Wheelchair-mounted robotic arm (WMRA) is a typical assistive robot system for the elder/disabled people. This paper studies the motion mapping from a human master arm to a heterogeneous slave WMRA for manipulation in activity of daily livings. The motion of the human master arm is measured via a wearable motion capture system

and then the motion captured data is transformed into the joint space of the heterogeneous slave robot arm using a corresponding joint-set motion mapping algorithm

and an autonomous obstacle avoidance algorithm is also developed. As a result

the motion of the heterogeneous slave WMRA can be easily controlled by the user via the motion of his/her arm. This method possesses two remarkable advantages that are vital to the application of the WMRA systems:the elder/disabled user can operate the WMRA system intuitively and conveniently

and the fatigue degree and error rate of the user can be significantly reduced; the safety of WMRA in an unstructured and constrained home environment can be significantly improved.