Swarm Intelligence

Swarm intelligence is the process used by groups of organisms or systems or agents to self organize into collective behavior that exhibits some level of intelligent activity. Swarms of insects often seem to act as an intelligent whole, even though the actual intelligence level of each bug is observed to be quite low. Most versions of swarm intelligence seem to involve some form of network communication and some form of simple rules for decisions, like the rules in cellular automata. The simple rules, when applied to many agents and magnified by “ripple” effect communication can create quite complex behavior with intelligent results for the group.

Dumb parts, properly connected into a swarm, yield smart results. – Kevin Kelly

The Genius of Swarms – [nationalgeographic.com]

“Ants aren’t smart,” Gordon says. “Ant colonies are.” A colony can solve problems unthinkable for individual ants, such as finding the shortest path to the best food source, allocating workers to different tasks, or defending a territory from neighbors. As individuals, ants might be tiny dummies, but as colonies they respond quickly and effectively to their environment. They do it with something called swarm intelligence.

Riders on a swarm – [economist.com]

Digital ants and birds, then, are good at thinking up solutions to problems, but Dr Dorigo is now working on something that can act as well as think: robots. A swarm of small, cheap robots can achieve through co-operation the same results as individual big, expensive robots—and with more flexibility and robustness; if one robot goes down, the swarm keeps going. Later this summer, he will be ready to demonstrate his “Swarmanoid” project. This is based on three sorts of small, simple robot, each with a different function, that co-operate in exploring an environment. Eye-bots take a look around and locate interesting objects. Foot-bots then give hand-bots a ride to places identified by the eye-bots. The hand-bots pick up the objects of interest. And they all run home.

All this is done without any pre-existing plan or central co-ordination. It relies on interactions between individual robots. According to Dr Dorigo, bot-swarms like this could be used for surveillance and rescue—for example, locating survivors and retrieving valuable goods during a fire.

SEE ALSO:
Self Organizing
Self Assembly
Mote Net Computing
Utility Fog
Smart Dust
Cellular Automata

Comments are closed.