Local Navigation Strategies for Multi-Robot Exploration: From Simulation to Experimentation with Mini-Robots

Abstract

This paper presents two local navigation strategies for multi-robot systems in performing exploration of unknown environments. The strategies are based on the frontier cell for mapping the environment and navigating in it. Additionally, two coordination strategies are used to solve possible conflicts among robots. Key criteria investigated are the efficiency and effectiveness represented by the completion time, travelled distance, the total steps, and the task distribution. The developed algorithms are tested in simulations as well as in experiments with Khepera III mini robots running on the Teleworkbench. To evaluate the robustness of the developed algorithms, tests are performed under different environmental configurations and varying numbers of robots. The results show the advantages of both proposed strategies in different situations such as environment types and starting positions.