A flow field approach to planning robot paths in parallel

Abstract

Automatic path planning is an important aspect of the development of autonomous robots. Path planning consists of deciding what motions a robot should execute in order to go from some initial configuration to a final configuration and doing so without colliding with physical objects present in its workspace. In order to embed real-time path planners in robot controllers the considerable computational burden which they present must be overcome. The motivation for the work described was to consider parallel processing as a means of achieving this. >