1

CHAPTER-1
INTRODUCTION
1.1 NEUROREHABILITATION ROBOT
Neurorehabilitation is a complex medical process which aims to aid recovery from a
nervous system injury, and to minimize and/or compensate for any functional alterations
resulting from it. Among many types of neurorehabilitation robots, there is a recent trend of
highlighting exoseleton robots because of the following advantages of over !nd "!ffector
#!!$ type robots. %wing to the close alignment of anatomical axes of the exoseleton robot,
all the human arm joints angles and tor&ues can be directly measured and individually
controlled and also computing the joint tor&ues. 'he relation between the joint angle and
tor&ue #i.e. the impedance or stiffness$ can be directly computed. (sing other type robots,
one cannot obtain elbow, wrist angles, tor&ues and impedance simultaneously. 'he )ange of
*otion #)%*$ with exoseleton robots might be larger than that with !! type robots.
+iagnosis, physical therapy and outcome evaluation are important and essential steps
of rehabilitation and are thus preferred to be integrated for effective treatment of complex
neurological impairments. %n the other hand, passive stretching reduce the joint/muscle
stiffness and to increase the muscle strength, and active movement training to recover the
motor functions. 'he existing robots have been used to evaluate the impairments post stroe
and the therapy on a single joint. ,or clinicians, it is infeasible to diagnose the changes in the
many +%,s and joints simultaneously and &uantitatively. 'hus to aid clinicians in planning
therapy by providing *-*+ diagnosis of passive/active impairments, a rehabilitation robot
with comprehensive measurements of relevant *-*+ variables is used.
.
1.2 INTEGRATED CAPABILITIES


Fig 1.1 !ssential /teps of Neurorehabilitation
0uantitative, objective and comprehensive *-*+ pre1evaluation capabilities aiding
diagnosis for individual patients.
/trenuous and safe passive stretching of deformed arm for loosening up
muscles/joints based on the robot1aided diagnosis.
Active movement training after the passive stretching for improving motor control
ability.
0uantitative and comprehensive outcome evaluation at the level of individual joints,
multiple joints.


2re !valuation
2assive /tretching
Active *ovement 'raining
%utcome !valuation
3
CHAPTER-2
LITERATURE REVIEW
Robust I!"ti#i$%tio" o# &u'ti-(oi"t Hu)%" A*) I)+!%"$! B%s! O"
D,"%)i$s D!$o)+ositio"- A &o!'i"g Stu,
*ulti1joint/*ulti1+egree of ,reedom #+%,$ human arm impedance estimation is
important in many disciplines. 4owever, as the number of joints/+%,s increases, it may
become intractable to identify the system reliably. A robust, unbiased and tractable estimation
method based on a systematic dynamics decomposition, which decomposes a *ulti15nput
*ulti1%utput #*5*%$ system into multiple /ingle15nput *ulti1%utput #/5*%$ subsystems,
is developed. Accuracy and robustness of the new method were validated through a human
arm and a .1+%, exoseleton robot simulation with various magnitudes of sensor resolution
and nonlinear friction. 'he approach can be similarly applied to identify more sophisticated
systems with more joints/+%,s involved.
A Robot #o* P%ti!"t-Coo+!*%ti.! A*) T/!*%+,
'his paper presents a new method of trajectory planning in rehabilitation robotics.
,irst were measured in healthy subject the pic to place trajectories while haptic robot is in
zero impedance space. 61spline approximation is used to mathematically define the measured
paths. 'his trajectory path serves as a central line for the rounding haptic tunnel. 5n addition
to radial elastic and damping force an optional guidance force can be applied along the tunnel
to reach the place point. 'he 61spline control points were observed around the robot and arm
worspace. 'he trajectory path defined with 61splines is compared with minimum jer and
minimum tor&ue defined trajectories. ,inally are compared the pic to place movements with
and without tunnel use in healthy subject and in stroe hemiplegic patient.
Sto$/%sti$ Esti)%tio" o# Hu)%" A*) I)+!%"$! u"!* No"'i"!%* F*i$tio"
i" Robot (oi"ts- A &o!' Stu,
7
'he basic assumption of stochastic human arm impedance estimation methods is that
the human arm and robot behave linearly for small perturbations. 5n the present wor, we
have identified the degree of influence of nonlinear friction in robot joints to the stochastic
human arm impedance estimation. 5nternal *odel 6ased 5mpedance 8ontrol #5*658$ is then
proposed as a means to mae the estimation accurate by compensating for the nonlinear
friction. ,rom simulations with a nonlinear 9ugre friction model, it is observed that the
reliability and accuracy of the estimation are severely degraded with nonlinear friction:
below . 4z, multiple and partial coherence functions are far less than unity; estimated
magnitudes and phases are severely deviated from that of a real human arm throughout the
fre&uency range of interest; and the accuracy is not enhanced with an increase of magnitude
of the force perturbations. 5n contrast, the combined use of stochastic estimation and 5*658
provides with accurate estimation results even with large friction: the multiple coherence
functions are larger than <.= throughout the fre&uency range of interest and the estimated
magnitudes and phases are well matched with that of a real human arm. ,urthermore, the
performance of suggested method is independent of human arm and robot posture, and
human arm impedance. 'herefore, the 5*658 will be useful in measuring human arm
impedance with conventional robot, as well as in designing a spatial impedance measuring
robot, which re&uires gearing.
