Distributed Multi-Robot Exploration Approach With Connectivity Maintenance

In this paper, we propose a novel distributed multirobot exploration algorithm capable of efficiently exploring unknown environments while ensuring uninterrupted communication with each other and with a base station, including multihop scenarios when direct communication is not feasible. The system is applicable in scenarios of emergencies like search and rescue operations where time and communication reliability is of utmost importance. Our approach introduces two distinct roles, explorers and supporters, to enhance efficiency and connectivity based on live assessments of network connectivity and exploration needs. Explorers perform frontier-based exploration keeping primarily focus on maximization of information gain through unexplored parts. Supporters adopt a flocking-inspired approach for their positioning, acting as relay points to ensure strong communication links. Decision making is distributed, with local information sharing so that each robot can autonomously come to a decision in the absence of centralized control. Results of our simulation indicate that the algorithm has reduced time required for exploration by 80% and time spent on redundant exploration by 74% as compared to Random walk exploration.