th
2011 4 International Conference on Mechatronics (ICOM), 17-19 May 2011, Kuala Lumpur, Malaysia
Design and Development of a Hand-glove Controlled
Wheel Chair
Rini Akmeliawati, Faez S. Ba Tis, Umar J. Wani.
Department of Mechatronics Engineering, Kuliyyah of Engineering.
International Islamic University Malaysia.
Kuala Lumpur, Malaysia.
Abstract— Wheelchairs are a way of reincarnating the purpose of
life in the lives of disabled people. Effective and efficient ways of
delivering a cost-effective and affordable wheelchair to the
common masses, which is not only at par with the present day
technology, but is much easier to use are presented herewith.
Replacement of the popular joystick stick controlled wheel chair
with a hand-glove control system for easier maneuvering by
bending the fingers, is discussed in this paper. Intended users
control the system by wearing an instrumented glove fitted with
flex or bend sensors for controlling the movement and direction
of the wheelchair. Uni-directional wireless communication exists
between the instrumented gloves and the controller which is
sandwiched between the user’s seat and the wheels. Initial design
results are also presented in this paper. The technologies
presented in this paper suggest a wide domain of possibilities to a
wide variety of users. In addition, it also aims at making a cost-
effective chair so that more hi-tech wheelchairs are made use of,
widely, by people with disabilities.
I. INTRODUCTION
Over 100 million people worldwide, with physical
disabilities require the assistance of a wheelchair but only a
small percentage of them actually own or have the liberty of
buying one. Although manual wheelchairs have proven to be
beneficial for the disabled but it has only served the purpose of
people with minor disabilities. Consider for a moment, a
person suffering from partial tetraplegia or sclerosis. It
becomes extremely difficult for such a person to maneuver the
chair, even within the confines of a house. Such a person would
require the assistance of other people, even to carry out his
daily chores. Another alternative would be an electric
wheelchair controlled by a joystick. Although the electric
powered wheelchair is a much improvised vehicle and an
easier-to-control device, but it might not help the cause of
severely disabled. Common experience shows that a joystick
requires a relatively large force which is more than the
threshold for severely disabled people. Taking into
consideration, the force and the whopping cost of wheelchairs
available in the market, this paper describes the design of a
cost-effective and easier-to-control wheelchair. The paper
presents a control-method to maneuver a motorized wheelchair
merely by the movement of fingers. It aims at incorporating the
modern ways of wheel chair dynamics and control and at the
same time making it cost effective, so that it is affordable to the
common masses.
978-1-61284-437-4/11/$26.00 ©2011 IEEE
The goal of this research is to develop a wheelchair system
which controls its movement by the mere bending of a person’s
fingers. Special type of sensors known as ‘flex-sensors’ are
embedded into a hand glove in order to achieve the desired
goal. In this research a prototype of an affordable and
technologically advanced wheelchair is to be designed and
developed. This is to aid the communication of severely
disabled people and enhance the maneuvering of the vehicle
with the use of hand movements. The proposed prototype will
be communicating wirelessly between the controller and the
plant and it will also replace the traditional joystick by the
implementation of user hand glove control.
The remaining part of this paper is organized as
follows. Literature review is presented in Section II while the
proposed methodology is elaborated in Section III. Initial
design results obtained are contained in section IV and V,
followed by the conclusion, which ends this paper.
II. LITERATURE REVIEW
The technological progress in the field of Wheelchair
technology is at its epoch. It’s not surprising to see that a lot of
work has been done in the field of wheelchair technology. First
there were the manual wheelchairs’, then came the electric
powered wheelchairs and now there are Smart Wheelchairs.
The design of a ‘Motorized Chair’ that transports a human
within the confines of a house is reported by Desai and Endrele
[1]. A ‘Motorized Chair’ consists of a chair with the two
motors and a joystick controller. A microcontroller outputs the
speed and direction to the motors. Most wheelchairs are
controlled from their back wheels. The front wheels respond
and rotate in accordance to the back wheels. As the controller is
moved, the micro-controller senses the movement, does the
calculations for power and direction and the motors move
accordingly. A motor-control circuit controls the speed and the
direction of rotation of motors.
H. Kazerooni et al.[2] report on the development of a
device called ‘The Magic Glove’. The ‘Magic Glove’ is an
instrumented glove with wide range of applications in the field
of robotics. The glove measures the force exerted by the wearer
of a material handling robot. Once the force is sensed and
measured, it is broadcasted to the controller. The function of
the magic glove is to allow the user to apply a minimal force on
an object while transporting it from one place to another. The
human or the wearer of the gloves just mimics the action to be
done, while the actual lifting and placing is left for the actuator
to be done. Radio Frequency (RF) technology is used to transit
