Information-Driven Collective Behavior in Active Matter Systems

When numerous objects with internal dissipation are considered together, they form what is known as active matter—a material that can exhibit unusual bulk properties resulting from broken symmetries at the (perhaps microscopic) scale of individual agents. A common model of such systems is the flock, or a group of self-propelled agents that exert aligning torques on each other (i.e. ‘flying spins’). However, many natural examples of flocks do not interact appreciably by exerting forces on each other, but instead by the exchange of information. For example, neighboring birds measure the orientation of their peers and choose a new heading for themselves.



Is a flock with such feedback control fundamentally different from a collection of ‘flying spins’? I will discuss an active matter model which isolates the influence of information on collective properties by prohibiting conventional sources of active work (e.g. self-propulsion) and serves as a many-body extension of the Maxwell demon thought experiment. This system converts local (uncorrelated) fluctuations into collective action at large scales though measurement and feedback control of agent-agent scattering cross-sections. Despite enforcing conservation of energy and momentum in every agent collision, scattering feedback control enables non-equilibrium behaviors such as propulsion, pattern formation, and spontaneous symmetry breaking. Local measurements can drive global alignment of a nematic flocking order parameter and hence the emergence of an anisotropic pressure tensor in the gas phase.



Using kinetic, hydrodynamic, and information theories as well as numerical simulation it will be shown that non-equilibrium effects scale with the magnitude of noise (both thermal and non-thermal), leading to pattern formation phenomena which do not degrade under nosier conditions. This model suggests that the ‘informational’ aspect of flocks and other complex active matter can play a consequential role in overall behavior, particularly for systems subjected to large magnitude external perturbations.