A step toward increasing automation in robot-assisted rehabilitation

This paper presents an intelligent control architecture for robot-assisted rehabilitation after stroke that has the potential to minimize the workload of the therapist and incorporate patient feedback into autonomous modification of the task presentation during therapy. In particular, the presented intelligent controller will be able to administer a therapy task, monitor safety issues, provide assessment of the progress, and incorporate patientpsilas feedback in order to make the necessary modifications of the task to impart effective therapy in an automated manner. Hybrid system modeling technique is used to design the intelligent control architecture. Experimental results with an unimpaired subject are presented to demonstrate the efficacy of the proposed control architecture.