Adaptive Robot-Assisted Telerehabilitation System Using Model Predictive Control and Digital Twin for Personalized Upper Limb Therapy

Abstract

Smart rehabilitation systems have revolutionized the delivery of upper limb therapy, with treatments that are more accurate, adaptive, and patient-focused. An advanced rehabilitation system integrating Model Predictive Control (MPC), Digital Twin visualization, and Augmented Reality (AR) interfaces ensures personalized and adaptive therapy. Real-time data are processed by the system with low latency of 10 ms, reducing X, Y, and Zaxes trajectory tracking errors. The system changes treatment from 80 units to 79.85 units according to the patient's level of fatigue (0.5) and enhances comfort and rehabilitation results. The Digital Twin module provides an immediate virtual representation of therapy progress, and AR interfaces enable remote monitoring and control. The integration of MPC, AR, and Digital Twin technologies offers an efficient and highly adaptive rehabilitation approach that is well-suited for clinical applications as well as for at-home therapy. The result indicate the system had the potential to enhance patient care, refine therapy accuracy, and maximize rehabilitation effectiveness.