Robot Motion Planning for Moving Objects

Abstract This paper describes a robot motion planning in a dynamic environment, that is, a trajectory generation to attain a working purpose with redundancy free to use under the restriction of time and space. Especially we focus on “Juggling” as the typical example of such a working purpose. To juggle several balls repeatedly by a real robot, we propose a parametric representation of the robot motion, a hierarchical learning control strategy, and a real-time modification of the end-tip trajectory using visual feedback. Experimental results obtained using a 2 d.o.f. direct-drive manipulator are also shown