Self-assembly Mechanisms for Evolutionary Robotics

Abstract

In recent study, an increasing amount of time has been spent researching complex systems due to the traditional Von Neumann architecture lacking sufficient efficiency. A specific complex system which has been growing in popularity is a system of swarm robots. Introducing evolutionary principles in such systems makes an attempt at collective computation, from simple elements, to resolve complex problems. To further add functionality to such systems, self-organizing techniques, such as self-assembly, have been introduced. Because of the lack of research literature within this paradigm, there are few comparisons of the different implementations of self-assembly mechanisms. This thesis outlines some mechanisms of self-assembly and explores the positive and negative implications these may have on evolutionary self-assembling robots. The results of the experiments conducted in this thesis show that self-assembly mechanisms such as the connection hardware and introducing local communication can influence the system's ability to self-assemble, and the behaviour of the self-assembled group.