A Simplistic Approach to Reactive Multi-Robot Navigation in

Multi-agent control is a very promising area of robotics. In applications for which it is difficult or impossible for humans to intervene, the utilization of multi-agent robot groups is indispensable. This paper presents a novel approach to reactive multi-agent control that is quite practical and elegant in its simplicity. The fundamental idea upon which this approach is based is that a multi-robot team can cooperate to determine the shortest, safest path through an unknown environment given only a very simple set of rules. In addition, the approach being presented in this paper can be successfully implemented using very simple sensors. In this algorithm, only one robot (the “scout”) is deployed into an unknown environment to randomly search the area for a goal location. The rest of the robots in the group wait at the designated starting location, referred to here as the “nest.” When the scout discovers the goal location, it sends a message to the other robots waiting in the nest containing the shortest straight-line path from the nest to the goal location. The other robots (the “nest workers”) are now able to actively seek the goal location. If the nest workers encounter obstacles during their journey, they simply avoid it and recalibrate their desired trajectory to account for the unexpected detour. When a nest worker finds the goal location, it sends another message to the group containing additional data gathered by it in its journey. This redundancy results in the emergence of a near optimal path from the nest to the goal through an environment containing random obstacles. The idea was implemented with two robots. The experimental results clearly show that the robots’ navigational ability through various unknown environments improves with time under the control of this algorithm. Furthermore, the simplicity of this approach makes implementation very practical and easily expandable to reliably control a group comprised of many agents.