Self-Organized Cooperation in Swarm Robotics

Abstract

Cooperation is a key concept used in multi-robot systems for performing complex tasks. In swarm robotics, a self-organized cooperation is applied, where robots with limited intelligence cooperate and interact locally to build up the desired global behavior. In this paper, we are studying a mobile object tracking scenario performed by a swarm of robots. The robustness, scalability and flexibility of swarm robots make it an attractive approach for missions like object tracking in complex and dynamic environments. As the individual robot capabilities are limited in swarm systems, the robots may not be able to track the mobile object continuously. This limitation is overcome using the robots communication capability. In order to increase the probability of object detection, we propose a greedy self-deployment strategy, where the robots are spread uniformly in the environment to be monitored. For detecting a moving target, the robots use a biologically inspired algorithm for collecting robots currently located in other regions to track the target. In such cooperative tasks the robots normally need to be time synchronized for simultaneous activation. A new proposal for time synchronization in swarm robots is introduced which exploits the mobility of the robots for handling possible disconnections in the network and synchronize them at the beginning of tracking time slots.