The Robot Evolution

MIT's Rodney A. Brooks is among researchers leading the charge to develop a smarter and more useful artificial creature.

The manufacturing industry is no stranger to robots.

Huge robot arms are commonplace in several industrial settings -- particularly automotive -- and primarily engage in long-run, repetitive tasks such as welding and assembly. They operate in splendid seclusion, with human employees urged to keep their distance from the immensely strong machines.

Then there are the intelligent robots of science-fiction movies and books, such as C3PO and R2D2 from the Star Wars movies, which seem almost human in their ability to reason and feel and interact with human beings.

In his latest book, "Flesh and Machines: How Robots Will Change Us" (2002, Pantheon Books), Rodney A. Brooks, director of the Artificial Intelligence Laboratory at the Massachusetts Institute of Technology (MIT) in Cambridge, suggests that the "science fiction fantasy," as he calls it, is not so far off.

"My thesis is that in just 20 years the boundary between fantasy and reality will be rent asunder," the native of Australia writes. "Just five years from now that boundary will be breached in ways that are as unimaginable to most people today as daily use of the World Wide Web was 10 years ago."

The implications of that breach for manufacturing are widespread, although significant challenges remain.

If anybody's predictions about the future of robotics are worth serious consideration, Brooks' are. He has dedicated his life to building intelligent robots and has a wealth of know-how to show for it. For example, Brooks, who also is the Fujitsu professor of computer science at MIT, built, along with his students, two prototypical humanoid robots called Cog and Kismet that are said to be capable of learning from experience.

He is co-founder, chairman and CTO of iRobot Corp., which manufactures robots and supplies developers with an industry-standard platform to create robot applications. The Somerville, Mass.-based company has received a lot of press recently. Earlier this year it introduced consumers to Roomba, a robotic vacuum cleaner that is both self-propelled and self-navigating. In mid-September a robot dubbed the Pyramid Robot and developed by iRobot, crawled through a narrow shaft in the Great Pyramid outside of Cairo and, on live TV, drilled a hole through a door it encountered and fed through it a tiny camera, revealing yet another door.

Brooks also has authored some 40 scientific articles and four books.

All that he has done to date, however, has not diminished Brooks' appetite to further push the boundaries of robotic development. In his latest tome, Brooks' says that in principle there is no reason why it should not be possible to build "a machine from silicon and steel that has both genuine emotions and consciousness," given both the fact that humans have emotions and his belief that humans are machines.

What may be missing are the mathematical techniques to make that leap, he surmises.

But what could increasingly intelligent robots mean to manufacturing? Potentially a lot -- someday. Brooks believes there will be imperatives from manufacturers for smarter machines -- robots that take instruction easily rather than ones that require hours of programming time; robots that can help in small-batch operations rather than ones that only make financial sense in continuous or nearly continuous fixed-automation settings with long runs.

More broadly, robots with social interaction skills could ultimately change the way production lines work.

Over-riding, or possibly propelling, the other imperatives for intelligent robots on the plant floor will be a drive to reduce costs. Brooks points out that many firms rely on low-cost manufacturing in underdeveloped countries. Those low-cost manufacturing locations keep changing as the introduction of industry slowly raises the standard of living.

"We might eventually run out of places to 'exploit,'" he says, which will give rise to calls for new low-cost solutions. (No insult to manufacturers is intended by the term 'exploit,' says Brooks, who notes that his own company manufactures in China to take advantage of its lower costs.)

The good news, he says, is that consumer robots "are starting to have traction."

It may be in that arena that technological advances occur that lower costs enough to bring smart robots into the manufacturing sector.

There are also significant challenges ahead in the development of smarter robots, which may help delay them making a significant impact on the manufacturing floor in the near future. For example, while computer vision is good at certain tasks, there also are many things it is not particularly good at, such as general object recognition.

"The truth of the matter is that we have no computer vision system that is at all good at recognizing that something is a cup, or a comb, or computer screen," Brooks says in "Flesh and Machines." "Our computer vision systems can do a few things with great skill, but still after forty years of effort they are not good at the things we human and many animals do effortlessly."

Secondly, robots lack the dexterity of the human hand, a primary ingredient in the types of manufacturing that have moved to low-cost locations.

"What we need is low-cost dexterous manipulation," Brooks says. "Right now we don't even have high-cost dexterous manipulation."

When will those challenges be met? "Not in two years or three years. But is it going to be there in 30 years or 40 years? I'm not quite prepared [to make a prediction]," Brooks says.

Hide comments

Comments

  • Allowed HTML tags: <em> <strong> <blockquote> <br> <p>

Plain text

  • No HTML tags allowed.
  • Web page addresses and e-mail addresses turn into links automatically.
  • Lines and paragraphs break automatically.
Publish