By H. Sayama
IEEE Computational Intelligence Magazine, 5.3 (2010): 43-49.
In this paper, Sayama introduces the concept of “swarm chemistry”, so to have swarms of situated agents divided in subsets. A different “recipe” (as collection of parameters regarding interaction rules among the components of the swarm) is assigned to different subsets of the swarm. Sayama demonstrates how mixing recipes allows the formation of shapes or notable patterns; by taking into account that each member of the swarm does not recognize whether its neighbouring components are instructed with the same parameters set or if they feature a different recipe, it is very interesting to note how awareness can emerge as a whole.