Evolutionary Swarm Robotics: A Theoretical and Methodological Itinerary from Individual Neurocontrollers to Collective Behaviors

In the last decade, swarm robotics gathered much attention in the research community. By drawing inspiration from social insects and other self-organizing systems, it focuses on large robot groups featuring distributed control, adaptation, high robustness, and flexibility. Various reasons lay behind this interest in similar multi-robot systems. Above all, inspiration comes from the observation of social activities, which are based on concepts like division of labor, cooperation, and communication. If societies are organized in such a way in order to be more efficient, then robotic groups also could benefit from similar paradigms. As Kube and Zhang (1993 ) have pointed out, “ Constructing tools from a collection of individuals is not a novel endeavor for humankind. A chain is a collection of links, a rake a collection of tines, and a broom a collection of bristles. Sweeping the sidewalk would certainly be difficult with a single or even a few bristles. Thus there must exist tasks that are easier to accomplish using a collection of robots, rather than just one. ” A multi-robot approach can have many advantages over a single-robot system. First, a monolithic robot able to accomplish various tasks in varying environmental conditions is difficult to design. Moreover, the single-robot approach suffers from the problem that even small failures of the robotic unit may prevent the accomplishment of the whole task. On the contrary, a multi-robot approach can benefit from the parallelism of operation to be more efficient, from the versatility of its multiple, possibly heterogeneous units, and from the inherent redundancy in using multiple agents ( Jones and Matari ć 2006 ). Swarm robotics pushes the cooperative approach to its extreme. It represents a theoretical and methodological approach to the design of “ intelligent ” multi-robot systems inspired by the efficiency and robustness observed in social insects in performing collective tasks ( Bonabeau, Dorigo, and Theraulaz 1999 ). Collective motion in fish, birds, and mammals, as well as collective decisions, synchronization, and social differentiation are examples of collective responses observed in natural swarms (for some recent 7 Evolutionary Swarm Robotics: A Theoretical and Methodological Itinerary from Individual Neurocontrollers to Collective Behaviors