Gaps
were also identified with respect to information placement. For some types of information, transit agencies do not appear to
be providing it using mediums customers desire. There are five primary types of information that are affected by these gaps,
including:

& Seating availability: More information should be provided by smartphone applications and DMS. This is particularly
important for real-time information for buses, as indicated by the information demand survey.
& Information for planned detours and special events: More information should be provided on DMS.
& Identification of unplanned service disruptions: More information should be provided on smartphone applications.
& Emergency information: More information should be provided on DMS.
& Information on availability and operational status of elevators/escalators: More information should be provided on smart-
phone applications.

In some cases, inconsistencies between information supply and demand reveal transit agencies are doing a better job at
providing real-time information than is desired by customers. For example, transit agencies provide information through
96 X.J. Harmony, V.V. Gayah / International Journal of Transportation Science and Technology 6 (2017) 8698

Table 1
Comparison between passenger and agency preferences for real-time information.

Information Type Comments Adequately

Arrival/departure times for next vehicle
Real-time location of transit vehicles
Seating availability
Number of cars on next train or type of next
bus
Real-time trip planning
Meets needs?
& The most desired type of information by passengers is the best provided Yes
information by agencies
& Information well provided for in many different ways
& Highly valued by passengers and provided by many agencies Yes
& The two most desired mediums for providing information are the two best
supported mediums provided by agencies
& Relatively low importance for passengers but more important for bus passen- No
gers; hence, too few agencies offer information
& More information needs to be offered on smartphones and DMS
& One of the least valued types of information and least provided by agencies Yes
& Information is provided through multiple platforms when it is available
+ Very important to passengers and many agencies provide information Yes
& Well provided for in most desired mediums for providing this information

Information for planned detours and special + Very important to passengers and many agencies provide information No
events & More information needs to be provided on DMS

Identification of unplanned service + Very important to passengers and many agencies provide information but No
disruptions more need to do so
& More information needs to be provided on smartphone applications

Emergency information + Somewhat important information that is well provided by agencies No

& Needs to be better provided on DMS

Information on availability and operational + Lower importance by passengers but it is relatively well provided by agencies No
status of elevators/escalators & Information needs to be better provided for on smartphone applications

Parking availability + One of the least important types of information and least provided by No
agencies
& Information needs to be better provided for on smartphone applications

Symbology: + Conclusions obtained directly from survey; & Comments obtained by comparing demand and supply surveys.

text messaging and phone calls more frequently than what is desired by respondents in the information demand survey.
Consequently, in these cases, transit agencies may have an opportunity to downsize the way they provide information in
these areas. However, if agencies do choose to do this they should recognize the impact this may have on other users
who may value this information, such as the elderly or sight-impaired.
To address these differences identified in the comparison between information supply and demand, several strategies are
proposed to improve the implementation of real-time information. These include the following:

& Improve public involvement: The information demand survey showed passenger characteristics could influence prefer-
ences while the information supply survey showed only 40% of agencies involve their customers in the design of systems.
Increased public involvement can help agencies tailor their systems for their customers.
& Reduce reliance on cookie-cutter smartphone application developers: As the previous strategy mentioned, passenger
characteristics can significantly influence RTTIS preferences. Consequently, systems need to be customized each different
transit catchment areas. This is also important when developing bus or train systems, as the preferences for each could be
different.
& Improve feedback for commuters: Ongoing dialogue between transit agencies and their customers could address con-
cerns like information accuracy and reliability. This will benefit both parties; passengers will be aware that agencies
are working on accuracy issues and agencies may receive more positive feedback from their customers.
& Provide different cost tiers for information: Information demand survey indicated those who are willing to pay for real-
time information value information differently to those who are not willing to pay. This could be used to develop pre-
mium information offerings for those that are willing to pay. Basic and universally demanded real-time information,
such as arrival and departure times for vehicles, could be provided for free, which ensures the vulnerable, low-income
passengers are still able to access the information. Additional information, such as seating availability, could be provided
at a price.
& Utilize alternative sources for the information: The source information provided by transit agencies indicates that there
can often be more than one way to gain information for the purposes of informing customers. The information demand
survey indicated some information is desired but not considered important. Consequently, cheaper alternative sources of
information could be used in these cases. For example, seating information could be provided through crowd sourcing
rather than through APC.
& Generate passive income through advertising on smartphone applications and websites: Both the information supply and
demand results indicated smartphone applications and websites are important and well-used. If transit agencies sell
advertising space on their websites or on their smartphone applications it would allow RTTISs to generate passive income.
As many smartphone applications and websites already generate income this way, this idea is feasible and should be
easily implementable from a technological perspective.
 X.J. Harmony, V.V. Gayah / International Journal of Transportation Science and Technology 6 (2017) 8698 97

