Magazine article Science News

Termitelike Robots Build Structures: Simple Guidelines Keep Machines Hauling and Placing Bricks

Magazine article Science News

Termitelike Robots Build Structures: Simple Guidelines Keep Machines Hauling and Placing Bricks

Article excerpt

Human construction crews, meet Robo-Termites. Like the mound-building insects, these squat little robots can erect complicated structures without an instruction manual.

Using just a few preprogrammed rules, some traffic laws and a stack of foam bricks, the bots get busy building towers, pyramids and fortress-like walls, Harvard computer scientist Justin Werfel and colleagues report in the Feb. 14 Science.

One day, such robotic builders might be able to take on risky human jobs such as building sandbag levees during floods. And like a termite or ant colony that gets stepped on, Werfel says, "it doesn't matter if some of the robots are lost--the rest will keep going."

Termites are the architects of the insect world. The itty bitty animals paste together dirt and chewed-up bits of wood to build mansion-style mounds, with open chimneys and labyrinths of snaking tunnels.

But perhaps most impressive, Werfel says, is the fact that termites can build at all. "It's frankly amazing," he says. "There's no central brain assigning them tasks, no one coordinating what they do." Instead, millions of insects working side by side each decide their next moves for themselves (SN: 5/9/09, p. 16).

Werfel and colleagues created computer programs that let robot construction crews act like independent insects. One program converted a building's blueprints into "traffic laws" for roaming robots to follow. The laws told the robots where they could move from any spot within the construction site, a checkerboard-like grid.

Another program gave the bots some basic guidelines to prevent gridlock: Always cruise counterclockwise around the structure, for example, and lay bricks one after another along a row, instead of skipping around. …

Search by... Author
Show... All Results Primary Sources Peer-reviewed

Oops!

An unknown error has occurred. Please click the button below to reload the page. If the problem persists, please try again in a little while.