Robot 0Ai! N!u*o*!/%bi'it%tio"
%ur goal is to apply robotics and automation technology to assist, enhance, &uantify,
and document neurorehabilitation. 'his paper reviews a clinical trial involving .< stroe
patients with a prototype robot1aided rehabilitation facility developed at the *assachusetts
5nstitute of 'echnology, 8ambridge, #*5'$ and tested at 6ure )ehabilitation 4ospital,
>hite 2lains, N?. 5t also presents our approach to analyze inematic data collected in the
robot1aided assessment procedure. 5n particular, we present evidence that 1$ robot1aided
therapy does not have adverse effects, .$ patients tolerate the procedure, and 3$ peripheral
manipulation of the impaired limb may influence brain recovery. 'hese results are based on
standard clinical assessment procedures. >e also present one approach using inematic data
in a robot1aided assessment procedure.
@
Sto$/%sti$ Esti)%tio" o# A*) &!$/%"i$%' I)+!%"$! u*i"g Roboti$
St*o1! R!/%bi'it%tio"
'his paper presents a stochastic method to estimate the multi1joint mechanical
impedance of the human arm suitable for use in a clinical setting, e.g., with persons with
stroe undergoing robotic rehabilitation for a paralyzed arm. 5n this context, special
circumstances such as hyper tonicity and tissue atrophy due to disuse of the hemiplegic limb
must be considered. A low1impedance robot was used to bring the upper limb of a stroe
patient to a test location, generate force perturbations, and measure the resulting motion.
*ethods were developed to compensate for input signal coupling at low fre&uencies
apparently due to human1machine interaction dynamics. +ata was analyzed by spectral
procedures that mae no assumption about model structure. 'he method was validated by
measuring simple mechanical hardware and results from a patientAs hemiplegic arm are
presented.
Usi"g So$i%'', Assisti.! Roboti$s to Aug)!"t &oto* T%s1 P!*#o*)%"$! i"
I"i.iu%'s Post0St*o1!
'his paper presents an application of a socially assistive robotics system to hands1off
post"stroe rehabilitation. >e validate the technical feasibility and efficiency of our system
in guiding, motivating, and administering an upper extremity rehabilitation tas. 'he robot,
which consists of a humanoid torso on a mobile base, monitors user performance on a wire
puzzle tas through a wearable inertial measurement unit and signals from the puzzle.
/moothness of stroe1affected limb movement is used as the evaluation metric. ,ive adults of
mild to moderate functional ability in the chronic phase of stroe recovery interacted with
our system over three separate days.
CHAPTER-2
S3STE& DESCRIPTION
B
2.1 &ETHODS USED IN NEUROREHABILITATION S3STE&
2.1.1 I"t!''iA*)- A" U++!* Li)b E4os1!'!to" Robot Fo*
N!u*o*!/%bi'it%tio"
5t was developed for clinicians to aid diagnosis and outcome evaluation as well as to
*-*+ assist physical therapy based on the robot aided diagnosis. ,or pre1evaluation,
physical therapy, and outcome evaluation the subject #forearm, hand$ were strapped to the
corresponding braces of the intelliarm mechanical axes.
'he intelliarm can independently control the following +%,s of human arm: elbow
,lexion1!xtension #,l1!x$ in horizontal plane, forearm 2ronation"/upination #2r1/u$ and
wrist ,l1!x. !ach +%, is driven by a servomotor placed on the corresponding human arm
joint axis.
/ince stroe survivors often develop pronation deformity of the forearm, it is
important to control and move the forearm in a proper range of pronation. ,or the controlled
movement of forearm a servo motor is used. 'he maximum output tor&ue and speed of
forearm driving system is 1<.. Nm and 7=.@deg/sec, respectively. 'here six1axis force/tor&ue
sensors are located at the elbow and wrist joints can measure three dimensional forces and
tor&ues at those joints. 'he )esistance 'or&ue measurement #)'$ for the wrist ,91!x and
forearm 2r1/u can be affected by the gravitational tor&ue of the plate #the plate is attached by
the wrist$. 'herefore the tor&ue can be obtained by subtracting the gravitational tor&ue from
the measured tor&ue, without attaching human arm to the intelliarm and using standard least
s&uare method.
'here are two mode of operation passive mode and active mode. 5n the passive mode
#i.e., robot drives human arm joints$ and the active mode #i.e., human arm drives the robot$.
