A decentralized fault-tolerant weights based algorithm for coordination of swarm robots for a disaster scenario

Abstract

Multi-robot systems are reliable and faster at searching and retrieving targets in times of crises as compared to single-robot systems. They are also better compared to manned missions as they are more resistant to hazardous conditions. In this work a method to deploy a swarm of robots to scan and retrieve the targets stuck in a disaster situation is deliberated. The multi-robot system developed is fault tolerant and follows a complete decentralized approach. In the proposed work, the basic elements in a real time search and rescue mission such as exploration of disaster area, target and obstacle detection, path-planning to reach the target in a coordinated way and tolerance for a few possible faults are accomplished using the “weights based algorithm” and the prototype testing done mimics the real time scenario. Communication among the robots in the swarm is implemented by using X-bee modules. The robots search the arena till either all the targets are found or the entire arena is explored and update a common map which renders the approach a decentralized behavior.