Characterizing Human Motor Adaptation and Learning When Using Different Occupational Exoskeletons

Abstract

Occupational exoskeletons offer promising solutions to reduce biomechanical risk factors and mitigate work-related musculoskeletal disorder risks from manual labor. However, understanding how individuals adapt to these devices over time remains limited, complicating effective design, training, and deployment. This dissertation addresses these gaps through studies examining motor adaptation and learning across diverse exoskeleton types and occupationally relevant tasks. Three separate studies were completed. First, adaptation to a whole-body powered exoskeleton was evaluated during stationary load handling. Novice users initially exhibited task completion times approximately 117% longer than experienced users, improving to 59% longer by the third session. Movement smoothness improved by 30%, and muscle activation patterns converged to those of experienced users over three sessions, yet novices retained up to ∼ 50% longer joint angular velocities. These results highlight the need for extended familiarization to achieve proficient and efficient use of complex multi-joint exoskeletons. Second, the neuromuscular effects of three different arm-support exoskeletons were assessed during lab-simulated overhead tasks. Participants recruited more muscle synergies in dynamic tasks (IQR: 2–5) than pseudo-static tasks (IQR: 1–3), with some task-exoskeleton combinations producing significant local changes in muscle recruitment strategies. Using these exoskeletons tended to shift synergy recruitment from primary arm-elevating muscles toward neck and back muscles, reflecting compensatory motor strategies for device-imposed biomechanical demands. Third, adaptation to passive and active back-support exoskeletons was examined during repetitive lifting over four sessions. Both exoskeletons increased trunk range of motion (by 11◦ to 20◦) relative to no device and reduced hip and knee range of motion by up to 50%. Ratings of perceived exertion for the lower back and legs decreased by approximately 70–75% among exoskeleton users by the final session, with continued reductions between sessions 3 and 4 indicating ongoing adaptation. Perceived mental workload decreased by nearly 50% over the session, with the active device producing less early workload than the passive device. Overall, This work advances understanding of exoskeleton adaptation across physical, neuromuscular, and perceptual domains, identifying factors that inform device design, training, and implementation to promote safer, more effective use in occupational settings.