Sensor-based modification of manipulator trajectories for applications in rehabilitation robotics

Abstract

This paper presents a strategy for the use of sensor information in a general purpose teleoperation system for rehabilitation. Current research in teleoperation and mobile robotics is reviewed and its relationship to the design of a rehabilitation manipulation system is discussed. Human models of reaching and grasping tasks are used as the basis for the development of sensor-based strategies for interaction with the environment. Integration of the strategies with the Cambridge University Robot Language is discussed, and quantitative results of experimental trials are presented.