Bilateral Upper Limb Rehabilitation



Current upper limb rehabilitation for stroke patients is based on a conventional physical therapy supervised by therapists. This research explores end-effector devices for bilateral upper limb rehabilitation. Through bilateral interaction schemes, it will encourage the patients to proactively carry out their rehabilitation therapy and remove the stress induced in physical therapies in the clinic. ReachHab is a bilateral upper limb rehabilitation robot for post stroke functional recovery. Patient’s left and right limb can be cooperatively controlled by using the bilateral robot. Universal robot is a 5-DOF commercially available serial robot that can be used as an upper limb end-effector. Patient’s impaired limb can be assisted by the universal robot (UR-1) that is master-slave controlled by another universal robot (UR-2).


The ReachHab is the appropriate choice for both hospital and home use. Different with other low-cost devices in the markets, it can provide an assistant force in the recovery process. It makes severe patients can take passive rehabilitation exercises under totally device leading and minor stroke patients take active rehabilitation exercises by their own intention. Some effective recovery tasks are achieved by analysing the conventional physical exercise for stroke. The research of human-machine interaction is conducted to reveal a method in interaction tasks design. Different recovery patterns are compared in this article to select the appropriate training model in upper limb rehabilitation. The approach built in this research offers a potential inspiration in upper limb rehabilitation device development. This device has been further improved for bilateral rehabilitation use in clinics, and clinical trials have been conducted on a number of stroke patients in the Jiaxing Hospital, China.


The objective of UR project is to design and implement a shoulder and elbow rehabilitation device using the Universal Robot. The device encompasses a mechanical structure which provides comfortable and effective interaction between patients and the Universal Robot and a user-friendly software interface for patient to operate. A mechanical end effector has been designed by taking into account requirements, safety and biomechanics. The software user interface enables the patient to exercise in different modes including patient active and passive motion using either one or both of the robots. Task oriented exercises and computer gaming has been incorporated to make the rehabilitation more efficient.

Performance of the device was satisfying which demonstrated one robot was able to perform accurate mirrored path when the other robot was moved by the user. Testing with healthy participants has been completed and positive feedback was received. For the self-controllable human robot bilateral interaction, it is to re-produce a mirrored motion which should be as close as possible to the original healthy limb motion. In the future, we will design self-controllable bilateral interaction schemes by capturing healthy limb movement using another robot. In addition, we will also design adaptive robot control algorithms to adjust the level of assistance accounting for the stroke patients’ individual needs.