An Adaptive Assistance Strategy Based on Handling Load and Kinematics to Control a Back Exoskeleton for Lifting Assistance

Handling loads are strongly associated with back muscle strain. However, most existing active back exoskeletons do not explicitly account for the influence of handling load on assistive torque. This study presents an adaptive assistance strategy for controlling an active back exoskeleton during lifting tasks. The strategy combines a handling-load signal obtained from four glove-mounted resistive pressure sensors with torso kinematic information measured using a trunk-mounted inertial measurement unit (IMU), enabling the system to adapt to different load weights. Unlike conventional kinematics-based assistance strategies, the proposed strategy uses the pressure-sensor-derived handling-load signal as a continuous input to modulate the load-compensation torque and provide greater assistance when heavier loads are lifted. Additionally, we developed an active back exoskeleton and deployed the adaptive assistance strategy onto the system. To evaluate the effectiveness of the adaptive assistance strategy, we recruited ten participants to perform lifting tasks under three conditions: without the exoskeleton, with the exoskeleton using either a kinematics-based or adaptive assistance strategy. Experimental results demonstrate the efficacy of the adaptive assistance strategy in reducing lumbar muscle activation. Specifically, the root mean square, average rectified value, and 90^th percentile value of muscle activation decreased by up to 34.94%, 40.89%, and 46.92%, respectively.