Towards a natural interface for the control of a whole arm prosthesis

Abstract

In the present study we illustrate a new concept for a user interface to control whole arm prosthesis. We extend the myo-electric control, taking into account the intended movement of the shoulder and integrating vision analysis to give a “visual control” to the user. While the control of the shoulder is partially obtained from pattern recognition of sEMG signals, the control of the elbow and wrist joints is only derived using the trajectory computed from the initial to the target position. We show results in simulation and discuss about future steps in developing the prosthesis.