Gait phase detection based on non-contact capacitive sensing: Preliminary results

Gait phase detection is essential to the control of lower-limb exoskeletons. In this paper, we present a non-contact capacitive sensing strategy for gait phase detection to replace foot pressure sensors. The designed capacitance sensing system can record signals of human muscle contraction from the leg. The electrodes are non-contact with the skin, which are fixed on the particularly designed cuffs. To evaluate the performance of the capacitance sensing on gait phase detection, two experiments are conducted on healthy subjects. With selected features and sliding window classification method, the proposed method obtains 98.3% average accuracy with the sensing cuff on the shank and 96.5% accuracy with the sensing cuff on the thigh for level walking tasks. The system also accurately recognizes the gait events (largest error rate smaller than 0.6%) when walking speed changes. The preliminary results indicate that the proposed sensing strategy is a promising solution to provide useful gait information for exoskeleton control.