the researchers no more talk about "Artificial Intelligence" and the new term "Artificial

Life" rules the roost in the recent conferences.

1.1.2 Biological Background

 Swarm Intelligence is biologically inspired by insect societies, where "complex

collective behavior may emerge from interactions among individuals that exhibit simple

behavior." [Bon99, p. 6] Some examples of this collective activity in insects are as listed

below.

* In Leafcutter ants (Atta), chains of ants work cooperatively in bringing cut leaves
 obtained far away, back to the nest.

* Hunting in army ants (Eciton) involves hundreds of thousands of workers.

* Wasps (Parachartergus, Epipona) and Termites (Macrotermes) can build very
 complex nests comprising concentric walls and ventilation ducts.

 1.2 Multiple Agents

 The word "swarms" is used more in biological sense as seen in the papers on insect

societies. But in the literature on artificial life, it is used interchangeably with another

phrase multiple agents. Hence, in this thesis too, although these two words mean the

same, the more appropriate one will be used according to the context.

1.2.1 Basics

 Agents can be defined as man-made entities, which can perform particular tasks.

Agents can be robots, as referred in this work and may also be software programs. The

following are few reasons for the need of multiple agents [KubOO]:

* Multiple agents are necessary to perform few tasks, which are impossible for a
 single robot. Robot soccer and box pushing can be good examples.

* The individual agents may be easier to design and hence, using multiple agents
 may be more cost-efficient than using one complex agent.

* Multiple agents are more flexible and robust compared to a single agent.