Capturing hand tremors with a fuzzy logic wheelchair joystick controller

We have designed and built a fuzzy logic wheelchair controller which minimizes the effect of Multiple Sclerosis and tremors. The aim of our project has been to give people with Multiple Sclerosis better control of an electric wheelchair by removing tremors from the joystick signal. The system intercepts the signal from the joystick and then passes it through the fuzzy logic controller. The fuzzy logic identify and eliminate erratic or unusual movements, employing a history mechanism to determine what "unusual" is. The fuzzy logic than outputs a signal which closely represents the intent of the user. This paper reports on the experiments conducted with our prototype wheelchair, using test volunteers with MS, as well as on the design of a new fuzzy controller. Also, we give a brief overview of the variety of recorded tremors. We show that those who have the most severe MS tremors benefit from the system, and are able to control the wheelchair safely.