,or the safety of patients, the joint/+%, angles and )'s are monitored by the central
digital controller in real1time, and if either in its out of range, the whole intelliarm system then
shut down in no time. A stop switch is given to the operator and to the subject to authorize
them to shut down the system at any time.
2.1.2 &u'ti-(oi"t P*!-!.%'u%tio" Fo* N!u*o)!$/%"i$%' C/%"g!s
C

'he *-*+ neuromechanical changes associated with the arm impairment post stroe
were characterized systematically by *-*+ stiffness1the individual joints stiffness and
cross1coupled stiffness between joints/+%,s "during controlled passive movements and loss
of individuation during the active movement.


Fig 2.1 'ypical angle1resistance tor&ue curve
5n the passive mode of operation to minimize reflex contributions and manifest the
passive mechanical changes of muscles/joints, the intelliarm passively moved one targeted
joint/+%,s at a time #,ig 3.1$ among the controlled +%,s of the subjectDs arm throughout its
)%* with a controlled speed and cycles, until its joint/+%, )',*res,reached its pre1
specified 2ositive pea )'#2)'$,*p, or negative )' ,*n ,#path 1 and 3 in fig3.1$;and if
*res reached wither *p or *n, then the movement direction was reversed after few seconds.
'he 2)%* of the targeted joint/+%, was determined from the measured *res and
angle , of the targeted joint/+%,. 6ecause of the hysteresis loop consist of two paths as
observed in the angle1)' as follows: positive end of the 2)%* # p1prm$ and negative end of
the 2)%* # n1prm$ in fig 3.1. ,or each joint/+%,, individual joint/+%, stiffness at p1prm
and n1prm #Ep and En respectively in fig 3.1$ was then derived by computing the slope of
the curve.
2.1.2 St*!"uous %" S%#! &u'ti-5oi"t I"t!''ig!"t St*!t$/i"g
F
'he movement and control of the elbow, wrist joints are closely coupled, because of
dozens of muscles and other soft tissues crossing the joints, and some crossing multiple
joints. 'hus for effective treatment of arms with excessive couplings, the elbow, wrist should
be treated together in a well1coordinated manner.
,rom the robot"aided multi1joint pre1evaluation aiding diagnosis, the joints/+%,s
with increased individual joint/+%, stiffness, excessive cross1coupled stiffness, large 8's,
and the associated arm postures were identified. 'he intelliArm then stretched either multiple
joints or +%, simultaneously or a joint/+%, individually in a safe manner by using the 5//
to reduce increased stiffness values of the joints/+%,s involved. 'he fingers are not directly
stretched, because of the possible coupling between the fingers and other joints.
2.1.6 &u'ti-5oi"t A$ti.! &o.!)!"t T*%i"i"g
After the controlled stretching reduced the excessive individual joint/+%, stiffness
and cross coupled stiffness, the neural command might be able to control the muscles better
and also to move the arm better.
+uring the active movement training, the intelliarm was made bac1drivable under
the 5*658 #5nternal *odel 6ased 5mpedance 8ontrol$. /ubjects were able to move their arm
freely with the intelliarm to match or trac targetDs displayed on monitor.
2.1.7 &u'ti+'! 5oi"t Robot Ai! Out$o)! E.%'u%tio"s
'he outcome evaluation was performed in terms of the biomechanical properties and
motor "control ability induced by the passive stretching and active movement training at the
multiple joints involved.
5n the passive mode, the elbow, wrist of the impaired arm of patients was moved by
the intelliarm throughout the )%*s individually or simultaneously under precise control.
5n the active mode, the patients were ased to move one of the impaired joints /+%,s
at a time and to move the multiple joints of the whole "arm simultaneously for functional
movements. 'he neuromechanical changes in the impaired arm after treatments were
evaluated using the data collected from the *-*+ passive and active movements.
=
2.2 BLOC8 DIAGRA& OF NEUROREHABILITATION
E9OS8ELTON ROBOT S3STE&
%N/%,,
/>5'84
2581B,FCCA
*%+!
/!9!8'5%N

E!?2A+
/!)G% *%'%)
#*1$
/!)G% *%'%)
#*.$
1<

Fig 3.2 Block Diagram of Neurorehabilitation Robot System
2.2.1 O+!*%tio" o# t/! N!u*o*!/%bi'it%tio" Robot S,st!)
'he central part is a microcontroller that has been used for this project is from 258
series. 258 microcontroller is the first )5/8 based microcontroller fabricated in 8*%/
#8omplementary *etal %xide /emiconductor$ that uses separate bus for instruction and data
allowing simultaneous access of program and data memory.
