New J. Phys. 18, 103005 (2016)

Credit: © ANDRZEJ WOJCICKI / CORBIS DOCUMENTARY / REINHARD DIRSCHERL

The concerted movements of birds and fish are captivating examples of collective motion. A number of models have been developed to describe such behaviour. However, which pattern a given ensemble of self-propelled particles — be they starlings, bacteria or robots — will eventually assume, and the stability of that pattern, remains difficult to predict. Now, Zhao Cheng and colleagues report that various 'pattern phase transitions' can be captured in a model that factors in the vision range of the particles and their tendency to avoid obstacles.

The authors found that for short vision ranges particles move, gas-like, in a widely uncorrelated fashion. Longer-sighted particles get locked into 'crystalline' patterns with identical inter-particle distances and, for even longer-ranging interactions, into liquid patterns with varying, but short distances between constituents. The liquid phase co-exists with various milling phases — stationary circulating patterns (as observed in nature; pictured) that change depending on the particles' obstacle-avoidance tendency, which defines the short-range interactions. Cheng et al. expect that these insights should enable, for example, changing the collective motion of vehicles or robots by simply tweaking a few dynamical parameters.