Meet The Champion Of The Micro Age

Kaigham J. "Ken" Gabriel has a tinkerer's passion for tiny machines. He predicts that microscopic sensors and motors eventually will be as ubiquitous as today's plastics and microchips, showing up in everything from phone switches to "smart" buildings.

If it happens, Gabriel will deserve a lot of the credit. At AT&T Bell Laboratories in the mid-1980s, he led one of the first teams of scientists to build tiny gears and motors from silicon. Now, as head of the micromachine effort at the Pentagon's Advanced Research Projects Agency (ARPA), the boyish, 37-year-old electrical engineer is trying to spread the gospel of the tiny devices to U.S. industry. "It's an unbelievably fun job," he says. "I can really have an effect."

TINY VALVE. Gabriel's operation offers a classic example of how ARPA works. Formerly known as DARPA, for Defense Advanced Research Projects Agency, it was renamed by the Clinton Administration to emphasize its nondefense roles. In general, the agency's technocrats scan the research landscape for developments with key military and commercial applications. Then they fund research and development both to prove the technology and to persuade industry to pick it up. It seems to work: ARPA gets credit for leading the development of everything from chipmaking techniques to massively parallel computers.

In the case of micromachines, Gabriel is spending $24 million in seed money over three years on a two-pronged strategy. One part involves direct grants. A dozen proposals--$4 million worth--have been funded, and an additional $5 million worth will start this month. For instance, ARPA largess has helped Menlo Park (Calif.) startup Redwood Microsystems Inc. fabricate a tiny valve on a silicon chip that can replace bigger, costlier metal valves. One possible use: controlling the flow of samples into a small gas chromatograph that detects poisons in the environment. ARPA also funded the development of a miniature vacuum pump by Westinghouse Electric Corp. and a cluster of tiny acceleration sensors by Draper Labs in Cambridge, Mass. Draper's accelerometers could guide weapons or, if hooked to a car's active suspension system, steady it on sudden stops, starts, and turns.

Gabriel's second thrust is to lower the cost of getting started in micromachines. Chip-fabrication lines are expensive. So last July, Gabriel made a deal with MCNC (Microelectronics Center of North Carolina), which was set up to stimulate high-tech business. The facility will build any scientist's micromachine for $500, with ARPA picking up the rest of the tab. ARPA kicked in about $25,000 for the first 23 projects, ranging from microgyroscopes to miniature wind-speed meters, which came off the line in March.

ACCESS. With that boost, Gabriel hopes, the technology can spread from universities and a few high-tech companies to mainstream industry--from construction to aerospace to makers of mundane valves and thermostats. "Opening access to more people will have a tremendous impact," he predicts.

Gabriel is waging an uphill battle to keep the U.S. ahead in micromachines. He ended up at ARPA largely because Bell Labs axed his research team in 1990--despite his argument that micromachines will be crucial to one of AT&T's main businesses, fiber-optic communications. "I can understand why they did it," he says, noting that AT&T wanted to bring research closer to the market. "But eventually they will have to get back in." As far as Gabriel is concerned, the sooner AT&T and others do that, the better.