5t consists of three servo motors is used for elbow ,l1!x, wrist ,l1!x and forearm 2r1
/u. /ervo motors are controlled by sending an electrical pulse of variable width, or 2ulse
>idth *odulation #2>*$, through the control wire. 'here is a minimum pulse, a maximum
pulse, and a repetition rate. A servo motor can usually only turns =< degrees in either
direction for a total of 1F< degree movement. 'he motorAs neutral position is defined as the
position where the servo has the same amount of potential rotation in the both the clocwise
or counter1clocwise direction. 'he 2>* sent to the motor determines position of the shaft,
and based on the duration of the pulse sent via the control wire; the rotor will turn to the
desired position. 'he servo motor expects to see a pulse every .< milliseconds #ms$ and the
length of the pulse will determine how far the motor turns. ,or example, a 1.@ms pulse will
,%)8!/'%)0(!
/!N/%)/
98+
+5/29A?
/!)G% *%'%)
#*3$
2%>!)
/(229?
11
mae the motor turn to the =<1degree position. /horter than 1.@ms moves it to < degrees and
any longer than 1.@ms will turn the servo to 1F< degrees, as diagramed below.
'he 98+ display used in this project is an1B/. alphanumeric display module 'he
-4+1B.A dot1matrix li&uid crystal display controller and driver 9/5 displays alphanumeric,
-apanese ana characters, and symbols. 5t can be configured to drive a dot1matrix li&uid
crystal display under the control of a 71 or F1bit microprocessor. /ince all the functions such
as display )A*, character generator, and li&uid crystal driver, re&uired for driving a dot1
matrix li&uid crystal display are internally provided on one chip, a minimal system can be
interfaced with this controller/driver. 'he advantages using this display are low power
operation.
A eypad is used for giving input to the system. 5t consists of buttons arranged in
matrix order. A mode selection switch is used for selecting the modes #passive mode and
active mode$. 5n the passive mode of operation the machine arm drives the patient arm and
in the active mode of operation the arm drives the machine arm. 'he default of the mode is
in passive. 'he passive stretching is of two types. ,irst one is individual joints stretching
and the second one is the multiple joint stretching.
,orce1sensing resistors consist of a conductive polymer, which changes resistance in
a predictable manner following application of force to its surface. 'hey are normally supplied
as a polymer sheet or in that can be applied by screen printing. 'he sensing film consists of
both electrically conducting and non1conducting particles suspended in matrix. 'he particles
are sub1micrometer sizes, and are formulated to reduce the temperature dependence, improve
mechanical properties and increase surface durability. Applying a force to the surface of the
sensing film causes particles to touch the conducting electrodes, changing the resistance of
the film. As with all resistive based sensors, force1sensing resistors re&uire a relatively simple
interface and can operate satisfactorily in moderately hostile environments. 8ompared to
other force sensors, the advantages of ,/)s are their size #thicness typically less than
<.@ mm$, low cost and good shoc resistance. 4owever, ,/)s will be damaged if pressure is
applied for a longer time period #hours$. A disadvantage is their low precision: measurement
results may differ 1<H and more.
1.
'he patient is entered into the machine; his /her hand is strapped to the
corresponding braces of the machine arm. 'hen the switch is turned on and it re&uires the
inputs for operation. 'he operator enters the stroe values # elbow ,l1!x, wrist ,l1!x and
forearm 2r1/u$ to the system through the eypad. 'he inputs are entered then machine
starts to run in the prescribed time period or the number of cycles. 'his is the passive
mode of operation, it reduces the joint /muscle stiffness. >hen the passive mode is
completed the operator switches the machine in to active mode. 5n this mode the patient
tries to move the arm in the prescribed direction. 5t improves the muscle strength of the
arm and this movement is sensed by the force /tor&ue sensor. 'his information is sent to
the controller, and processed it .'he operation is continuously monitored by the controller
in real time and if either of them is out of its range, the whole system is then shut down. A
stop switch is given both to the operator and the patient to authorize them to shut down the
system at any time.
2.2.2 S!*.o )oto*
/ervo motors have been around for a long time and are utilized in many applications.
'hey are small in size but pac a big punch and are very energy1efficient. 6ecause of these
features, they can be used to operate remote1controlled or radio1controlled toy cars, robots
and airplanes /ervo motors are also used in industrial applications, robotics, in1line
manufacturing, pharmaceutics and food services.
'he servo circuitry is built right inside the motor unit and has a positionable shaft,
which usually is fitted with a gear #as shown below$. 'he motor is controlled with an electric
signal which determines the amount of movement of the shaft.
