Social Odometry: A Self-Organized Distributed Location Algorithm

In this paper, we study a new approach to multi-robot localization based on local communication which confers the robots the possibility to learn from the others. By communicating with the rest of the group, robots are able to correct, localize and achieve tasks they could not solve by their own. We use a foraging task as test bed where a nest and prey areas must be identified by the robots. Once both areas have been located, robots must forage from nest to prey endlessly. Each robot has an estimate of its own location and an associated confidence level that decreases with the distance travelled. The algorithm guides a robot to its goal by imitating estimated locations, confidence levels and actual locations of its neighbors. We evaluate the performance of different local communication algorithms by observing the number of times robots go from nest to prey and come back. We extend the performance analysis to different quantitative measures about the intrinsic recruitment and self-organized process.