Emergent Intelligence
By ai-depot | August 16, 2002
Looking into biologically inspired models of ant colonies, this article shows how artificial systems can recapture their emergent properties in order to efficiently solve optimisations problems. Theoretical and practical explanations are provided.
Written by Danilo Benzatti.
Introduction
Nature has inspired researchers in many different ways. Airplanes have been designed based on the structures of birds’ wings. Robots have been designed in order to imitate the movements of insects. Resistant materials have been synthesized based on spider webs.
After millions of years of evolution all these species developed incredible solutions for a wide range of problems. Biologically inspired systems have been gaining importance and it is clear that many other ideas can be developed by taking advantage of the examples that Nature offers.
Some social systems in Nature can present an intelligent collective behavior although they are composed by simple individuals with limited capabilities. The intelligent solutions to problems naturally emerge from the self-organization and indirect communication of these individuals. These systems provide important techniques that can be used in the development of distributed artificial intelligent systems.
The collective behavior of ants has been attracting my attention for some time. It’s really interesting and curious how a colony composed by simple individual ants can show such fascinating results. Since I’m interested in pathfinding and memory representation, I always find myself dealing with graph search problems. It is known that the ability of ants in finding the shortest route between the nest and a food source can be used to solve graph problems such as the Traveling Salesman Problem (TSP), so it can be also applied to simple graph search.
During the last few weeks I decided to try a quick experiment using these ideas. The result is a graph search system based on the behavior of a colony of ants which was implemented using multi-agents and genetic algorithms.
By presenting this system, this essay will provide an overview about the use of these different Artificial Intelligence techniques. Although there are many subjects involved, the ants’ colony can be an intuitive and didactic example. This may be useful as a good starting point to learn these techniques or may be also a tutorial on how to implement a system using concepts from different AI fields together.
Tags: none
Category: tutorial |
