Collective Complexity out of Individual Simplicity

The concept of Swarm Intelligence (SI) was first introduced by Gerardo Beni, Suzanne Hackwood, and Jing Wang in 1989 when they were investigating the properties of simulated, self-organizing agents in the framework of cellular robotic systems [Beni89]. Eric Bonabeau, Marco Dorigo, and Guy Theraulaz extend the restrictive context of this early work to include “any attempt to design algorithms or distributed problem-solving devices inspired by the collective behavior of social insect colonies”, such as ants, termites, bees, wasps, “and other animal societies.” The abilities of such systems appear to transcend the abilities of the constituent individuals. In most biological cases studied so far, the robust and capable high-level group behavior has been found to be mediated by nothing more than a small set of simple low-level interactions between individuals, and between individuals and the environment. The SI approach therefore emphasizes parallelism, distributedness, and exploitation of direct (agent-to-agent) or indirect (via the environment) local interactions among relatively simple agents. The title, Swarm Intelligence: From Natural to Artificial Systems, summarizes the content and the structure of the book well: each of the central chapters starts by presenting experimental results of one or several biological studies (foraging, division of labor, clustering and sorting, nest building, cooperative transportation), then describes a model for explaining these results, and moves on to discuss engineering outcomes in the form of algorithms or collective robotic systems that have or could have been inspired by the biological examples. One of the strengths of this monograph is undoubtedly this extensive use of models as a quantitative and abstract interface for the implementation of natural principles in artificial systems. Without an adequate level of description provided by models, it would be difficult, if not impossible, to understand the collective behavior of natural systems or to explore parametric ranges, which can be of interest for engineering purposes, though not necessarily useful in Nature. While the biological examples and their related modeling are well-structured, the book tends to be less systematic in the myriad of engineering case studies that are associated with them. The global picture is far from being exhaustive and unitary: the book suffers from the current youth and rapidly developing nature of the SI field. The field currently lacks mature and sound methodologies to transfer biological mechanisms into useful engineering algorithms or to choose an adequate level of description for modeling. Bioinspiration is a process fully dominated by the intuition and imagination of a few researchers. A full body of theory for designing and describing such distributed systems is still missing.