the signals from the glove to the microcontroller. Zimmerman
et al. [3] also report on the development of a similar hand to
machine interface device. The device essentially is a glove that
provides real-time gesture, position and orientation
information.
Various methods for the advancement of wheelchair
technology have been discussed and analyzed herewith. It is
apparent that the control of a wheelchair can be further
improved by the combination of different technologies such as
the incorporation of an instrumented glove, hence replacing the
need for joysticks and creating an independent controller,
which is easier-to-control and portable at the same time, for the
end user.
III. METHODOLODY AND WORKING
The methodology consists of a thorough study and analysis
of electric powered and joy-stick controlled wheelchairs, and
the control law used to maneuver these vehicles. Also,
incorporated in the methodology is a study of the working of
the bend sensors popularly known as flex-sensors. Figures 1
and 2 show the block diagrams representing the complete
system. It consists of a flex-sensor circuit, a wireless
transmitter and receiver combo, microcontrollers, a motor-
driver circuit and motors.
Figure 1. Block Diagram (On the Hand Glove)
The arrangement consists of two sub-systems. One is the
hand-glove and the other the lies beneath the seating base of
the chair.
A. On the Hand-glove
The bending of the flex-sensors attached to the hand-glove
initiates the process. As the fingers are bent and hence the flex
sensors, a voltage is developed across the constructed flex
sensor circuit. This voltage is fed into the Arduino Funnel I/O
Board which comes with an inbuilt X-Bee wireless transmitter-
socket and a microcontroller. The microcontroller does the
necessary calculations (works as an analog to digital converter
and converts the data into serial data) and sends the signals
serially through the X-Bee transmitter, using the Zigbee
wireless communication protocol.
The glove further consists of two components, a bend
sensing circuit and a transmitting circuit. The bend sensing
circuit is essentially a simple voltage divider circuit. A resistor
is connected in parallel with the flex sensor, across the supply
voltage. When the user bends his finger, the resistance of the
sensor increases. The more the bend, the more the resistance
and hence the voltage across the sensor varies. This varying
voltage is fed into the microcontroller and is given by,
ୖϐౢ౛౮
˜ൌ ୗ୳୮୮୪୷ (1)
ୖϐౢ౛౮ ାୖ
Wheel
Figure 2. Block Diagram (Wheel Chair Side)
‫ ݁ݎ݄݁ݓ‬ϐ୪ୣ୶ ൌ ܴ݁‫ݎ݋ݏ݊݁ݏݔ݈݁ܨ݄݁ݐ݂݋݁ܿ݊ܽݐݏ݅ݏ‬
ܸௌ௨௣௣௟௬ ൌ ܵ‫݁݃ܽݐ݈݋ܸݕ݈݌݌ݑ‬
 ൌ ܴ݁‫ݎ݋ݏ݊݁ݏݔ݈݁ܨ݄݁ݐ݋ݐ݈݈݈݁ܽݎܽ݌݊݅݀݁ݐܿ݁݊݊݋ܿ݁ܿ݊ܽݐݏ݅ݏ‬
B. Underneath the Wheel Chair  ൌ ™ ൌ ʹ͸ሺͳǤ͹Ͳͷሻ ؆ ͶͲ (5)
On the other end, beneath the wheelchair, another circuit is
constructed that consists of an X-Bee receiver and its shield. ‫ ܲ݁ݎ݄݁ݓ‬ൌ ܴ݁‫ݎ݁ݓ݋ܲ݀݁ݎ݅ݑݍ‬
The shield acts as interface platform between X-Bee receiver ஺ ൌ ‫ݐ݄݃݅݁ݓ݊ܽ݉ݑ݄݋ݐ݁ݑ݀݁ܿݎ݋ܨ‬
and Arduino ATMEGA328P board. The receiver receives the ஺ ൌ ܴ݁ܽܿ‫݈݄݁݁ݓݎܽ݁ݎ݄݁ݐ݊݋݁ܿݎ݋݂݊݋݅ݐ‬
signals sent by the transmitter and passes them onto the ‫ ݎ‬ൌ ܴܽ݀݅‫݈݄݁݁ݓݎܽ݁ݎ݄݁ݐ݂݋ݏݑ‬
Arduino ATMEGA328P board which has another built-in ݉௪ ൌ ‫݈݄݁݁ݓ݄݁ݐ݂݋ݏݏܽܯ‬
microcontroller. The microcontroller evaluates and analyses the ௥ ൌ ‫݁ܿݎ݋݂݈ܽ݊݋݅ݐܿ݅ݎܨ‬
signals, based on the instruction codes which have been ܶ ൌ ܶ‫݁ݑݍݎ݋‬
programmed into it and accordingly sends the signals to its
PWM output and the H-bridge circuit or the motor driver. The
function of the motor driver circuit is to control the direction of
rotation of the motors. Through the H-bridge the motors Rear
receive the instructions and act consequently, having complete wheel The Frame
control over the speed and direction of rotation of the motors.
Hence, the wheelchair begins to move according to the
movement or bending of the user’s fingers.
Front
Once the voltage is received by the microcontroller, it A Wheel
needs to be transmitted over to the other side of the system, B
which is the wheelchair, underneath which lies the main
controller board, the Arduino ATMEGA328P, along with the
motor-drivers and other circuitry. This is done by the Rear Wheel ܹீ
transmitting circuit present on the hand-glove, hence realizing
wireless communication between the chair and the glove. The
A
wireless transfer protocol used is Zigbee, since it is easier to
use and the wheelchair doesn’t require a long-range
communication. A
B
The most significant component in this design is the flex
sensor, because it requires negligible force to bend the sensor,
which in turn causes the motion of the wheelchair. This small Figure 3. Two-dimensional Side view of the wheel chair.
force overcomes the limitation provided by the traditional
joystick control of wheelchairs making it easier for people with
severe disabilities to maneuver the vehicle. The flex sensors are B. Design of Belt
embedded into the instrumented gloves, with each finger
having a different function. The bending of the index finger
causes the chair to move forward. As the finger is bent more,
the voltage across the divider-circuit increases and so does the Shaft
speed of the vehicle. The rest of the fingers are used to move
the chair right, left and backwards. There is a tolerance
Motor ࡲ૚
provided for the abnormal bending because of the tendency of
the fingers to follow the one which is bent, although vaguely. If
all four fingers are bent at a time, the microcontroller is
programmed such that it will halt all the operations and the
chair will come to a standstill.
IV. INITIAL DESIGN CALCULATIONS

