Essay On Swarm Robotics

Gerardo Beni was in 1989 one of the first who used the term “swarm intelligence”, he used it to denote a class of cellular robotic systems (Beni, 2005). The term then was embraced more by the social insect studies and by optimisation studies that used the social insect metaphor, and was losing much of its original robotics context (Bonabeau et al., 1999). The term “swarm robotics” emerged in the recent years as the application of swarm intelligence to multi-robot systems, with emphases on physical embodiment of the entities and realistic interactions among the entities and between the entities and the environment (S ̧ahin, 2005) .

A swarm can be defined as a collection of interacting organisms, the therm swarm-

11

Chapter 2. Background

ing is in particular used for a collection of insects but can also be used for other organisms that exhibit swam behaviour. Swarm behaviour is the collective behaviour of similar size and type of organisms collaborating together, although more specific names are used for different kinds of animals. For example, a collective of birds is called a flock (Figure 2.1a), fish swim in schools (Figure 2.1b), and the therm herd is generally applied to grazing mammals (Figure 2.1c). The individual organisms in these swarms are

Particle swarm optimisation (Kennedy and Eberhart, 2001) was first intended to simulate the movements, or the social behaviour of a bird flock or a fish school. The original algorithm was adapted and evolved to a simple and efficient optimisa- tion algorithm which uses particles for a population based stochastic optimisation technique. There are several other optimisation algorithms inspired on nature, like glowworm swarm optimisation (Krishnanand and Ghose, 2005) and intelligent water drops (Shah-Hosseini, 2009), but the currently the most promising ones are based on the foraging behaviour of social insects like ants and