Numerical Simulation-Based Design of a Pneumatic Finger Rehabilitation Robot for Tele-Rehabilitation

Abstract

Hand motor function rehabilitation after stroke or traumatic injury requires repetitive, task-specific training, which is often limited by therapist availability and clinical resources. This study presents a lightweight, modular, and wearable pneumatic robotic arm for finger rehabilitation, designed to support tele-rehabilitation applications. The system employs a four-link mechanical structure that accommodates variations in finger length and enables natural flexion and extension without the need for individual customization. Motion control is achieved using a discrete-time proportional-integral-derivative (PID) controller with aerodynamic drag compensation, ensuring stable and accurate actuation under compressible air dynamics. A stage-specific pressure strategy is implemented, applying 0.1 MPa for early mobilization and 0.3 MPa for intensive training, enabling up to 80° of finger bending within 2.5 s. Network-induced latency and sensor delay are explicitly modeled in the control loop, and their effects on response time and tracking accuracy are evaluated through numerical simulations. Simulation results demonstrate a motion tracking error below 2°, with control errors remaining bounded under network latencies up to 50 ms, confirming real-time responsiveness suitable for remote rehabilitation scenarios. These findings support the feasibility of a scalable, cost-effective, and clinically viable pneumatic rehabilitation platform for individualized hand therapy and tele-rehabilitation deployment.