A. Dynamic Analysis of the Rear Wheel ࡲ૛

Figure 3 shows the side view of a typical wheel chair
structure with a load acting on it. It is used to derive and
calculate the necessary specifications for power and torque. Rare
Pulley 1(attached to the
Wheel
gear head)
஺ ൅  ݉௪ ݃ ൌ ஺ (2)
Pulley (w) (attached to
௥ ൌ  Ɋ௞ ‫ כ‬஺ ൌ ͺ͹ܰ (3) the shaft)

 െ  ୰ ‫ ” כ‬ൌ Ƚ (4)
Figure 4. Side view of the belt-shaft assembly.
Diameter for Pulley 1, ‫ܥ݁ݎ݄݁ݓ‬ҧ ൌ ‫ܽݎ݀ܽ݋݈ܿ݅݉ܽ݊ݕܦ‬ ܽ‫݃݊݅ݎܾ݄ܽ݁݁ݐ݂݋݃݊݅ݐ‬
‫ܮ‬෨ ൌ ‫݃݊݅ݎܾ݂ܽ݁݋݂݁݅ܮ‬ሺ݈݈݅݊݉݅݅‫ݏ݈݁ܿݕܿݏݏ݁ݎݐݏ݊݋‬ሻ
†ଵ ൌ
ୢ౭
ൌ
଺
ൌ ʹǤͷ‹ (6) ܲ ൌ ‫ܽ݋݈݃݊݅ݎܾܽ݁ݐ݈݊݁ܽݒ݅ݑݍܧ‬ܽ݀
୥౨ ଶǤହ
Based on the preceding calculationss, the designation for
With the above values, the angle used to descrribe arc contact bearings A and B is ͸ͲͲͶǤ
and the belt length can be expressed as,
V. RESU
ULTS
ୢ ିୢ
Ƚൌ •‹ିଵ ౭ భ  ൌ ͺǤ͵ͺ ι
(7) Figure 5 shows the plot of the resistance
r of the flex sensor
ଶେౚ
and the bending of fingers. As the fingers are bent, the
resistance of the sensors increases linearly
l causing an increase
‫ ׎‬ൌ  ͳͺͲι െ ʹሺȽሻ ൌ ͳͺͲι െ  ͳ͸Ǥ͹͸ι ൌ ͳ͸͵Ǥʹ
ʹͶι  (8)
in the speed of the vehicle. The plot
p of velocity against the
஠ ஠ሺୢ౭ ିୢభ ሻ஑ bending angle is shown in Figure 6, for the no-load condition.
 ൌ  ඥሺሺʹୢ ሻଶ െ  ሺ†୵ െ †ଵ ሻଶ ൅ ሺ†୵ ൅ †ଵ ሻ ൅ 
ଶ ଵ଼଴

