Active Matter with Intent: Clog Control in Excavating Collectives

Ensembles of self-propelling elements can form clusters, clogs and jams. In certain biological and robotic systems, clog mitigation is important for collective task completion. To discover principles by which individual and group behaviors facilitate clog control in such “task-oriented” active matter, we studied tunnel excavation in granular media using fire ants, autonomous robots and two models: cellular automata (CA) and a one worker model. We used tools from the study of dense particulate ensembles to provide insight into how different excavation strategies modulate congestion dynamics. These tools elucidated how participant idleness in robots and modelled ant systems reduced tunnel density and decreased the frequency of catastrophic clogs and how selective “retreats” reduced jam dissolution time for large clogs. In simulations, the maximum flux in an experimentally derived tunnel width (2 ant body width(BW)) occurred at a particular tunnel density. Experiments with ants in tunnels(2 BW) excavated by them in laboratory revealed that the ants selected a similar density, suggesting that the ants optimize the tunnel flux and thus excavation efficiency by implementing workload inequality and retreats, without need for global control.