Path planning for planar articulated robots using configuration spaces and compliant motion

This paper presents a path-planning algorithm for an articulated planar robot with a static obstacle. The algorithm selects a robot part, finds a path to its goal configuration by systematic configuration space search, drags the entire robot along the path using compliant motion, and repeats the cycle until every robot part reaches its goal. The planner is tested on 11000 random problems, which span dozens of robot/obstacle geometries with up to 43 moving parts and with narrow channels. It solves every problem in seconds, whereas randomized algorithms appear to fail on all of them.