Conclusions

The purpose of this project was to evaluate and strategize the implementation of Real-Time Transit Information Systems
(RTTISs) through an examination of information supply and demand. For information demand, the most valued types of
information were found to be information about the location of the vehicle while the least valued information relates to
other information about the vehicle itself, like seating availability. In terms of media preferences, smartphone applications
were found to be the preferred medium for receiving information with Internet/websites and dynamic message signs follow-
ing in second and third place. The way information is displayed depends on the purpose of the information as well as how it
is presented. Finally, it was found that different demographic and socioeconomic groups could influence preferences for real-
time information.
The information supply survey found that approximately 69 percent of agencies offer real-time information. The provi-
sion of information was influenced by agency characteristics. The primary constraints for providing real-time information
were found to be similar to what was found in the real-time information literature. Funding is the largest issue while staffing
needs are the second largest. Finally, although real-time information literature mentions the importance of providing infor-
mation for the differently-abled, the survey found differently-abled user requirements were one of the least important fac-
tors influencing the implementation of RTTISs.
A comparison between the surveys found that the information currently being provided by transit agencies is mostly sim-
ilar to the information most valued by transit passengers. When there were differences between supply and demand it was
generally because agencies were not providing information in the form most preferred by transit users. Clear examples were
identified where RTTISs can be improved simply by changing the information medium. Through these findings and strate-
gies, real-time transit information can be better understood, especially considering the differences between information sup-
ply and demand. This information can hopefully be used to better develop and prioritize investment in real-time information
systems.
Several opportunities for future work have been identified, including:

& Focusing on riders served by agencies surveyed: This research focused on agencies and transit users from across the U.S.
Due to sample size issues, responses from riders could not be matched directly to individual agencies (or the group of
agencies surveyed) in a statistically significant way. Further work should aim to increase the sample size of transit riders
served by the set of agencies surveyed.
& Differences between socioeconomic and demographic groups: Some of the findings indicated demographic or socioeco-
nomic differences that are difficult to explain. Consequently, future research could focus on some of these groups and
identify why they may have particular preferences when it comes to real-time information.
& Reported versus observed differences: Real-time information research has found that there may sometimes be differences
between how someone self-reports the influence of real-time information and how it actually affects them. Consequently,
these relationships may need to be confirmed with other data.
& Transit mode specific differences: The survey found that there were differences between people who primarily used dif-
ferent transit modes. Future research could target these differences and ask different questions depending on whether
someone was seeking information in a bus system, a train system, or some other transit system.

Acknowledgements

The authors would like to acknowledge the help of Hugh Mose who assisted with the information supply survey design
and survey data collection.

References

American Public Transportation Association. 2007. A Profile of Public Transportation Passenger Demographics and Travel Characteristics Reported in On-
Board Surveys. Washington, DC.
American Public Transportation Association. 2013. APTA Surveys Transit Agencies on Providing Information and Real-Time Arrivals to Customers.
Washington, DC.
American Public Transportation Association. 2015. Update on Public Transportation Agencies Providing Real Time Data. Washington, DC.
Arabikhan, F., Postorino, M. N., Dupont-kieffer, A., & Gegov, A. 2016. Gender-based analysis of zones of tolerance for transit service quality. In:
Transportation Research Board 95th Annual Meeting, Washington, DC.
Beul-Leusmann, S., Jakobs, E. M., & Ziefle, M. 2013. User-centered design of passenger information systems. In: IEEE International Professional
Communication Conference, Vancouver, BC, pp. 18.
Brakewood, C., Barbeau, S., Watkins, K., 2014a. An experiment evaluating the impacts of real-time transit information on bus riders in Tampa, Florida.
Transport. Res. Part A Policy Practice 69, 409422.
Brakewood, C., Macfarlane, G.S., Watkins, K.E., 2014b. The impact of real-time information on bus ridership in New York City. Transport. Res. Part C
Emerging Technol. 53, 5975.
Buehler, R., 2005. Promoting public transportation comparison of passengers and policies in Germany and the United States. Transport. Res. Record J.
Transport. Res. Board 2110, 6068.
Caulfield, B., OMahony, M., 2007. An examination of the public transport information requirements of users. IEEE Trans. Intell. Transp. Syst. 8 (1), 2130.
Ferris, B., Watkins, K., Borning, A. 2010. OneBusAway: results from providing real-time arrival information for public transit. In: Proceedings of the SIGCHI
Conference on Human Factors in Computing Systems, pp. 110.
98 X.J. Harmony, V.V. Gayah / International Journal of Transportation Science and Technology 6 (2017) 8698