13

Fig 2.2 /ervomotor
/ervos are controlled by sending an electrical pulse of variable width, or 2ulse >idth
*odulation #2>*$, through the control wire. 'here is a minimum pulse, a maximum pulse,
and a repetition rate. A servo motor can usually only turns =< degrees in either direction for a
total of 1F< degree movement. 'he motorAs neutral position is defined as the position where
the servo has the same amount of potential rotation in the both the clocwise or counter1
clocwise direction. 'he 2>* sent to the motor determines position of the shaft, and based
on the duration of the pulse sent via the control wire; the rotor will turn to the desired
position. 'he servo motor expects to see a pulse every .< milliseconds #ms$ and the length of
the pulse will determine how far the motor turns. ,or example, a 1.@ms pulse will mae the
motor turn to the =<1degree position. /horter than 1.@ms moves it to < degrees and any longer
than 1.@ms will turn the servo to 1F< degrees, as diagramed below.
Fig 2.6 2ulse >idth >ave ,orm
17
>hen these servos are commanded to move, they will move to the position and hold
that position. 5f an external force pushes against the servo while the servo is holding a
position, the servo will resist from moving out of that position. 'he maximum amount of
force the servo can exert is called the tor&ue rating of the servo. /ervos will not hold their
position forever though; the position pulse must be repeated to instruct the servo to stay in
position.
.2.6 Li:ui C*,st%' Dis+'%, ;LCD<
Fig 2.7 /tructure of 98+ +isplay
'he main principle behind li&uid crystal molecules is that when an electric current is
applied to them, they tend to untwist. 'his causes a change in the light angle passing through
them. 'his causes a change in the angle of the top polarizing filter with respect to it. /o little
1@
light is allowed to pass through that particular area of 98+. 'hus that area becomes darer
comparing to others. ,or maing an 98+ screen, a reflective mirror has to be setup in the
bac. An electrode plane made of indium1tin oxide is ept on top and a glass with a
polarizing film is also added on the bottom side. 'he entire area of the 98+ has to be
covered by a common electrode and above it should be the li&uid crystal substance. Next
comes another piece of glass with an electrode in the shape of the rectangle on the bottom
and, on top, another polarizing film. 5t must be noted that both of them are ept at right
angles. >hen there is no current, the light passes through the front of the 98+ it will be
reflected by the mirror and bounced bac. As the electrode is connected to a temporary
battery the current from it will cause the li&uid crystals between the common1plane electrode
and the electrode shaped lie a rectangle to untwist. 'hus the light is bloced from passing
through. 'hus that particular rectangular area appears blan. 'he surfaces of the electrodes
that are in contact with the li&uid crystal material are treated so as to align the li&uid crystal
molecules in a particular direction. 'his treatment typically consists of a thin polymer layer
that is unidirectional rubbed using, for example, a cloth. 'he direction of the li&uid crystal
alignment is then defined by the direction of rubbing. !lectrodes are made of a transparent
conductor called 5ndium 'in %xide #5'%$.
2.2.6.1 (HD1=2A- 1= 9 2 A'+/%"u)!*i$ LCD &ou'!
'he -4+1B.A dot1matrix li&uid crystal display controller and driver 9/5 displays
alphanumeric, -apanese ana characters, and symbols. 5t can be configured to drive a dot1
matrix li&uid crystal display under the control of a 71 or F1bit microprocessor. /ince all the
functions such as display )A*, character generator, and li&uid crystal driver, re&uired for
driving a dot1matrix li&uid crystal display are internally provided on one chip, a minimal
system can be interfaced with this controller/driver.
'he -4+1B.A character generator )%* is extended to generate .<F @IF dot
character fonts and 3.@I1< dot character fonts for a total of .7< different character fonts.
'he low power supply #..CG to @.@G$ of the -4+1B.A is suitable for any portable battery1
driven product re&uiring low power dissipation.
1B

Fig 2.= .19ine by 1B18haracter +isplay
2.2.7 Fo*$!>To*:u! S!"so*
,orce1sensing resistors consist of a conductive polymer, which changes resistance in
a predictable manner following application of force to its surface. 'hey are normally supplied
as a polymer sheet or in that can be applied by screen printing. 'he sensing film consists of
both electrically conducting and non1conducting particles suspended in matrix. 'he particles
are sub1micrometer sizes, and are formulated to reduce the temperature dependence, improve
mechanical properties and increase surface durability. Applying a force to the surface of the
sensing film causes particles to touch the conducting electrodes, changing the resistance of
the film. As with all resistive based sensors, force1sensing resistors re&uire a relatively simple
interface and can operate satisfactorily in moderately hostile environments. 8ompared to
other force sensors, the advantages of ,/)s are their size #thicness typically less than
<.@ mm$, low cost and good shoc resistance. 4owever, ,/)s will be damaged if pressure is
applied for a longer time period #hours$. A disadvantage is their low precision: measurement
results may differ 1<H and more.
