10.1 Introduction

The development of a multicellular adult organism from the single egg cell is an amazing process. The complexity of the adult is vastly greater than that of the fertilised egg there are many cells instead of just one, these cells display a range of different behaviours, and yet there is a co-ordination of behaviour to produce an identifiable unified entity, the organism, rather than an unco-ordinated soup of cells. But no new information has been provided during the course of development to allow the creation of complexity from (relative) simplicity.

This self-organisation of relatively simple parts into a complex and functional whole is a process which could be exploited for artificial systems. Computational methods, based on biological development, can be applied to a range of problems and would be particularly well suited to problems which are sufficiently large, complex, unpredictable and distributed where attempts to exert explicit central control are unsatisfactory. In such cases, the ability of local, autonomous cells to continue to function and build an acceptable multicellular whole in the absence of global knowledge can ease or remove the burden of central control. There are many telecommunications problems of this type which represent potential applications of developmentally inspired computing. Network growth, routing of data across networks, control of software agents and their environments, and allocation of resources (hardware, software or human) to maintain system performance could all benefit from technology which is, at some level, inspired by development.