Assist-As-Needed Electromyography-Based Control for a Wearable Ankle Robotic Orthosis

Abstract

The challenge of addressing the critical public health issue of lower limb disability has led to the exploration of robotic assistive devices for rehabilitation, as a complement to conventional therapies. However, achieving optimal synergy between robotic activity and human effort remains a persistent challenge. Even the optimal control of Assist-as-Needed (AAN) remains to be solved. This study addresses this challenge by proposing an AAN electromyography (EMG)-based control strategy to automatically assist the walking motion throughout the entire gait cycle. The AAN EMG-based control strategy was integrated into an active orthosis to provide automatic ankle assistance. Preliminary results from a healthy male participant demonstrate that the ankle orthosis increased its plantar flexion assistance by 28% when the user's ankle joint torque decreased by 6% and his EMG signals from Gastrocnemius Lateralis and Tibialis Anterior decreased by 41% and 29%, respectively. In addition, the results show that the ankle orthosis was able to perform the gait pattern when there was no user participation, demonstrating the system's AAN capacity.