Wireless Head Gesture Controlled Robotic Wheel Chair for Physically Disable Persons

A robotic wheelchair is assumed to be capable of doing tasks like navigation, obstacle detection, etc. using sensors and intelligence. The initial part of the work was development of a cap controlled wheelchair to test and verify the gesture operation. Following that, a real time operating wheelchair was developed consisting of mode changing option between joystick control mode and head gesture control mode as per as the user’s requirement. The wheelchair consists of MPU6050 sensor, joystick module, RF module, battery, dc motor, toggle switch and Arduino. The movement of the head is detected by MPU6050 and the signal is transmitted to the microcontroller. Then the signal is processed by controller and motion of wheelchair is enabled for navigation. The wheelchair was capable of moving left, right, forward and backward direction. The speed of the wheelchair was 4.8 km/h when tested. Design objective of the wheelchair included cost effectiveness without compromising safety, flexibility and mobility for the users.


Introduction
Disability is a curse for a country particularly developing country. It is regarded as an embarrassment to the family the disabled people belongs to. This negligence of the society makes a bar between the persons with disabilities and the others in the society on normal economic, social and political activities in their families, communities, essential services and education. According to World Health Organization (WHO), "More than one billion people in the world live with some form of disability, of which nearly 200 million experience considerable difficulties in functioning" [1]. The amount of persons with disabilities is so high that it requires special consideration. WHO estimated the prevalence of the disabled person at 10.5% which is more common in the rural areas of the country [2]. The total rate of the physically disabled population is 27.5%. In children below 18 years, the prevalence of disability is estimated at 6% and for the age group above 18 years, it is 14%. It can be regarded as 3.4 million children with disabilities and 10.2 million adults with disabilities. And about 4.6% of people get disability due to spinal cord injury. To help physically challenged persons, there are many assistive technologies & devices available [3]. Among the approaches to deal with disability issue, electric wheelchair is very much reliable. Mohammed Ismail. B et al. [3] developed a simple cost effective wheelchair with minimum assistance to the person suffering with Quadriplegia or Paraplegia. They proposed a microcontroller based control system by employing simple head gestures for directions like left, right, front, and back which is done by Micro-electromechanical systems (MEMS) technology. Rushikesh T. Bankar et al. [4] proposed a Field Programmable Gate Array (FPGA) based gesture recognition system on a wheelchair. The system was developed based on basically works on the principle of acceleration. The accelerometer is connected to ATMEGA328 Microcontroller which controls the direction of wheelchair. Shayban Nasif et al. [5] proposed a device which worked based on the Head Gesture Recognition technique using Acceleration sensor. Acceleration sensor is used for the head gesture recognition and RF (radio frequency) module is used for the smart wireless control. Jigmee Wangchuk et al. [6] designed a cost effective reliable robotic wheelchair which had capabilities of navigating, detecting obstacle and moving automatically by utilizing sensors which was controlled by head gestures. The head motion is calculating sensor value. Wheelchair avoids obstacles by ultrasonic sensor. Ankit Kunti et al. [7] proposed a micro-controller based smart wheelchair with different head gestures of the user. Ultrasonic sensor has been used as the medium of obstacle detector. G Azam et al. [8] proposed a voice controlled wheelchair which employs a voice recognition system which consists of microcontroller, microphone, voice recognition processor, motor control interface board to move the wheelchair. Users can operate this wheelchair by simply speaking in front of microphone. Kazi Ehsanul Karim et al. [9] proposed automated wheelchair with vertically adjustable seat which is microcontroller based. The fabrication required microcontroller, motor driver, DC motors and battery. Functions of the motor are controlled by microcontroller and motor driver by using Pulse Width Modulation (PWM). Trinayan Saharia et al. [10] proposed a joystick controlled wheelchair where microcontroller was used to process all the command. The controller sends digital signal to the motor driving Integrated Circuit (IC) and this IC control the movement of wheelchair. The input is sent from joystick which basically are the required commands needed to move the wheelchair. However, it is observed that most of the wheel chairs men-tioned are not accessible for mass people in the developing countries like Bangladesh.
In this paper, we are focusing on fabricating a wireless head gesture controlled smart wheelchair with joystick control. It will develop the self-assistive skill and they will be much more independent on controlling their movement. DC motor is used for the mobility of the wheelchair and sensor is used to sense the movement of head. We use simple sensors so that it can be cost effective & there is a joystick module as a second control system for user interface. The motor we used are very powerful, but rest of the components are easy to access & user friendly. This wheel chair can be controlled both head gesture & joystick by switching their mode in a toggle switch.

Methodology
The model consists of transmitting part and receiving part. The block diagram of the proposed model consisting with transmitter and receiver section is shown in Figure 1. RF transmitter and RF receiver are used in this case. The MPU6050 sensor senses the rotation on X-axis and Y-axis. Based on that, the RF transmitter sends messages to RF receiver. If the sensed rotation is at positive X-axis, the transmitter sends specific data. That means motor driver will drive the wheels at left. If the sensed rotation is at negative X-axis, the transmitter sends another specific data. That means motor driver will drive the wheels at right. If the sensed rotation is at negative Y-axis, the transmitter sends another value. That means motor driver will drive the wheels to forward. RF receiver is interfaced with Arduino mega2560 which accepts the messages from transmitter and sends them to Arduino mega2560. Arduino mainly converts these messages into American Standard Code for Information Interchange (ASCII) code and then this code is decoded. DC motors are driven by BTS7960 motor driver. Led acid battery is used for the purpose of mobility.
Prototype design of the wheelchair is discussed below.

