Controllability Characterizations of Leader-Based Swarm Interactions

© 2012, Association for the Advancement of Artificial Intelligence (www.aaai.org). All rights reserved.The definitive version of this paper is available at: http://www.aaai.org/ocs/index.php/FSS/FSS12/paper/view/5543/5832Presented at the AAAI Symposium on Human Control of Bio-Inspired Swarms, Arlington, DC, Nov. 2012In this paper, we investigate what role the network topology plays when controlling a network of mobile robots. This is a question of key importance in the emerging area of humanswarm interaction and we approach this question by letting a human user inject control signals at a single leader-node, which are then propagated throughout the network. Based on a user study, it is found that some topologies are more amenable to human control than others, which can be interpreted in terms of the rank of the controllability matrix of the underlying network dynamics, as well as, measures of node centrality on the leader of the network