go to HorizonZero HorizonZero 02 vertical line layout graphic français >  

printer friendly version of article  >

readings : transferences: ants & robots
View this article in flash  requires flash 6 >

Ants and Robots
by C. Ronald Kube

Go to the ant, you lazybones;
consider its ways, and be wise.
Without having any chief
or officer or ruler,
it prepares its food in summer,
and gathers its sustenance in harvest.
Proverbs, 6: 6-8

It was the mid-1960's. I was eight years old. I was obsessed by robots. I did not have the knowledge or skills to build the creatures I saw in my mind (or on television), but I did know how to use scissors and glue. My first robot had arms and legs, a body, and a head, all made from cardboard. It wasn't pretty, and it couldn't do anything except stand.

Nonetheless, I was very proud of my robot. It was useless, but it served a purpose: it expressed my desire to create. But that was not enough. Beyond this simple desire lay the urge to create something useful.

Creation, whether artistic or robotic, can always seek inspiration in nature. What fascinated me earliest as an adult scientist was the study of nature's higher examples of intelligence; the exploration of how human intelligence came to be. If this could help us to simulate and eventually build intelligent robots, they might one day wake up and help us to make the world a better place to live and create.

Like most children, I loved looking at ants. Ants are pretty smart as a group. They can build well-structured nests with well-defined shapes, all without some master planner to tell them what to do. How do they do that? Where is their group intelligence?

Each individual ant seems simple at first, but collectively they can accomplish great feats, such as that of group transport. While transporting food back to their nest, some ants push and others seem to pull, sometimes against each other. Still others run back and forth carrying nothing at all. Despite this apparent confusion, their food always makes it back to the nest.

For thirty years science tried to build an intelligent robot without much success. In fact, a simple ant was more successful at navigating its environment than the most sophisticated robot ever built. The smartest robots could follow a night-time garden path by carefully tracking its outside edges with cameras, contrasting its surface against the green grass. But then they would go wandering off course as soon as the sun came out, chasing the hard shadows made by trees. Silly robots.

I was a thirty-one-year-old scientist when my colleagues and I finally remembered the ants. We wondered if a system of many simple robots could somehow be organized like a swarm of ants to accomplish something useful. We decided to build them and test them to see if a collective intelligence would emerge. Synergy might create an intelligence whose sum was greater than its parts.

This time we built our robots out of actuators, motors, and circuits. They didn't have eyes; they had sensors. And, ugly as they were, they worked just like the ants. Our robots were no longer useless: now they could push a box between two places. Like ants, some pushed against each other while some just wandered around. Although the path was never straight, the box always managed to get moved to the right place: a small but real accomplishment.

Some years later, NASA was thinking about building space-based solar power systems on the moon. They flew robot scientists to Washington to spend three days speculating with space scientists about how vast collections of robots might be used to build enormous solar cell arrays. A robot construction crew would mine the raw materials from the lunar surface, transform them into the necessary components, and then assemble them into kilometre-long structures. It was very inspiring.

Hoping to learn their secrets, we still listen to nature and the ants. If we are to discover sources of energy that will serve humanity without waste, we must consider their ways. But if science is to achieve this useful end, perhaps it must also recognize the beauty that art perceives in nature. Together, art and science might someday recreate the intelligence of nature, and understand that what we seek may be found underfoot, with the lowly ant.

Ronald C Kube is an adjunct professor in the Department of Computing Science at the University of Alberta. His research focuses on collective robotics and intelligent systems.

back to top back to top  

 

Valid XHTML 1.0!
Valid CSS!