Electric Wheelchair with Avoiding Obstacle Feature using LiDAR and CNN Method

Wheelchairs are one of the important medical instruments for the mobility of patients with disabilities to carry out several activities. Wheelchair has been significantly developed from the structures to features such as electric motors, controllers, and location detection system to improve the quality of safety and comfort. However, aspects of the smart obstacle avoider are currently being developed. In this study, an Advanced Driver-Assistance System (ADAS) has been developed using 2D LiDAR as an area mapping sensor, and Convolution Neural Network (CNN) to control a wheelchair embedded in a Raspberry Pi single board computer. The experimental results show that the success rate in avoiding obstacles automatically is between 76.6% to 100%.