MIT model explains how the brain can learn novel tasks while still remembering what it has already learned.
MIT researchers and their colleagues have won three of this year's R&D 100 Awards, which honor the 100 most technologically significant new products and processes. Two of the groups have now won the award two years in a row.
MIT as a whole has won more R&D 100 Awards-15-than any other university since 1963, when the awards began. The closest contender is Johns Hopkins, with six.
Winners are chosen by the editors of R&D Magazine and a panel of more than 75 experts in a variety of disciplines. The award-winning MIT inventions for 1995 and the researchers involved are:
- A device for fast and accurate machining of parts. Developers are Alexander H. Slocum, the Alex and Brit d'Arbeloff Associate Professor of Mechanical Engineering and president of Aesop, Inc.; New Way Machine Components; Weldon Machine Tool, Inc.; Wilbanks International, Inc., and K. Jung, Inc. Professor Slocum and colleagues also won an R&D 100 Award last year.
- A new technology for aircraft surveillance developed by the Lincoln Laboratory Air Traffic Control Surveillance Group.
- A hazardous-waste detector for monitoring smoke stack emissions. The Microwave Plasma Continuous Emissions Monitor was developed by Paul P. Woskov, principal research engineer at the Plasma Fusion Center (PFC); Daniel R. Cohn, PFC division head and senior research scientist; David Y. Rhee, PFC technical staff; Paul Thomas, PFC technical supervisor; Jeffrey E. Surma of Pacific Northwest Laboratories; and Charles H. Titus of T&R Associates. (A story on the detector appeared in Tech Talk on September 27.) Drs. Woskov, Cohn, Titus, and Surma also won an R&D 100 Award last year (Tech Talk, September 28, 1994).
Machine tools, or the machines that shape, cut, drill, grind or polish solid parts, "are the very heart of a manufacturing economy, because all products are made with these tools," said Professor Slocum.
Now he and his colleagues have designed a device that can make machine tools faster and more accurate. It's "the next link in the evolutionary chain for machine tools," they say.
Dubbed the HydroGuideTM, the device is a ceramic platform supported by a very thin film of pressurized water. The part to be machined is attached to the platform; the platform then moves past the machine tool that gives the part its finished shape.
The HydroGuide belongs to a family of devices known as linear hydrostatic bearings. Such bearings provide a number of advantages over other kinds of bearings for machine-tool applications. For example, because there is a film of liquid between the bearing platform and the rail it moves over, there is no mechanical contact between the two. And that means no friction. "Friction introduces inaccuracies in a machined part, because it's not predictable and therefore it doesn't allow [the bearing platform] to move exactly where you command it to move," said Nathan R. Kane, a graduate student in mechanical engineering working with Professor Slocum.
To date, however, hydrostatic bearings have not been extensively used in machine tools. That's because their disadvantages-costliness, for one-outweigh their benefits. "With the HydroGuide, we've made a hydrostatic linear bearing that's much more practical for use in machine tools," said Mr. Kane.
For example, the HydroGuide is less expensive than conventional hydrostatic bearings because of a unique manufacturing process (patent pending) invented by Professor Slocum and colleagues. Further, because it is modular, or composed of standardized parts, it's much easier for designers to incorporate it into a machine tool. (Other hydrostatic bearings must be custom-designed for specific machine tools.)
The HydroGuide is also the only hydrostatic bearing that uses water rather than oil as the liquid between the bearing platform and the rail it moves over. This is environmentally friendly and saves money since it's much more expensive to dispose of used oil.
Finally, the HydroGuide is faster than other hydrostatic bearings, which have a speed limit of about 3/10 of a meter per second. The HydroGuide can operate at up to two meters per second.
The HydroGuide is already proving itself in the workplace. In 1994 Weldon Machine Tool asked Professor Slocum to design the HydroGuide into one of its grinding machines. Nine months later the resulting grinder won a "Best of Show" award at the International Machine Tool Show. "Using the HydroGuide, the cylindrical parts [Weldon] grinds were four times straighter," Mr. Kane said.
Work on the HydroGuide is supported by the National Science Foundation and Professor Slocum's Alex and Britt d'Arbeloff Professorship, as well as the corporate sponsors who were co-winners of the award.
Fifty years ago, MIT research was key to the development of radar for tracking aircraft, so it is appropriate that the Institute has now developed a technology to extend such surveillance into the satellite age.
The new technology takes advantage of the Global Positioning System (GPS), a network of satellites that can accurately determine the position of a given object to within 10 meters. Dubbed GPS-Squitter, the technology determines a plane's position via GPS, then squitters, or broadcasts, that position-plus the plane's identification-to all listeners.
Those listeners include not only air-traffic controllers, but other planes. As a result, it allows aircraft to see each other, a feat that is not practical for all aircraft today because the radar equipment to do so is very expensive. "Now for only a few thousand dollars, you can have a receiver on any plane to pick up squittered information from other planes," said Dr. Steven R. Bussolari, leader of the Lincoln Laboratory group that developed GPS-Squitter.
In addition to being an added safety measure, the ability of planes to see each other could pay off in other ways. For example, currently there is no aircraft surveillance over many ocean flyways because there is no place to put the radar ground station that sends and receives the pulses of energy that determine a plane's position.
"Air traffic controllers keep planes separated over the ocean by sending them off about 100 miles apart," Dr. Bussolari said. "This extra separation ensures they won't get close enough to hit, since the controllers can't track their progress on radar." If planes could see each other, "you could safely pack them in much tighter to really take advantage of the airway."
GPS-Squitter has other advantages over radar. For example, the ground stations that receive the squittered information (not to be confused with the smaller receivers on each plane) are much less expensive than radar ground stations (about $100,000 versus $4-5 million).
Dr. Bussolari noted that GPS-Squitter will also improve surveillance of planes and other vehicles that are on the runway. He explained that a big plane can sometimes shadow a smaller plane, blocking the radar signals that would ordinarily have alerted controllers to the presence of the smaller plane. GPS-Squitter solves this problem because each aircraft sends out its own signals, and GPS-Squitter ground stations are so inexpensive that several could be located around the airport so signals can't be blocked.
Two major demonstrations of GPS-Squitter were conducted in 1994. One tested surface surveillance of aircraft at Logan International Airport; the other tested air surveillance of helicopters servicing oil platforms in the Gulf of Mexico. Both demonstrations were successful, producing performance results that matched theoretical predictions. A third demonstration, to test low-cost collision avoidance receivers for private aircraft, will take place in December in the Los Angeles area.
The work is supported by the Federal Aviation Administration.
A version of this article appeared in MIT Tech Talk on November 8, 1995.