Modelling the emergence of coordinated collective motion by minimizing dissatisfaction

Coordinated collective motion (CCM) has been recently studied using agent-based simulations by applying three behavioural rules: repulsion, attraction and alignment. But these rules are so similar to the expected group behaviour that it can hardly be labelled emergent. We developed an agent-based model that produces CCM using a set of low-level dyadic interaction rules. The agents change their positions with regard to other agents in order to minimize their own dissatisfaction with their inter-individual distances. To test the emergence of CCM, several simulation experiments were performed. The results show that the agents were able to achieve CCM after a few thousand time steps, and that the bigger the area perceived by them, the more coordinated and cohesive the group motion became. An increased memory span and capacity to remember other agents' identities improved cohesion and coordination. The relationship with biological referents is discussed.