Prepare for invasion of chatterboxes

November 17, 2005, vol. 34, no. 6
By Barry Shell



Document Tools

Print This Article

E-mail This Page

Font Size
S      M      L      XL

Related Stories

One day, when Richard Vaughan was 10, he took a break from building lego models to laze in the sun on the lawn with a cat. “The cat was looking at me and I wondered how I appeared from the cat's point of view. Big, I supposed.”

Vaughan knew that cat vision was different. They are very sensitive to motion and they don't see reds well. “I tried very hard to imagine what it would be like to see that way and also what it would be like to have whiskers and to be able to steer my ears.” He wanted to know how cats worked.

Now Vaughan is an assistant professor of computing science at SFU and he builds robots for a living. Just as he did with the cat, he wonders how a robot perceives its environment and in particular how it relates to other robots. Everything Vaughan does with robots has to do with interactions. “One robot is not interesting to me,” says Vaughan who created the SFU autonomy lab in the school of computing science to investigate long-lived robot swarms. Funded by the Natural Sciences and Engineering Research Council (NSERC) and SFU grants, work began in January 2004 shortly after Vaughan arrived at SFU and by spring 2006, 30 little robots called chatterboxes will be running around the place. Vaughan wants to find out how they should behave. For example, when do they need to recharge their batteries?

The $1,000 critters are 12 centimetre cubes that can see with infra-red eyes and communicate with flashing lights and sound - hence the name chatterbox.

Up until now, Vaughan and his graduate students have been simulating robot swarms with software called Player/Stage, an open source robot control and simulation package Vaughan co-authored. Player is probably the most widely used robot control interface in the world.

Vaughan's robots are designed to work together on projects with no human intervention. One of their first jobs will be to map the TASC1 building where they live. They will wander all over the place, taking measurements, but above all they will have to remember to get recharged. Two large mother robots will deliver electricity to the chatterboxes. Vaughan wonders, “Should a chatterbox running low on power madly drive around looking for a mother, or should it just stand still and scream for a fill up? Should it compete with other robots for mother's attention? Should it be aggressive?”

Even though he is a computer scientist, Vaughan does not expect his robots to rely only on logic. “Animals - including humans - do not act in optimal ways,” says Vaughan, “but they often act as if they are optimizing things.” Vaughan looks to animal behaviour for robot programming ideas and he finds aggression to be very useful - especially aggressive display behaviour. Ultimately, he hopes his swarm of chatterboxes may be able to return the favour. He believes one day biologists will observe robot interactions to understand how animals work.

Search SFU News Online