Bio-inspired Navigation of Autonomous Robots in Heterogenous Environments

Thomas Schmickl, Christoph Möslinger, Ronald Thenius, Karl Crailsheim
International Journal of Factory Automation, Robotics and Soft Computing 3 (2007), 164-170

Abstract:

Swarms of autonomous robots demand for simple, robust and flexible algorithms for navigation and communication. Biological evolution has developed behaviors in animals which are efficient and robust. Inspired by the trophallactic behavior (mouth-to-mouth feedings) of social insects, we developed a simple local-neighbor communication strategy that allows a swarm of autonomous robots to make optimal collective decisions concerning the navigation of individual robots [21, 19]. In this article we present a novel elaboration of this distributed algorithm, that allows the robot swarm to collectively avoid unsuitable terrain by forming trails of robots that circumvent such areas. We demonstrate the key features of this new algorithm, analyze its performance in several environmental situations and show some interesting solutions found by the robot swarm in complex environments.

Arena with dirt (blue), dump (yellow) and unsuitable terrain (green); robot paths


Download PDF:
Projects: