J Intell Robot Syst (2012) 66:321–342 DOI 10.1007/s10846-011-9620-2 Decentralized Cooperative SLAM for Sparsely-Communicating Robot Networks: A Centralized-Equivalent Approach Keith Y. K. Leung · Timothy D. Barfoot · Hugh H. T. Liu Received: 14 December 2010 / Accepted: 13 July 2011 / Published online: 18 August 2011 © Springer Science+Business Media B.V. 2011 Abstract Communication between robots is key to performance in cooperative multi-robot sys- tems. In practice, communication connections for information exchange between all robots are not always guaranteed, which adds difficulty in per- forming state estimation. This paper examines the decentralized cooperative simultaneous local- ization and mapping (SLAM) problem, in which each robot is required to estimate the map and all robot states under a sparsely-communicating and dynamic network. We show how the exact, centralized-equivalent estimate can be obtained by all robots in the network in a decentralized manner even when the network is never fully connected. Furthermore, a robot only needs to consider its own knowledge of the network topol- ogy in order to detect when the centralized- equivalent estimate is obtainable. Our approach is validated through more than 250 min of hard- ware experiments using a team of real robots. The K. Y. K. Leung (B ) · T. D. Barfoot · H. H. T. Liu University of Toronto Institute for Aerospace Studies, 4925 Dufferin St., Toronto, ON, M3H 5T6, Canada e-mail: keith.leung@robotics.utias.utoronto.ca T. D. Barfoot e-mail: tim.barfoot@utoronto.ca H. H. T. Liu e-mail: liu@utias.utoronto.ca resulting estimates are compared against accurate groundtruth data for all robot poses and landmark positions. In addition, we examined the effects of communication range limit on our algorithm’s performance. Keywords Networked robots · Decentralized state estimation · Finite sensing and communication · SLAM · Autonomous agents 1 Introduction A cooperative multi-robot system is beneficial in many applications. It allows for the implementa- tion of complex strategies that require more than a single robot. Multiple robots can also provide a certain degree of redundancy to ensure the com- pletion of tasks should a portion of the multi-robot team become disabled. Communication and the mutual exchange of information are key perfor- mance factors for many cooperative multi-robot systems. However, the limitation of communica- tion range and its impact on a multi-robot system have only occasionally been the focus of research. In this paper, we examine the cooperative de- centralized simultaneous localization and map- ping (SLAM) problem, in which we require each robot to estimate the map (landmarks) and the state of all robots in a sparsely-communicating