Robust Rule-based Heuristic Assistance Strategy for a Semi-Active Shoulder Exoskeleton Used in Overhead Work

Semi-active shoulder exoskeletons strike a balance between active and passive exoskeletons for reducing shoulder load during overhead work. Current assistance strategy focuses on assisting with static tasks, neglecting the requirement of dynamic tasks, e.g. overhead wall painting.

This study presents a robust rule-based heuristic assistance strategy for a semi-active shoulder exoskeleton to improve task adaptability. Different from previous studies, the robust rule-based heuristic assistance strategy determines assistance by weighing both angular velocity and displacement to accommodate both static and dynamic tasks. A robust detection mechanism is also proposed to improve its stability under non-uniform and abrupt movements, making it better suited for real-world applications.

Experiments on the outputs of the robust rule-based heuristic assistance strategy showed that minimal support (Level 0) was provided for tasks below the overhead region, whereas increased assistance was offered for overhead tasks (Levels 1 to 3), with static tasks (Level 3) receiving greater support than dynamic ones (Levels 1 and 2). Results showed that applying the robust detection mechanism can achieve a reduction of up to 60.3% in the switching times the assistance level changes. Electromyography experiments showed that the exoskeleton with our strategy reduced muscle activation with reductions of up to 52.4%.

These results indicate that the proposed exoskeleton can provide effective assistance for both static and dynamic overhead tasks.

Our study provides valuable insights into the practical utility and robustness of exoskeletons, thereby accelerating their application in occupational scenarios.