Design and control of a wearable lower-body exoskeleton for squatting and walking assistance in manual handling works

Several manual-handling works (e.g. load-lifting, carrying, carpet-laying, etc.) involve repetitive lower limb movements (RLLM) such as squatting, kneeling, or walking which are notable risk factors for work-related musculoskeletal disorders of the lower-limb and lumbar spine. Wearable exoskeletons can augment workers' strength and minimize muscular activation in RLLM, thus enhancing efficiency and improving workers health. In this paper, we present the design and control of a wearable lower-body exoskeleton to assist RLLM. The design of the exoskeleton is anthropomorphic with six-degrees of freedom (6-DoF) on each leg: 1-active and 1-passive DoF on the hip-joint, 1-active and 1-passive DoF on the knee-joint, and 2-passive DoFs on the ankle-joint. The active DoFs are actuated by bi-directional brushless DC motors to permit sagittal movements. The exoskeleton controller is designed hierarchical with a low-level linear quadratic gaussian (LQG) torque controller, a middle-level user-input torque estimator based on dual extended Kalman filter (EKF), and a novel high-level supervisory algorithm for movement detection and synchronization of the exoskeleton with the user. Initial evaluation of the prototype is conducted on five healthy participants recruited to perform repetitive load-lifting and carrying manual-handling tasks. Electromyogram sensors are placed at the participants' right Vastus Intermedius and right Gastrocnemius to extract muscle activity signals. Results show that the exoskeleton could provide torque assistance by amplifying participants' input joint torques and could minimize muscular activation in the two muscles by more than 36% respectively in all movement tasks.