Physical human-robot interaction: Increasing safety by robot arm’s posture optimization

Abstract

To have robot manipulators working alongside with humans is a necessity in service robots. Obviously, in these robotics applications, human safety has precedence over precision and repeatability, which are the most important qualification of the conventional industrial manipulators. The safety measures can be taken either in the hardware or in the software or in both. This work by using a redundant manipulator aims at providing a safety measure through controlling the self-motion of the manipulator. The self-motion of the manipulator is controlled to change the posture of the manipulator to minimize or maximize the forces it can exert along a given direction. In this way, by knowing the location of the human or a delicate piece that it should not harm, manipulator’s posture is optimized to exert the minimum amount of forces during an unexpected collision. The control algorithm for this objective is described in this paper and it is evaluated through simulation tests on a redundant lightweight robot manipulator.