From robot uprising to rise of the dead — Lessons in engineering and biology from a new type of active particles

Cooperation is vital for the survival of a swarm. No single bird is faster than a jet plane, and no single fish is faster than a speed boat — humans beat individual animals in air, land, and sea. But, when animals cooperate and swarm, they beat us since biblical times. The technological gap in engineering artificial swarms suggests a gap in our understanding of emergent behaviors in decentralized, distributed systems. Inspired by non-equilibrium statistical mechanics I will shed light on a hallmark of agile cooperation in nature — cooperative transport. Starting from the microscopic description of mechanically self-propelled particles, I will show that particles are expected to rotate by external forces. Coarse graining of the microscopic dynamics will lead to a set of equations of motion of Force Aligning Active Brownian Particles, which extends and generalizes existing models. Force-alignment is captured by a charge-like parameter with units of curvature that sets the tendency of the particle to align with or against an external force. Particles with a negative “active-charge” will turn in opposition and move against an external force or a repulsive potential. I will then argue that this minimal description can capture geotaxis — a widespread phenotype in individual insects that tend to climb uphill (i.e. move against the external force). I will then proceed to show how a swarm of Force Aligning Active Brownian Particles, spontaneously commence in the cooperative transport of a larger payload. Surprisingly, in both experiments in robotic swarms and numerical simulations, cooperative transport improves with increasing payload size. Using the effective equations of motion I will derive a non-linear dynamical system where cooperative transport is expected as a spontaneous symmetry breaking. I will present the resulting geometrical criterion for cooperative transport as the interplay of the payload’s geometrical curvature, and the “active- charge” of the self-propelled particles. Our findings offer new design rules for distributed robotic systems and shed light on cooperation in natural swarms.