A robust satisficing feedback strategy for autonomous navigation

The authors present a robust algorithm for dynamic navigation of autonomous mobile robots (AMRs) using local feedback information. The structure of a previously proposed algorithm (see D. Feng and B.H. Krogh, 1989) is reviewed which uses the concept of a satisficing strategy to make steering decisions while the AMR is moving. The original feedback algorithm is modified to guarantee collision avoidance in the presence of limited sensor range, sensor and servo errors, and computational delays for processing online sensor data to update the internal map of the local visible environment. Details of the approach are presented for an omnidirectional AMR and a conventionally steered AMR. Simulation examples demonstrate the robustness of the feedback algorithm for realistic sensor and servo errors and computational delays