Myo-signal based intuitive driving interface Development and evaluation

Abstract

This study proposes an electromyogram (EMG)-based driving interface, MyoDrive, which uses two wrist motions to accelerate, brake, and steer in a simple manner. The horizontal angle of the right wrist is used for steering, and the vertical angle of the left wrist is used for speed control. Each angle was estimated from the EMG signals with a simple regression method. Conventional wheel/pedal interface and the proposed interface, MyoDrive, were compared with three tasks-slalom, speed tracking, and emergency stop to investigate their intuitive controllability. The proposed interface showed better learnability and response than wheel/pedal interface, but was behind in steering and speed control performance. In the speed tracking task, as the target speed increased, the deviation of MyoDrive increased whereas the deviation of Wheel/Pedal decreased in the slow region. It meant that Wheel/Pedal controlled speed more precisely. However, MyoDrive showed faster response both in the transition state and emergency.