You are on page 1of 5

Journal of Scientific & Industrial Research

444
Vol. 69, June 2010, pp. 444-448

J SCI IND RES VOL 69 JUNE 2010

Low cost prototype development of electronic knee


Neelesh Kumar1*, Sanjeev Soni1, Amod Kumar1 and B S Sohi2
1
2

Central Scientific Instruments Organisation (CSIO), Chandigarh 160 030, India

University Institute of Engineering & Technology (UIET), Chandigarh 160 014, India
Received 20 January 2010; revised 16 March 2010; accepted 17 March 2010

This study presents design and development of an electronic knee, which uses attached sensors, microcontrollers and
electro pneumatic valve arrangement to emulate near natural walk. Being all sensors and mechanical assembly developed
indigenously, prototype would be a low cost electronic knee.
Keywords: Embedded system, Microcontroller, Prosthetic

Introduction
With increase in road accidents, machine calamities
etc, 5.5 lakhs people in India have lower limb
amputations every year and need some kind of
prosthesis1. Low cost prosthetic products are not able
to perform coordinated and complex movements of a
knee joint2,3. Sophisticated electronic knee are very
costly (OttoBock costs Rs 18-20 lakhs). With the
advent of integration of technology, numerous sensors,
microcontrollers and accurate feedback controls,
orthotic and prosthetic devices will no longer be
separate, lifeless mechanisms, but intimate extensions
of human body structurally, neurologically and
dynamically4. Advances in biomedical engineering have
led to introduction of computer-controlled, variabledamping prosthetic knees5,6. With advent in complex
feedback sensor and control system, device designs
are equipped with high degree of adaptability and
reliability7.
This paper describes indigenously designed and
developed a low cost electro-pneumatic knee by
integration of indigenously developed electrogoniometer and force resistive sensor (FSR) in order
to control swing phase dynamically.
Experimental Section
Electronic knee consists of electro-goniometer and
FSR as sensor mechanisms8,9. Knee design is based
*Author for correspondence
E-mail: neel5278@rediffmail.com

on controlling a swing phase using a pneumatic cylinder


mechanism attached with drive (motor) control mechanism
to its flow control valve. Required energy to extend knee
into new gait cycle is provided by a spring mechanism.
Electro-Goniometer

An electro-goniometer that measures angles between


body segments (Fig. 1) was developed. Material used for
fabrication was aluminum (thick, 2 mm). A single turn
potentiometer of 1 K was used. For testing of developed
electro-goniometer, experiments on individuals were
performed to achieve a suitable calibration factor. It was
observed that there is considerable change in knee flexion
angle range with walking speed.
Sensor Modules

Sensors are employed to measure walking phase and


walking speed of patient wearing prosthetic. To record
knee flexion angle, an electro-goniometer was designed
and developed. In order to determine heel strike and toe
off, which gives information about walking phase (swing
or stance), FSR sensor was fabricated and tested. FSR is
a polymer film device, which exhibits decrease in
resistance with an increase in force applied to active
surface. Its force sensitivity is optimized for use in human
touch control of electronic device. FSR was developed
using carbon impregnated conductive polyether
polyurethane foam (surface resistivity, 104-105 /square;
density, 45-55 kg/m3). It is sandwiched in two pieces of
general purpose PCBs (Fig. 2a). For a simple force to
voltage conversion, FSR device is tied to a measuring

KUMAR et al: LOW COST PROTOTYPE DEVELOPMENT OF ELECTRONIC KNEE

445

a)

b)
Fig. 1Electronic Goniometer: a) Mechanical layout of goniometer fixing clamp and potentiometer holder
assembly; b) Placement on human body

Fig. 2FSR sensor: a) Developed sensor; b) Placement of FSR sensor

resistor in a voltage divider configuration. Developed FSR


accurately gives heel strike and toe contact time and
phase information (Fig. 2b). FSR gave accurate and
repeatable results (Table 1). Developed force-sensitive
resistors were able to sustain large forces exerted while
using as foot sensors.
Electro-pneumatic Control

In present arrangement, a pneumatic cylinder at knee


joint is used to adjust swing phase resistance. Developed
electro-pneumatic cylinder (Fig. 3) contains a needle
valve to adjust resistance in knee joint. This control valve

is driven by a stepper motor. Valve adjusts flow air in


cylinder, which in turn builds required resistance.
Controller Unit

ATMEGA32L microcontroller (8-bit) was used for


processing information and controlling electronic knee.
Sensors for sensing angle and walking phase are
embedded in knee. Microcontroller in real time scans
information from various sensors for knee flexion angle
and walking phase. It combines both inputs and matches
for corresponding walking speed. This knowledgebase
was developed by extensive trial on variety of individuals.
Based on walking speed detected, it adjusted flexion

446

J SCI IND RES VOL 69 JUNE 2010

resistance of cylinder by giving control pattern to stepper


motor, which controls opening and closing of flow control
valve.
Control Algorithm

