• upper limb
• shoulder
• task-oriented training
• exoskeleton
• spasticity
• stroke
• elbow
• wearable robotics
• robot

Task-oriented movements can improve strength and coordination of upper limb in post-stroke patients, leading to a major functional recovery. This study focused on the development of a multi-joint task-oriented control system for two powered shoulder-elbow exoskeletons, developed at the BioRobotics Institute, Scuola Superiore Sant’Anna (Pisa, Italy). A finite-state machine has been developed for real-time recording of functional trajectories to be used for replicating task-oriented movements in rehabilitation scenarios. Trajectories are fed back to the low-level controllers of the exoskeleton, to either guide passively the user arm along via position control, or provide assistance during active movements, via torque control. An experimental procedure has been designed to verify the safety of the control system on healthy subjects and a post-stroke patient (Fugl-Meyer score = 60). Different metrics related to smoothness, execution time and velocity of the movement were extracted offline. In addition, single-joint quasi-static movements were performed to isolate the subjects’ behavior in terms of joint torque from external interactions (e.g. gravity, friction).
Statistical analysis (Wilcoxon rank sum test) showed no significant differences for all metrics and for all joints between healthy subjects and the patient. This may be due to the small number of subjects included in this pilot study and to the high residual mobility exhibited by the patient, which makes the metrics comparable between the two groups. For future studies, a larger pool of subjects will be analyzed, to evaluate the effectiveness of the proposed method and extracted parameters for assessing the patients’ condition and the course of the therapy.