Noncontact Capacitive Sensing Based Human Motion Intent Recognition

Abstract

Accurate and reliable recognition of human motion intent is an important issue in the field of tri-co robotics. A noncontact capacitive sensing approach is proposed which can measure muscle contraction signals by freeing human skin from contacting to metal electrodes. The method is proposed to address the issues in biological-signal-based human motion sensing. We then introduced four case studies of noncontact capacitive sensing. The first study involved locomotion transition recognition on transtibial amputees wearing lower-limb robotic prostheses. The second one is the initial study on new capacitive sensing electrodes with liquid metal electrode, the purpose of which is to build a soft elastic capacitive sensing front-end for more compatible dressing. By comparison the last two studies are applications of the capacitive sensing method on human upper limb motion recognition to prove the feasibility of the new method on upper limb motion sensing. The third one is the forearm discrete motion pattern recognition and the last case is the continuous grasp force estimation with the capacitance signals. The results of the studies suggest the promise of the new sensing method. In future works, the noncontact capacitive sensing approach will be further studied for the control of wearable robots and collaborative robots.