The Shape Of Robots To Come?
Call him the Pablo Picasso of robotics. Joseph Michael's world, after all, is cubist. By patching together robotic cubes of different sizes, the founder of Robodyne Cybernetics Ltd. in London envisions new families of smart machines for a host of jobs--from fighting fires and inspecting the inside of pipes in chemical plants to building structures in space or products in factories. That includes cloning themselves, manufacturing new generations that would become ever cheaper and smaller--tiny enough, ultimately, to battle disease inside the human body.
Mixing sizes of cubes can yield endless variations. Some resemble insects, others look like forklifts. And the clusters can morph when the cubes slide over and around one another, using grooves on their surfaces. Each robot cube would get its own microchip brain, a communications system for talking to the other cubes in a cluster, and a motor. Most would have ports for attaching mechanical fingers, tools, or wheels.
Michael calls the concept "fractal shape-changing robotics." He has been honing it since high school, after his family emigrated to Britain from India 27 years ago. He got serious about the idea after graduating from the University of Hull in 1984 and filed for his first patent in 1994. Since then, he has plowed every spare penny into his robots--close to $200,000 in total.
An independent software engineer, Michael says he makes "a very good middle-class living" writing programs. But you would hardly know it from the two houses in North London where he lives with his parents and a brother. Machine tools, motors, and other robot parts litter seven rooms and a basement workshop, along with two dozen prototype cubes ranging in size from 10 inches per edge down to 5 inches.
TEXANS INTERESTED. Marketing, though, has gotten short shrift. So while Michael has won prizes at six industrial competitions since 1995--most recently at last May's Invention/New Products Exposition show in Pittsburgh--his work is not well known. Robot experts who have toured Robodyne's Web site (www.robodyne.com) point out that similar, so-called reconfigurable robots capable of repairing themselves, have become a hot topic, and some say Michael brings little new to the party. But Jing Xiao, director of the National Science Foundation's robotics program, says she is "quite impressed" by the versatility of Michael's fractal approach.
The market may soon resolve the disagreement. NanoTechnology Development Corp. (NTDC), a Houston startup, has purchased the first fractal-robot license. Its initial goal is a bridge-building system for use in the wake of natural disasters. A bunch of big cubes would be trucked to a collapsed bridge. The robots would "crawl" across the river or the caved-in section of the bridge or a landslide (table), then lock together to form a rigid structure.
Bridge-building is just for starters, though, says NTDC President Michael S. Forbes. The ultimate payoff will come when the robots are outfitted with tools and installed in factories to make smaller copies of themselves. A factory with 1-cm workers, say, would crank out 1-mm clones until there were enough of them to create a duplicate, but smaller, factory for making 0.1-mm robots. The final result would be 1-micron cubes, just a tiny fraction of the width of a human hair.
Clusters of such robots could be programmed to navigate blood vessels and excise cancer cells, says Michael D. McDonald, NTDC's chief technology officer and head of advanced research at Access Multimedia Technology. Cheap robot factories the size of a grain of sand would also serve up a cornucopia of goods. "Anything you want, you'll be able to replicate" at home, he adds.
McDonald is no wet-behind-the-ears idealist. He was Control Data Corp.'s man in charge of the supercomputers at NASA's Houston facility in the late 1960s when NASA was racing to fulfill President John F. Kennedy's pledge to put a man on the moon. He was lured out of retirement by the promise of nanotechnology--building almost anything by manipulating individual atoms.
A key champion of nanotech is K. Eric Drexler, chairman of the Foresight Institute in Palo Alto, Calif. He envisions an atomic-scale "assembler" even smaller than Michael's 1-micron cubes. While such a robot is still in the future, Michael says his scheme can be launched today, since many applications don't need the ultimate 1-micron cubes.
INVESTORS SOUGHT. For example, furniture, appliances, and even cars might be designed faster with the help of fractal 1-mm cubes. When managers need to see a physical model, the computer-aided design system could send the contours to a car-size cluster of cubes, which would shift to reproduce the design.
But such dreams hinge on shaping NTDC into a moneymaker. So far, it has not gotten past square one. The company needs to raise $300,000 to fund its first cube factory. NTDC made a stock offering on the Internet last summer--up to 200,000 shares at $5 each. A mere $5,000 has been collected, admits Forbes. Partly that's because the shares are registered for sale only in New York State. Forbes, who has sunk almost $100,000 of his own money into the project, expects approval soon to expand the online offering to other states.
If NTDC can't raise the funds, Michael says he will take his vision to Japan. If that fails, he vows to scrape together $75,000 to construct a bridge-building prototype himself. Despite the setbacks, Michael still sees a cubist future.