1C
'or&ue is a twisting force, usually encountered on shafts, bars, pulleys, and similar
rotational devices. 5t is defined as the product of the force and the radius over which it acts. 5t
is expressed in units of weight, times, length, such as lb.1ft. and N1m. Another way to
measure tor&ue is by way of twist angle measurement or phase shift measurement, whereby
the angle of twist resulting from applied tor&ue is measured by using two angular position
sensors and measuring the phase angle between them. ,inally, if the mechanical system
involves a right angle gearbox, then the axial reaction force experienced by the inputting
shaft/pinion can be related to the tor&ue experienced by the output shaft. 'he axial input
stress must first be calibrated against the output tor&ue. 'he input stress can be easily
measured via strain gage measurement of the input pinion bearing housing. 'he output tor&ue
is easily measured using a static tor&ue meter.
2.2.? &o! S!'!$tio"
'here are two operating modes used in the system,
2assive stretching #passive mode$
Active training #active mode$
5n the passive mode of operation, machine arm drives the impairment arm in the
prescribed level of the input.
5n the active mode of operation, patient arm drives the machine arm .5n this mode the
patient arm tries to move #or apply forces to $the machine arm, and it is sensed by using some
force/tor&ue sensors .'his is the training mode of operation.
,or mode selection a switch is used. 5t has two level ,high and low, high indicate in
the active mode and low indicate the passive mode. 'he default position of the switch is in
the passive mode. 'he operator changes the switch in passive mode or active mode.
1F
2.2 BLOC8 DIAGRA& OF THE &ODIFIED E9OS8ELTON S3STE&
%N/%,,
/>5'84
258 1B,FCCA
*%+!
/!9!8'5%N
),5+
)!A+!)
J6!!
(N5'
G%58!
8%N')%9
/!)G% *%'%)
#*1$
/!)G% *%'%)
#*.$
/!)G% *%'%)
#*3$
,%)8)/'%)0(!
/!N/%)/
1=
Fig 2.? 6loc +iagram of the (ser 8ontrolled !xoseleton /ystem
Fig 2.@ 28 with Application
2.2.1 O+!*%tio" o# t/! Us!* Co"t*o''! E4os1!'!to" S,st!)
'he bloc diagram of proposed system is shown in fig 3.@. 'he proposed system
consists of a voice control unit, Jbee units and ),5+ that not present in the existing system.
As mentioned in the chapter 1, the present system needs an operator to provide the inputs to
the system and the user have no control in the system. 'hese two drawbacs are overcome in
the proposed system.
'o eliminate the need of an operator it provides a user id to every patient. >hen a
user login to the machine, a signal corresponding to this id is sent to the database of the
hospital or clinic through the Jbee unit. 'he database contains all the information about the
patients #pre1evaluated values$ .5f it is a valid id, and then informationDs corresponding to that
id is sent bac to the machine. 'hese values are loaded into the controller and machine starts
to run. 'he proposed system also provides a voice control to the user to set the number of
cycles of operation. 'he mode selection switch is used to select the passive mode or active
98+
+5/29A?
2%>!)
/(229?

J6!!
(N5'

*AJ 3.
28
.<
mode. 'he passive stretching reduces joint/muscle stiffness and active training improves the
muscle strength.
2.2.2 9B!!
Fig 2.A Jbee module
5t is a wireless communication standard designed for point1to1point and star
communications at over1the1air baud rates of .@< bit/s. J6ee *odules are available in two
form1factors; 'hrough14ole and /urface *ount. All J6ees #with the exception of the J6ee
FBF92$ are available in the popular .<1pin 'hrough14ole form1factor. 8ertain J6ee modules
are also available in a 3C1pad /urface *ount design, which is popular for higher volume
applications due to the reduced manufacturing costs of /*' technology. J6ee *odules
typically come with several antenna options, including (.,9, 286 !mbedded, >ire, and
)2/*A.
.1
'he J6ees can operate either in a transparent data mode or in a pacet1based
application programming interface #A25$ mode. 5n the transparent mode, data coming into
the +ata 5N #+5N$ pin is directly transmitted over1the1air to the intended receiving radios
without any modification. 5ncoming pacets can either be directly addressed to one target
#point1to1point$ or broadcast to multiple targets #star$. 'his mode is primarily used in
instances where an existing protocol cannot tolerate changes to the data format. A'
commands are used to control the radioDs settings. 5n A25 mode the data is wrapped in a
pacet structure that allows for addressing, parameter setting and pacet delivery feedbac,
including remote sensing and control of digital 5/% and analog input pins.
2.2.2 RFID &ou'!
Fig 2.1B ),5+ *odule
)adio1fre&uency identification #),5+$ is the wireless non1contact use of radio1
fre&uency electromagnetic fields to transfer data, for the purposes of automatically
identifying and tracing tags attached to objects. 'he tags contain electronically stored
information. /ome tags are powered by and read at short ranges #a few meters$ via magnetic
..
fields #electromagnetic induction$, and then act as a passive transponder to emit microwaves
or (4, radio waves #i.e., electromagnetic radiation at high fre&uencies$. %thers use a local
power source such as a battery, and may operate at hundreds of meters. (nlie a bar code, the
tag does not necessarily need to be within line of sight of the reader, and may be embedded in
the traced object.
