Experimental validation of a motion generation model for natural robotics-based sit to stand assistance and rehabilitation

Abstract

Sit to stand (STS) is an important daily activity that requires high joint torques and neuromuscular coordination. Robotics-based assistive devices have a great potential to increase independence and quality of life for elderly and disabled people. Comparison between motion kinematic of unassisted and assisted STS transfer shows that commercial lifters result in an abnormal pattern of motion. Therefore, the minimum shoulder jerk (MSJ) criterion is used in this paper to model the natural motion during STS transfer. In order to assess the ability of the MSJ criterion to replicate the natural STS motion, a testbed consists of a 2-DOF robotic arm is developed. This testbed is used to compare motion pattern and required assisting force when a subject is assisted through MSJ trajectory and the one that is obtained from a commercial lifter. The simulation and experimental results prove that the proposed approach can achieve a natural pattern of motion. Also, the required assisting force in the case of MSJ trajectory is lower than that of the commercial lifters which can reduce fatigue due to repeated lifting. The method of trajectory generation presented in this paper is shown to be easy to apply and suitable for real-time implementation. Thus, the results of this paper may be used to enhance the existing strategies of STS assistance and rehabilitation.