It took researchers in the field of optimization and robotics quite a lot of time to recognize that a decentralized, but internally directed search can yield better optimization when dealing with large solution search spaces. Whereas genetic algorithms perform a highly parallel and relatively mostly undirected search only being slightly nudged in the right direction by a crossover operator that favours "better" search results, particle swarm optimization and other methods associated with it (Ant Colony optimization, Swarm Robotics, etc.) rely on the ingenuity of the agent/ant designer to provide a set of searching/traversing directives to the large number of "swarming" particles in order to get optimal results. Within the articles presented here, we will discuss the necessary paradigm shift for swarm agent/ant/particle design necessary to researchers in this field, and how the interesting nature of these swarms can then be used to create emergent behaviour in computer games.  

Ant Colony Optimization, Swarming Behaviours, Swarm Intelligence and other particle-swarm based models useful in game development are discussed here.