A Simplistic Approach To Reactive Multi Robot Navigation In Unknown Environments
William T. MacKunis, Daniel Raviv · 2020
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, autonomous robot groups is indispensable.In this paper, a novel approach to multi-agent control is analyzed in which concepts from several different engineering disciplines are unified within a simplistic yet robust framework.While many recently researched strategies for multi-agent control require cumbersome calculations and complex sensor layouts, the approach presented here enables a multiagent group to perform very accurate navigational tasks using minimal sensory equipment and simple geometrical calculations.This approach incorporates various concepts from geometry, trigonometry, sensor fusion, software/hardware integration, and RF communication in order to explore a method that can be used to simplify the overall process of coordinating multi-robot navigation in unknown (unmapped) environments.Further, the purpose of this paper is to present a possible new direction for engineering education in the area of multi-agent robotics.Simplicity is one of the most important aspects of a multi-agent control scheme.Many recently researched strategies to multi-agent control tend to increase the complexity of their respective systems for the purpose of attaining superior results.The research presented here utilizes the concept of Occam's Razor or the principle of parsimony, which states: "It is vain to do with more what can be done with less."This concept is extremely profound from an engineering standpoint, and engineering education would be greatly enhanced by emphasizing this axiom.In this paper, a multi-agent robot group is shown to demonstrate extremely efficient navigational patterns based upon a control scheme that incorporates the concept of Occam's Razor.The fundamental idea upon which this approach is based is that a multi-robot team can cooperate to determine the shortest 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.The idea was implemented with two robots.In simulation, groups consisting of over sixty agents were tested, and it was shown that a near optimal (shortest) path emerges through the environment without mapping the environment.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 Page 8.116.