A radio1fre&uency identification system uses tags, or labels attached to the objects to be
identified. 'wo1way radio transmitter1receivers called interrogators or readers send a signal
to the tag and read its response.
),5+ tags can be passive, active or battery1assisted passive. An active tag has an on1board
battery and periodically transmits its 5+ signal. A battery1assisted passive #6A2$ has a small
battery on board and is activated when in the presence of an ),5+ reader. A passive tag is
cheaper and smaller because it has no battery. 4owever, to start operation of passive tags,
they must be illuminated with a power level roughly three magnitudes stronger than for
signal transmission. 'hat maes a difference in interference and in exposure to radiation.
'ags may either be read1only, having a factory1assigned serial number that is used as a ey
into a database, or may be read/write, where object1specific data can be written into the tag
by the system user. ,ield programmable tags may be writing1once, read1multiple; KblanK
tags may be written with an electronic product code by the user. A tag with no inherent
identity is always threatened to get manipulated.
),5+ tags contain at least two parts: an integrated circuit for storing and processing
information, modulating and demodulating a radio1fre&uency #),$ signal, collecting +8
power from the incident reader signal, and other specialized functions; and an antenna for
receiving and transmitting the signal. 'he tag information is stored in a non1volatile memory.
'he ),5+ tag includes either a chip1wired logic or a programmed or programmable data
processor for processing the transmission and sensor data, respectively.
.3
CHAPTER 6
SOFTWARE C HARDWARE DETAILS
6.1 HI-TECH C
'he 451'!84 8 8ompiler for 2581</1./1B *8(s is a free1standing, optimizing
AN/5 8 compiler. 5t supports all 2581<, 2581. and 2581B series devices, as well as the
25817<<< device and the enhanced *id1)ange 258L *8( architecture.
'he command1line driver is called 2588M and is the application that can be invoed
to perform all aspects of compilation, including 8 code generation, assembly and lin steps.
!ven if we use an 5+! to assist with compilation, the 5+! will ultimately call 2588.
Although the compiler applications can be called explicitly from the command line,
using 2588 is the recommended way to use the compiler as it hides the complexity of all the
internal applications used and provides a consistent interface for all compilation steps. 2588
distinguishes source files, intermediate files and library files solely by the file1type, or
extension. 451'!84 8 8ompiler for 2581</1./1B *8(s supports a number of special
features and extensions to the 8 language which are designed to ease the tas of producing
)%*1based applications.
'he macro assembler included with 451'!84 8 2)% for 2581</1./1B *8( ,amily
assembles source files for 258 1</1./17/1B/1C *8(s. 5t describes the usage of the assembler
and the directives #assembler pseudo1ops and controls$ accepted by the assembler in the
source files. Although the term NassemblerO is almost universally used to describe the tool
.7
which converts human1readable mnemonics into machine code, both NassemblerO and
NassemblyO are used to describe the source code which such a tool reads.
'he application name of the liner is 495NE. 5n most instances it will not be
necessary to invoe the liner directly, as the compiler driver, 2588, will automatically
execute the liner with all necessary arguments. (sing the liner directly is not simple, and
should be attempted only by those with a sound nowledge of the compiler and lining in
general. 'he compiler often maes assumptions about the way in which the program will be
lined. 5f the aspects are not lined correctly, code failure may result. 5f it is absolutely
necessary to use the liner directly, the best way to start is to copy the liner arguments
constructed by the compiler driver, and modify them as appropriate. 'his will ensure that the
necessary startup module and arguments are present.
6.2 PIC 8IT 2 DEVELOP&ENT PROGRA&&ER>DEBUGGER
'he 258 it . +evelopment 2rogrammer/+ebugger is a low1cost development
programmer. 5t is capable of programming most of *icrochip. ,lash microcontrollers and
serial !!2)%* devices. 'he 258 it . also may be used to debug selected devices 'he 258
it . +evelopment 2rogrammer/+ebugger can program microcontroller devices that are
installed in an application circuit using 5n18ircuit /erial 2rogramming #58/2$. 58/2 re&uires
five signals:
G22. 2rogramming Goltage; when applied, the device goes into 2rogramming
mode.
58/289E or 2P8. 2rogramming 8loc; a unidirectional synchronous serial
cloc line from the programmer to the target.
58/2+A' or 2P+. 2rogramming +ata; a bidirectional synchronous serial data
line.
G++. 2ower /upply positive voltage.
G//. 2ower /upply ground reference.
.@
'he application circuit must be designed to allow all the programming signals to be
connected to the device without distorting the programming signals. 5n addition to the
258it. 2rogrammer application, the 258 it . +evelopment 2rogrammer/+ebugger may be
used with *29A6L 5+!, the free integrated development environment. *29A6 5+! allows
the 258 it . to be used as an in1circuit debugger as well as a programmer #for selected
devices only$. 5n1circuit debugging allows us to run, examine and modify our program while
the device is embedded in the target hardware.