 ൌ ͵͹Ǥ͸Ͳ‹ (9) R
e
The centrifugal force acting on the belt is giveen by, s
i
s
୵౰ ୳మ
ୡ ൌ  ሺ ሻ ൌ ͲǤͲͲͲͳͻŽ„ (10) t
୐ ୥ a
n
ρಘ‫׎‬
୊భ ି୊ౙ c
ൌ  Ԗభఴబι‫כ‬౩౟౤ಊ (11) e
୊మ ି୊ౙ

(K
ଵ ൌ ͺͲǤ͵Ž„ǡ ଶ ൌ ͷŽ„ (12) -
O
‫ ܮ݁ݎ݄݁ݓ‬ൌ ‫ ݐ݈ܾ݄݁݁ݐ݂݋݄ݐ݃݊݁ܮ‬ h
m
‫ܨ‬௖ ൌ ‫ݐ݈ܾ݄݁݁ݐ݊݋݃݊݅ݐܿܽ݁ܿݎ݋݂݈ܽ݃ݑ݂݅ݎݐ݊݁ܥ‬ s
‫ ׎‬ൌ ܹ‫ ݈݁݃݊ܽ݌ܽݎ‬ )
‫ ݑ‬ൌ ‫ ݕݐ݅ܿ݋݈݁ݒݐ݈݁ܤ‬
݀ଵ ൌ ‫ͳݕ݈݈݁ݑ݂ܲ݋ݎ݁ݐ݁݉ܽ݅ܦ‬
݀௪ ൌ ‫ݕ݈݈݁ݑ݂ܲ݋݁ݎ݁ݐ݉ܽ݅ܦ‬ሺ‫ݓ‬ሻ
‫ܥ‬ௗ ൌ ‫ ݏݕ݈݈݁ݑ݌ ݄݁ݐ݊݁݁ݓݐܾ݁݁ܿ݊ܽݐݏ݅݀ݎݐ݁݊݁ܥ‬ Bendiing Angle
Figure 5. Plot of Sensor Resistan
nce (vs) Bending Angle
C. Shaft Design

஠ୢయ ୗ౯
Ʉୱ ൌ  య
(13) V
ଷଶට୑ାቀ ቁ୘మ e
ర
l
o
య మ ାቀయቁ‫כ‬ଶ଺మ
c
ସ‫כ‬ଷଶට଼Ǥ଻ସ
† ൌ ඨ
ర i
 ൌ ͲǤͲͳͻ ؆ ʹͲ
 (14) t
஠‫כ‬ଶଽହ୑୔ୟ
y
‫ߟ݁ݎ݄݁ݓ‬௦ ൌ ݂ܵܽ݁‫ݎ݋ݐ݂ܿܽݕݐ‬ (
݀ ൌ ‫ݐݐ݂݄ܽݏ݄݁ݐ݂݋ݎ݁ݐ݁݉ܽ݅ܦ‬ m
‫ ܯ‬ൌ ‫ݐ݊݁݉݋ܯ݃݊݅݀݊݁ܤ‬ /
ܶ ൌ ܶ‫݊݋݅ݏݎ݋‬ s
)

D. Design of Bearing
For Bearings A and B,

෨ ൌ ͸Ͳ ‫ כ ୵ כ‬୦୭୳୰ୱ (15)

భ
ത = ൫෨൯ౣౡ ‫  כ‬ൌ ʹ͹͵Ǥͻ
 (16)
Bending An
ngle
Figure 6. Plot of Velocity (vs) Ben
nding Angle (Under No Load)
 Figure 7 shows the plot of the linear velocity of the wheel
chair subjected to weight of 60 kg. The speeed increases with
the increase of bending angle. However, iit is reduced in
comparison to the no-load condition.

V
e
l
o
c
i
t
y

(
m
/
s
)

Figure 9. Instrumented Glove with Flex

F Sensors and a Wireless
Bending Angle transmitter
Figure 7. Plot of Velocity (vs) Bending Angle (U
Under Load)
VII. REFERE
ENCES
[1] Keyur Desai, John D. Enderle. Motorized
M Chair. University of
VI. CONLUSION Connecticut.
In this paper, a novel method for wheell chair control is [2] H. Kazerooni, D. Fairbanks, A. Chen, G. Shin (2006). The Magic Glove
discussed. A prototype of the system has been developed which University of California at Berkeley,Beerkeley, California.
is believed to provide better control to peoople with severe [3] T.G.Zimmerman et al., “A Hand Gesture G Interface Device.” Proc.
Human Facttors in Computing System ms and Graphics Interface, ACM
disabilities in comparison to the traditionall joystick-control Press, New York, April 1987.
because of the lesser amount of force requireed to manipulate
the hand-glove in contrast to the joystick. Further research
includes the integration of an obstacle avoidannce and collision
detection system. There is also a huge scopee to improve the
braking mechanisms of the system on whhich research is
underway.

Figure 8. The Prototype