A fast, parallel method for computing configuration space maps is presented. The method is made possible by recognizing that one can compute a family of primitive maps which can be combined by superposition based on the distribution of real obstacles. This motion planner has been implemented for the first three degrees-of-freedom of a Puma robot in *Lisp on a Connection Machine with 8 K processors. A six degree-of-freedom version of the algorithm which performs a sequential search of the six-dimensional configuration space, building three-dimensional cross sections in parallel, has also been implemented.>

The authors describe a method for coordinating the trajectories of two robot manipulators so as to avoid collisions between them. It is assumed that the robots' environment is known and that the robots' paths can thus be planned in advance but that there may be significant variations in the execution time of some of the path segments. The goal is to allow the motions of each manipulator to be planned nearly independently and to allow the execution of the path segments to be asynchronous. The coordination is achieved by introducing explicit coordination commands into the path. The key problems in coordinating trajectories are to avoid collisions between the two robots and to avoid deadlock, that is, situations where each manipulator is waiting for the other to proceed.

We describe a robot system capable of locating a part in an unstructured pile of objects, choose a grasp on the part, plan a motion to reach the part safely, and plan a motion to place the part at a commanded position. The system requires as input a polyhedral world model including models of the part to be manipulated, the robot arm, and any other fixed objects in the environment. In addition, the system builds a depth map, using structured light, of the area where the part is to be found initially. Any other objects present in that area do not have to be modeled.