Inclusive and seamless control framework for safe robot-mediated therapy for upper limbs rehabilitation

Robot-based rehabilitation requires not only the use of a suitable robot, but also an optimal strategy to guarantee that the interaction forces with the patient fit his or her impairment level. In this work, an inclusive and seamless control framework for upper limb rehabilitation robots is presented and validated. The proposed control framework involves 1) a complete set of training modes (assistive, corrective and resistive) that can be adapted to the needs of the different states of the patient's recovery, and 2) three different advanced controllers (position, force, impedance) to track safely the force and motion references defined by the aforementioned training modes. In addition, the proposed framework allows one to tune the parameters critical to the safety of the user, such as the maximum interaction forces or the maximum speed of the robot movement. In order to validate the proposed control framework, a set of experiments have been carried out in the Universal Haptic Pantograph (UHP) upper-limb rehabilitation robot. Results show that the proposed control framework for robot-mediated therapy works properly in terms of adaptability, robustness, and safety, which are crucial factors for use with patients.