Page Text

ANOTHER WAY OF GETTING TOGETHER

text alex MArishian photography bjarne winkler

Understanding the behaviour of SWARMS could lead to some of this century’s biggest technological breakthroughs.

he Danes call it “Black Sun”.

Each spring, more than a million starlings pass through Denmark’s Tønder Marsh on their northbound journey home. In the early evenings, they swarm together in vast numbers, producing voluminous black forms of staggering complexity in the skies overhead. Although a swarm may consist of tens of thousands

of individual birds, the overall effect is of a single super-organism, whimsically forming and fading in the hour before sunset. While the so-called Black Suns are a vivid illustration of the phenomenon, starlings are by no means alone in exhibiting the kind of highly coordinated, self-organising group behaviour

scientists refer to as “swarm

intelligence”. Schools of f i sh, hives of bees and especially colonies of ants have all come under the gaze of researchers in the fast-growing, interdisciplinary f i eld of swarm theory. Using powerful computers to model the

behaviour of swarms, these researchers are applying their f i ndings to the study of a wide range of phenomena, from traff i c jams to mobile communications networks to the mechanisms of the human immune system. Swarm theory, they say, will help drive some of the 21st century’s most critical technologies. Dr Christian Jacob, director of the Evolutionary and Swarm Design (ESD) research group at the University of Calgary, uses the human brain as the starting point for understanding the intelligence of swarms. “A neuron seems to be a rather simple processing device,” says Jacob, “much simpler than the processors we use in our

computers.” But while a computer typically features only a small number of processors, the human brain is a network of literally billions of neurons. “These neurons are not even synchronised by a centralised clock, as is the case of computers. Yet when it comes to vision, language or pattern recognition, our brains’

processing capabilities by far outperform computers.” In much the same way that the brain’s work is distributed across a vast, decentralised network of neurons, the complex intelligence of the swarm is distributed across a vast, decentralised network of organisms. One of the many implications for the theory is in

the f i eld of robotics, where

126 Another Magazine

researchers around the world are investigating whether a decentralised network of relatively simple robots could achieve an intelligence greater than the sum of its parts. Imagine a swarm of self-organising micro-robots

travelling through the body, adapting and readapting itself to a range of tasks, from unclogging arteries to performing surgery. It may sound like science f i ction, but for Professor Jacob, it’s just 10 to 20 years away. Indeed, the I-SWARM project, a consortium of 10 European research institutes, aims to produce

a thousand millimetre-sized robots capable of acting together in a collaborative swarm, by 2008. For all its potential applications, Professor Jacob is quick to note that swarm intelligence is not just of the future. “Looking at our world through the eyes of swarm intelligence,” he observes, “one can recognise that everything

we see, hear, touch is the result of emergent effects from interactions on a lower level. Our cells determine what our body looks like. The building blocks encoded in the cells’ genes determine what they do. Our human interactions determine how society progresses (or not). I tell my students that the world is one

big swarm system .”