Prototype Design of Wheelchair
For better output, at first we made a prototype in which we mainly focus on the head gesture control system. In this prototype we use small dc motor, an Arduino, RF module & L298N motor driver module. We have used RF module to reduce the use of wire and avoid complex circuitry.
In Figure 2, the transmitting section with Arduino Nano is shown. The sensor is also connected in the transmitter. The Vcc & GND pin of both sensor & transmitter is connected with the +5 V & GND pin of Arduino and D12 is used for transmitter. A4 & A5 is used for the sensor data.
In Figure 3, we see that there are a receiver, a motor driver (L298N), a battery, two dc gear motor & an Arduino mega2560. Here is a 12 V supply which is given into 12 V port of motor driver module. The enable & input pin is connected into PWM section of the arduino. Both the output is connected into the motor.
Moreover arduino powered from the motor driver 5 V output port. There is a battery which gives power to the whole circuit (Table 1).    Nano is a small, complete, and breadboard-friendly board based on the ATme-ga328P (Arduino Nano 3.x). It has more or less the same functionality of the Arduino Uno, but in a different package. The receiver section of the prototype wheel chair is shown in Figure 4.
The receiver section consists of Receiver, Arduino Mega, L298N motor and wheel. The receiving signal will be analyzed by Arduino Mega which is connected to L298N motor driver. According to signal, the motor driver connected to the wheels will control the four directions which are moving left, right, forward and backward ( Figure 5).
In the prototype design, an arduino nano is connected with the sensor & transmitter. The arduino is powered by a battery and the whole system is mounted in the cap. There is a delay which makes the cap more effective from random tilt of head. That means if anyone move their head randomly the gesture won't work. To work the gesture, one must keep the head to the desired direction for a certain time.

Hardware Implementation
After completing the prototype design and tested properly, we implemented the real wheel chair. But we use some upgraded motor, motor driver module and more powerful battery for the real structure. We use stainless steel pipe to make the frame. The back wheel we used in this structure is cycle wheel. Caster wheel is used as front wheel. The schematic diagram is shown in Figure 6.   The total height of the wheelchair from the ground is 44''. Receiver section is shown in Figure 7.
In Figure 7, we see that there is joystick module, motor driver (2), Receiver, Motor, Battery & a toggle switch. Two motor driver modules are connected with the digital PWM pin of the Arduino. There is also a connection of the 24 V battery in both motor drivers. Both the motors are also connected into the motor driver. To give supply in the Arduino we use a small battery. For the transmitter we use same circuit that used for the prototype design. We use the similar components that used for prototype designing in the real wheel chair, but we use upgraded dc motor, motor driver module, battery, wheel & joystick module is also added. The list of components used for real wheel chair is shown in Table 2.
To understand the algorithm of the project there is a flow chart which describes the whole algorithm. The flow chart is shown in Figure 8.
Here we see that, the whole algorithm is mainly depended into the toggle switch mode. If it is high, then the joystick control algorithm is enabling. But if Table 2. List of components used for real wheel chair.  the toggle switch is off, that means the mode is low, then head gesture control system is on. Then depending on the sensor values, the wheel chair will work.
Based on the algorithm, the hardware implementation of the real wheel chair is shown in Figure 9.   The output of left, right, forward and backward direction with real image is shown Figure 10.
After interfacing the joystick module with the microcontroller, we can expect four outputs from the wheelchair. As 2-axis joystick has been used, we can get right, left, forward and backward directions for different values in the x-axis and y-axis. From the table given above, we can see that for negative values in the x-axis, the wheelchair will move to the right. For positive values in the x-axis, the wheelchair will move to the left. For positive values in the y-axis, the wheelchair will move backward. And for negative values in the y-axis, the wheelchair will move forward. The joystick value is shown in Table 4.
For smoothness we gradually increase the speed of the motor. After taking several reading we find out that the average time required to reach in its maximum speed which we provided is 3.8 seconds. We also calculate the velocity by taking a fixed distance of two (2) meter.   The cost analysis of the wheelchair is shown in Table 5. Total cost of wheelchair is BDT 41,000 that is equivalent to USD 500 only excluding the labour cost which is very cheap compared to the chair available in the market.

Conclusion
In this paper, we develop a wheel chair which is cost effective for a disabled person to lead his or her day to day life. We use a simple sensor that makes it cost effective & there is a joystick module as a second control system for user interface. The motor we used are very powerful, but rest of the components are easy to access & user friendly. This wheel chair can be controlled both head gesture & joystick by switching their mode in a toggle switch. Before implementing the real wheel chair, we develop a prototype cap controlled wheelchair and tested the operation and to work gesture properly, one must keep the head to the desired direction for a certain time. After that, real time Robotic Wheel Chair is developed that has mode changing options which are joystick control mode and head gesture control mode as per as the user's requirement. The controller processes the signal and enables the motion of wheelchair for its navigation. The implemented wheelchair functioned properly and moved left, right, backward and forward direction according to the program given into microcontroller. The speed of the wheel chair is 4.8 km/h. Moreover, the wheelchair is designed in a cost-effective way but ensures safety, flexibility, and mobility for the users.