Interacting Swarm Sensing and Stabilization

TL;DR

This paper develops a theoretical framework to predict when colliding swarms of mobile agents form stable milling states, supported by experiments with micro UAVs, advancing understanding of swarm interactions.

Contribution

It introduces a novel analytical approach to determine the parameters leading to stable milling in colliding swarms, based on a limit-cycle oscillation model.

Findings

Predicted critical interaction coupling for stable milling.

Identified saddle-node bifurcation as the underlying mechanism.

Experimental validation with micro UAVs shows agreement with theory.

Abstract

Swarming behavior, where coherent motion emerges from the interactions of many mobile agents, is ubiquitous in physics and biology. Moreover, there are many efforts to replicate swarming dynamics in mobile robotic systems which take inspiration from natural swarms. In particular, understanding how swarms come apart, change their behavior, and interact with other swarms is a research direction of special interest to the robotics and defense communities. Here we develop a theoretical approach that can be used to predict the parameters under which colliding swarms form a stable milling state. Our analytical methods rely on the assumption that, upon collision, two swarms oscillate near a limit-cycle, where each swarm rotates around the other while maintaining an approximately constant density. Using our methods, we are able to predict the critical swarm-swarm interaction coupling (below which two colliding swarms merely scatter) for nearly aligned collisions as a function of physical swarm parameters. We show that the critical coupling corresponds to a saddle-node bifurcation of a limit-cycle in the constant-density approximation. Finally, we show preliminary results from experiments in which two swarms of micro UAVs collide and form a milling state, which is in general agreement with our theory.