Autonomous navigation in a dynamic environment
Henry C. Davies, Ali E. Kayaalp, Saied Moezzi · Proceedings of SPIE, the International Society for Optical Engineering/Proceedings of SPIE · 1991
Navigation of a mobile robot in the presence of dynamic obstacles requires fast data collection routemonitoring and route replanning. A method for route replanning in the presence of dynamic obstacles and an extendible autonomous vehicle navigation architecture based on this method is presented. At the center of our architecture is the Dynamic World Modeling subsystem which maintains an internal representation of the environment. It continuously assimilates time-varying stereo-based range data to build an accurate model of the world. The primary user of this information the Navigator subsystem plans a path for the vehicle to follow and insures that this path remains valid by computing detours when environmental changes block the planned path. A preliminary version of our autonomous navigation system utilizing a near real-time stereo range detection module is operational in our laboratory. It is used to navigate a Denning MIRV-2 mobile platform and executes in a distributed fashion on a network of Sun workstations.