CHAPTER 7
RESULT
7.1 OUTPUT
'he experiment had conducted in two hands #A)* A and A)* 6$. 'he reference used
for each arm is 3< deg. 'he weight of A)* A is less than A)* 6. 'he experimental result
shows in the table. >e can see that A)* A has more degrees of rotation compared to A)*
6.

T%b'! 7.1 !xperimental )esult
'he evaluation capabilities aiding diagnosis can provide valuable information on
which joints and which +%,s have significant changes and which joints lose independent
.B
control and what are the abnormal couplings. 'herefore the clinicians had much awareness
on the type, intensity and duration of therapy of each patient.
CHAPTER =
CONCLUSION
'he user controlled exoseleton robot system overcomes the drawbacs of the
existing robot system. /uch as it avoid the need of an operator and provide a user control. 5t
aiming to support clinicians and patients in all four steps of neurorehabilitation with the
following novel integrated capabilities: &uantitative, objective and comprehensive *-*+
pre1evaluation capabilities aiding diagnosis; strenuous and safe passive stretching of
deformed arm; active training after passive stretching for improving the control ability; a
comprehensive outcome evaluation at the level of individual joints/+%,s and multiple
joints/+%,s.
'he evaluation capabilities aiding diagnosis can provide valuable information on
which joints and which +%,s have significant changes and which joints lose independent
control and what are the abnormal couplings. 'hus, the clinicians can mae more informed
decision on the type, intensity and duration of therapy of each patient. 'his can be realizing
with the same robot by utilizing its passive stretching and active movement training
capabilities. 5t also improves peopleDs health because the body is meant to wor in an upright
posture and helps people psychologically.
.C
REFERENCES
Q1R 4yung1/oon 2ar, 9i10un Shang, /ang 4oon Eang, ?i1Ning >u and ?upeng )enO
+eveloping a *ulti1-oint (pper 9imb !xoseleton )obot for +iagnosis, 'herapy, and
%utcome !valuation in Neurorehabilitation,O 5!!! 'ransactions %n Neural /ystems
And )ehabilitation !ngineering, Gol. .1, No. 3, *ay .<13.
Q.R 6ai S, 8hung /. P, -ohnson *. !, )ogers *.>, )oth !. -, Shang 9.10 and Gan)ey,
N5ntelligent stretching for anle joints with contracture/spasticity,O 5!!! 'ransaction
Neural /ystem )ehabil. !ng., vol. 1<, no. 3, pp. 17="1@C, /ep. .<<..
Q3R 6uchanan '. /, +ewald -. 2. A, Piven -. +, 2ope 2. / and )ymer >.S, NAbnormal
muscle co1activation patterns during isometric tor&ue generation at the elbow and
shoulder in hemi paretic subjects,O 6rain, vol. 11F, pp. 7=@"@1<, 1==@.
Q7R 6urgar 8.P, /hor 2.8, Gan der 9oos 4.,.*,O+evelopment of robots for rehabilitation
therapy: 'he 2alo Alto GA//tanford experience,O-.)ehabil. )es.+evelop. vol.3C,
ppBB31BC3, .<<<.
Q@R 8hang 2.4, -in * and Eang /.4,OA solution to the accuracy/robustness dilemma
impedance control,O5!!!/A/*! 'rans.*echatron, vol.17, no.3, pp..F.1.=7,
-un..<<=.
QBR 8hang 2.4, 4ogan N, Eang /.4, Erebs 4.5 and 2ar E, N/tochastic estimation of
human arm impedance under nonlinear friction: An experimental validation,O
5!!!/A/*! 'rans.*ecatron., vol.1F, no.., ppCC@1CFB, Apr..<13.
QCR 8hang 2.4 and Eang /.4, N/tochastic estimation of human arm impedance under
nonlinear friction in robot joints: A model study,O- Neurosci.*ethods, vol.1F=, pp.=C1
11., .<1<.
QFR Puidali, Nef .' and )iener ),NA)* in 5551Arm therapyexoselton with an ergonomic
shoulder actuation,O Application 6ionics.6iomech., volB, pp.1.C117., .<<=.
.F
Q=R Eang /.4 and Shang 9.0, N)obust identification of multi1joint human arm impedance
based on dynamics decomposition: A *%+!95NP /'(+?,O 5n 2roc.33
rd
Annu.5nt.8onf.5!!! !ng.*ed.6iol./oc., 6oston, *A, .<11, pp.77@3177@B.
Q1<R /humway18oo A and >oollacott *. 4,N *otor 8ontrol: 'heory and 2ractical
Applications, O .nd ed. 2hiladelphia, 2A: 9ippincott >illiams >ilins, .<<1, 8h. B.