Motion planning for a robot and a movable object amidst polygonal obstacles

The authors address the motion planning problem for a robot and a movable object within polygonal obstacles. Motion planning in this context appears as a constrained instance of the coordinated motion planning problem for two robots. The object cannot move itself, it only does when the robot grasps it. A topological property that characterizes the existence of solutions in the subspace of configurations where the robot touches the object is discussed. This property provides a general resolution scheme which is applied to the special case of a convex polygonal robot moving in translation with a convex polygonal object amidst polygonal obstacles. The algorithm has been implemented, and experimental results are discussed