The concept relies on packaging micro-electromechanical systems, devices with micron-scale features dozens of times narrower than a human hair. An optimal sensor package would take up about 100 cubic microns, said John Manobianco, a staff scientist with the aerospace sciences and engineering division of Ensco, a systems integration and research company.
"One dimension or more of the probe might be 100 microns, or at that scale, but we might have to make a portion of it slightly larger," Manobianco said from his Cocoa Beach office. "I try to think now along the lines of military (radar-reflecting material), which is about 2 centimeters long but only microns thick."
As a cloud of such probes dispersed on wind currents inside a storm, they would measure atmospheric pressure, temperature, humidity and other factors with real-time, 3-D resolution not possible with radar or satellite sensors, Manobianco told United Press International.
Computer models show the sensors would stay aloft for days if released from several miles up in Earth's atmosphere, he said, and other missions are possible.
"(NASA's) planetary exploration people can envision, if we can make the probes small enough, that this would be a great payload for a ... satellite expedition to a remote planet," Manobianco said. "Some members of the military are interested in battlefield surveillance and reconnaissance with this technology."
The next step in developing Global Environmental MEMS Sensors is building about 1,000 of them for systems integration and network communications tests, Manobianco said. Meteorological applications could be available in about a decade, he added, while military versions would arrive sooner.
The very ambitious concept faces daunting technical challenges, said Tom Haddock, vice president of technology commercialization at Ardesta, a research and venture capital company in Ann Arbor, Mich., that specializes in MEMS and other "small tech" applications.
Machining features at the micron scale is done regularly, he told UPI via telephone, but systems integration issues, such as self-contained power sources, will demand a lot of work.
Initial GEMS designs include batteries, Manobianco said, but solar power and even miniaturized fuel cells -- which generate electricity as hydrogen and oxygen combines to form water -- are candidates for later iterations.
Regardless of the power source, another issue is providing communications between the sensors themselves and whatever system collects the data, Haddock said. The sensors' tumbling in wind or cloud turbulence would prevent "pointing" the devices at each other, so they would have to generate a strong radio signal in all directions to pass along their information, he said.
MEMS devices are produced in large quantities using processes that also create computer chips, lowering the per-device cost, Haddock said. If GEMS designs ease integration challenges by including measurements tens or hundreds of times larger than 100 microns, however, their cost will rise accordingly, he said.
Global Positioning System satellites could provide the sensors with necessary highly accurate location information, but designers have yet to include an entire GPS system on a computer chip, said Paul Ostdiek, technology manager of the engineering and technology branch at Johns Hopkins University's Applied Physics Laboratory in Laurel, Md.
Despite the challenges, applications such as monitoring a single thunderstorm could be developed with enough resources, Ostdiek told UPI via telephone. Having such a lofty goal as GEMS is very helpful in encouraging research, he added.
"(The sensor) is basically a 100-cubic-micron spacecraft without any propulsion," Ostdiek said. "The meteorological folks would love this ... replacing all their field work (gathering data) with 'smart dust' is a very powerful concept."
(Reported by UPI Science and Technology Writer Scott Burnell in Washington.)