Trajectory Prediction in Upper-Limb Robotic Rehabilitation and its Applicability to Post-Stroke Patients: A Preliminary Analysis

Rehabilitation based on repetitive, goal-oriented movements appears effective for post-stroke patients, with robotic systems providing controlled and consistent exercises. However, despite advancements in robotic rehabilitation platforms, recent studies have questioned their cost-effectiveness and functional impact compared to traditional therapies. This underscores the need to reduce costs and improve therapeutic outcomes. A potential approach to enhance robotic therapy is integrating patient-specific motion prediction, enabling robots to adapt to individual movement patterns and anticipate patient actions during rehabilitative tasks. Leveraging these techniques could allow robotic platforms to achieve adaptive assistance, creating personalized protocols that align with each patient's needs and progress. Human intention estimation and trajectory prediction have been extensively studied in various robotics domains, from autonomous driving or surveillance to scenarios characterized by a closer human-robot interaction, such as service robotics. By delineating the current shortcomings, the outcomes of this work highlight future investigation in the field of human motion prediction for the challenging applicative scenario of upper limb robotic rehabilitation of post-stroke patients, who often experience significant variability in motor control and recovery, analyzing how the capability to anticipate patient movements could enhance its therapeutic efficacy. The study concludes by presenting the concept of a novel robotic platform prototype that could enhance rehabilitation therapy in this scenario.