First Robot run by living cells
Slime mould used to create first robot run by living cells
Alok Jha, science correspondent
Wednesday February 15, 2006
The Guardian
Ever worried that the terrifying cyborgs that fill sci-fi stories might one day become a reality? Perhaps the latest research by Klaus-Peter Zauner of Southampton University will cause a stir: the engineer has invented a robot that is controlled by living cells.
The cells in question are a specially grown type of "slime mould" that naturally shies away from light.
Dr Zauner grew a star-shaped sample of the slime mould and attached it to a six-legged robot (with each point of the star attached to a leg) to control its movements.
Shining white light on to a section of the single cell organism made it vibrate, changing its thickness. These vibrations were fed into a computer, which then sent signals to move the leg in question. Pointing beams of light at different parts of the slime mould means that different legs move. Do it in an ordered way and the robot will walk.
The work came out of a collaboration with scientists at Kobe University in Japan, who had been studying ways of using biological cells in robots. Dr Zauner himself had been trying to use individual molecules - rather than instructions from computer programs - to control the functions and movements of robots.
"The long-term vision that I have is that this technology we're after is going to be somewhere between living cells and molecules," he said. "Molecules need infrastructure around them to work and so on. The cell provides that infrastructure but provides a lot more complexity."
Biological cells have been integrated into electronic circuits before but only for use in sensors. Dr Zauner's work is the first time that cells have been used to control movement in robots. "What is very attractive to us is the fact that cells can self-repair and self-restructure, all the things that you can't achieve with conventional technology," he said.
Using biological cells provides some autonomy to the robot's movements. "In a conventional computer we specify a program and if the computer doesn't do exactly what we want ... there's an error."
A good analogy for the biological approach is to think of a group of people working together. "They all do their own things in different ways and, to make an effective team, you let them do what they are best at or what they naturally do and try to get them into a directed purpose for your company's goal," Dr Zauner said. "If you apply very rigid rules on them, they become very inefficient. We have the same problem when we approach the molecular scale. If we impose a type of strict programming that you use in computers at the molecular scale, it will not work well."
Autonomy is a useful trait. It is much easier to trap a robot moving according to the rules of a computer program than a living organism. "Robots are much more easily trapped by unexpected contradictory situations," Dr Zauner said. "If you try to put a bug in a box you'll find that they're very clever in overcoming such things and finding a way out, getting beyond their normal behaviour."
Giving robots a window on to biological functions gives them a handle on unpredictable complex environments.
According to Dr Zauner, using biological cells is still on the fringes of robotics research: "We have made the very first step in an interesting new direction."
guardian.co.uk
