An embedded risk prediction system for wheelchair safety driving

Development of intelligent wheelchairs can increase mobility and independence of impaired individuals. As there exist dangerous driving risks such as improper driving postures, moving too fast or on rough road, it is useful to monitor the wheelchair user’s driving conditions and, particularly, predict potential driving risks. This paper proposes a risk prediction system for the wheelchair safety driving. A novel designed smart cushion is used to evaluate dangerous drivings risks. Our cushion is able to better combine the pressure sensors and accelerometer and it can detect sitting postures, wheelchair accelerations, and terrain conditions. In addition, we propose a prediction system to monitor the wheelchair driving status. Features are extracted and a fuzzy inference system is used to quantify the dangerous driving risks. Warnings or intervention control strategies will be trigged to increase wheelchair driving safety.