Abstract
A method has been developed to plan feasible, collision free paths for two spatial robots working cooperatively in a known static environment. The method also applies to single robots. The approach is c-space based and performs selective rather than exhaustive mappings. A novel, divide-and-conquer style heuristic guides the selective mapping process. The heuristic requires no precomputations, is immune to degeneracy problems, and does not involve any robot, task, or environment specific assumptions. The approach may be applied to cooperating redundant robots without requiring solution of inverse kinematics for a redundant robot. Postprocessing is applied to a safe path to improve its efficiency. Results are illustrated via example problems.
