Abstract
This paper presents a novel upper limb rehabilitation system for stroke patients, combining a patient-specific exoskeleton with an integrated Augmented Reality (AR) platform. By leveraging advanced technologies such as gesture recognition and AR tools (e.g., Microsoft HoloLens), the system aims to enhance functional recovery and user engagement. A custom-designed exoskeleton prototype has been developed and tested, emphasizing precise joint trajectory control, kinematic accuracy, and adaptability to individual user needs.
Validation has been conducted through bench tests and human-in-the-loop experiments, focusing on control performance, ergonomic design, and safety. Moreover, the development and testing of the AR-assisted exoskeleton system were carried out with continuous input from a physical therapist, ensuring that clinical expertise was embedded from the conceptual stage onward. These results support future integration into customized rehabilitation protocols for better neurofunctional outcomes by confirming the viability and clinical potential of intelligent, user-adaptive exoskeleton systems.