In complete system, critical part is control algorithm,


which processes information from sensors and then
operates valve. Control of electronic knee joint consists
of different modes of operation in control algorithm as

follows: i) calibration mode; ii) training mode; ii) execution


mode; iv) backup mode (safe mode); and v) fail safe
mode. Software codes for each mode are written and
tested. Artificial knee, powered with ON, operates in
execution mode (Fig. 4), which consists of slow mode,
normal mode and fast mode. In slow mode, knee detects
a particular range of sensor inputs and adjusts valve
position to offer damping required in slow walking.
Power Requirements

Device is powered with ENIX Li-Ion rechargeable


battery (15 V, 2.2 Ah). Control algorithms are designed

Fig. 3Mechanical layout of electropneumatic cylinder

Fig. 4Execution mode flow chart

Table 1FSR results for two different subjects


Subject weight
kg

Walking speed

Toe contact

Heel strike

Time, s

Amplitude, V

Time, s

Amplitude, V

48

Slow
Fast

1
0.7

3.90
4.20

0.95
0.64

3.78
3.90

72

Slow
Fast

0.65
0.62

4.01
4.05

0.85
0.59

4.03
4.01

KUMAR et al: LOW COST PROTOTYPE DEVELOPMENT OF ELECTRONIC KNEE

447

Fig. 5Prototype of developed knee joint

in such a way so that control valve will operate only


when patient changes his walking mode from slow to
normal and normal to fast. Total current drawn by
electronic knee while operating control valve is less than
200 mAmp. In all other conditions, when only sensors
give signals to controller, system draws 65 mAmp current
at 5V level.
In order to provide accurate control and save power
saving, mode change is decided only when change
conditions are met for more than two times
consecutively. As prosthetic gait is slow varying activity,
this approach would save power effectively in comparison
with normal operation of electronic knee. This means
knee can easily operate for full day walking activity of
amputee without any need of recharging. Battery status
of knee is checked by controller and it generates audio

alarm for low battery. In low battery condition, knee will


run in slow speed walking mode and provide maximum
stability to patient. After low battery, device needs to be
plug-in for charging. It will take roughly 2 h for full
charging.
Results
This study presented indigenously designed and
developed a low cost electro-pneumatic knee (length, 40
cm; wt, 1600 g; flexion angle, 60; weight-bearing
capacity, 100 kg). In prototype of knee joint (Fig. 5),
socket and sache foot was procured from a rehabilitation
center. Socket was fabricated according to residual
stump of patient. For every patient, due to difference in
knee to toe length, some minor adjustments are required.

448

J SCI IND RES VOL 69 JUNE 2010

Experimental analysis of indigenously developed and


tested electro-goniometer and FSR showed similar results
as compared to standard available sensors. Component
and fabrication cost (Rs 50,000/-) of developed prototype
is very low as compared to available electronic knee
from Otto Bock or Endolite (Rs 3.5-20 lakhs). It can be
lower when it goes for production.

Conclusions

A prototype with patient selected walking speed


through switch is tested and fabricated. Performance
quantification of knee will be done by questionnaire
methods. By incorporating more number of different
sensor mechanism for measuring kinematics and kinetics,
a near natural knee can be designed but this will raise
cost of prosthetic.
References
1

Hatwell M S, Oderkerk B J, Ari S C & Inbar G F, The


development of a model reference adaptive controller to control
the knee joint of paraplegics, IEEE Trans Automat Cont, 39
(1991) 683-691.

Irby S E, Kaufman K R & Sutherland D H, Electronically


controlled long leg brace, in Biomed Engg Conf, vol 29 1996,
427-430.
Zlatnik D, Steiner B & Schweitzer G, Finite-state control of a
trans-femoral (TF) prosthesis, IEEE Trans Cont Syst Technol,
10 (2002) 408-421.
Previdi F & Carpanzano E, Design of a gain scheduling controller
for knee-joint angle control by using functional electrical
stimulation, IEEE Trans Cont Syst Technol, 11 (2003) 310-325.
Dollar A M & Herr H, Active orthoses for the lower-limbs:
challenges and state of the art, in Proc 2007 IEEE 10th Int Conf
on Rehabilitation Robotics (The Netherlands) 12-15 June 2007.

Cheoltaek K, Ju-Jang L & Xinhe X, Design of biped robot with


heterogeneous legs for advanced prosthetic knee application,
in SICE-ICASE Int Joint Conf, (Bexco, Busan, Korea) 18-21
Oct 2006.

Andrysek J & Chau G, An electromechanical swing-phase


controlled prosthetic knee joint for conversion of physiological
energy to electrical energy: Feasibility study, TBME-004652006.R1, 2006.

Lee J-W & Lee G-K, Gait angle prediction for lower limb
orthotics and prostheses using an EMG signal and neural
networks, Int J Cont Automat Syst, 3 (2005) 152-158.

Berry D, Microprocessor prosthetic knees, Physical Med


Rehabilitat Clinics of North America, 17 (2006) 91-113.