Fricker, R.D., Schonlau, M., 2002. Advantages and disadvantages of internet research surveys: evidence from the literature. Field Methods 14 (4), 347367.
Gooze, A., 2013. Real-Time Transit Information Accuracy: Impacts and Proposed Solutions. Georgia Institute of Technology.
Harmony, X.J., 2016. Evaluating the Implementation of Real-time Transit Information through an Examination of Information Supply and Demand. The
Pennsylvania State University.
Hickman, M.D., Wilson, N.H.M., 1995. Passenger travel time and path choice implications of real-time transit information. Transport. Res. Part C Emerg.
Technol. 3 (4), 211226.
Maclean, S., & Dailey, D. 2002. Measuring the utility of a real-time transit information system. In: Proceedings The IEEE 5th International Conference on
Intelligent Transportation Systems, Singapore, pp. 846850.
Papangelis, K., Sripada, S., Corsar, D., Velaga, N., Edwards, P., & Nelson, J. D. 2013. Developing a real time passenger information system for rural areas. In:
Human Interface and the Management of Information. Information and Interaction for Health, Safety, Mobility and Complex Environments, vol. 8017,
Berlin, pp. 153162.
Schonlau, M., van Soest, A., Kapteyn, A., Couper, M., 2009. Selection bias in Web surveys and the use of propensity scores. Soc. Methods Res. 37 (3), 291318.
Steinfeld, A., Zimmerman, J., Tomasic, A., Yoo, D., Aziz, R.D., 2011. Mobile transit rider information via universal design and crowdsourcing. Transport. Res.
Rec. 2217, 95102.
Tang, L., Thakuriah (Vonu), P., 2012a. Will psychological effects of real-time transit information systems lead to ridership gain? Transport. Res. Rec. J.
Transport. Res. Board 2216, 6774.
Tang, L., Thakuriah, P.V., 2012b. Ridership effects of real-time bus information system: a case study in the City of Chicago. Transport. Res. Part C Emerg.
Technol. 22, 146161.
Transportation Research Board. 2003a. TCRP Report 92: Strategies for Improved Traveler Information, Vol. 1., Washington, DC.
Transportation Research Board. 2003b. TCRP Synthesis 48: Real-Time Bus Arrival Information Systems. Transit Cooperative Research Program. Washington,
DC.
Transportation Research Board. 2006. TCRP Synthesis 69: Web-based Survey Techniques. TCRP Synthesis 69: Web-Based Survey Techniques, Washington,
DC.
Transportation Research Board. (2011). TCRP Synthesis 91: Use and Deployment of Mobile Device Technology for Real-Time Transit Information,
Washington, DC.
van der Meij, H., Gellevij, M., 2002. Effects of pictures, age, and experience on learning to use a computer program. Tech. Commun. 49 (3), 330339.
van Hees, M., 1996. User instructions for the elderly: what the literature tells us. J. Tech. Writing Commun. 26 (4), 521536.
Venkatesh, V., Morris, M.G., Davis, G.B., Davis, F.D., 2003. User acceptance of information technology: toward a unified view. MIS Quarterly 27 (3), 425478.
Watkins, K.E., Ferris, B., Borning, A., Rutherford, G.S., Layton, D., 2011. Where Is My Bus? Impact of mobile real-time information on the perceived and actual
wait time of transit riders. Transport. Res. Part A Policy Practice 45 (8), 839848.
Yoo, D., Zimmerman, J., Steinfeld, A., & Tomasic, A. (2010). Understanding the space for co-design in riders interactions with a transit service. In:
Proceedings of the 28th International Conference on Human factors in Computing Systems, Atlanta, GA, pp. 17971806.