Evolutionary Swarm Robotics

Abstract

We have argued that self-organising behaviours represent a viable solution for controlling a swarm robotic system, and that evolutionary robotics techniques are a valuable design tool. There are multiple reasons why self-organisation should be sought for. Among these, the properties of decentralisation, flexibility, robustness and emergence that pertain to self-organising systems and that are highly desirable for a swarm of autonomous robots (see Section 3.3). However, if everything seems to fit in nicely, the problems arise while trying to design a behaviour that can be considered self-organising. In fact, the features that determine the behaviour of a self-organising system are not explicitly coded anywhere, while the design of a control system requires exactly the definition of behavioural rules for each robot of the system. The design problem - treated in detail in Section 4.1-consists in filling the gap between the desired global behaviour of the robotic system and the individual rules that govern each single robot. © 2008 Springer-Verlag Berlin Heidelberg.