The satellite industry could experience its biggest revolution since it joined the ranks of commerce, thanks to some of the smallest machines in existence. Researchers are performing experiments designed to convince the aerospace industry that microelectromechanical systems (MEMS) could open the door to low-cost, high-reliability, mass-produced satellites.
MEMS combine conventional semiconductor electronics with beams, gears, levers, switches, accelerometers, diaphragms, microfluidic thrusters, and heat controllers, all of them microscopic in size [see "A Quick Glance at MEMS"]. "We can do a whole new array of things with MEMS that cannot be done any other way," said Henry Helvajian, a senior scientist with Aerospace Corp., a nonprofit aerospace research and development organization in El Segundo, Calif.
All satellites require a basic set of subsystems to do useful work in orbit [see figure]. A power source is one, typically a solar cell array backed up with batteries. Also necessary are a communications system to receive commands and return information, internal sensors to gauge the satellite's state, and a control unit to coordinate all subsystem activities. Satellites with more complex missions will also require systems to determine the spacecraft's orientation and position and propulsion systems to control both.
If all the functions of these subsystems can be performed by MEMS, then "you could start thinking of fabricating satellite subsystems like we make CMOS chips for laptops," said Helvajian. Satellites could be constructed by stacking wafers covered with MEMS and electrical components. The result would be a "1-kg-class satellite, or picosat, that could be mass produced...you could put up a constellation of hundreds of these little guys in low Earth-orbit [LEO] and get, say, incredible weather information," he continued.
The same was said by Thomas George, who supervises the MEMS Technology Group, part of the Jet Propulsion Laboratory (JPL), in Pasadena, Calif. Such satellites, having negligible mass, size, and power consumption requirements, can be easily piggybacked on conventional satellites or launched using smaller and cheaper launch vehicles, he told IEEE Spectrum.
Some of the advantages of using MEMS-based satellites were spelled out by Ernest Robinson, a distinguished engineer at Aerospace. Low launch costs and high resistance to radiation and vibration headed the list. It costs about US $10 000/kg to put an object into LEO. Clearly, smaller, lighter satellites will cost less to launch. MEMS-based satellites also promise to be cheaper to develop and fabricate than conventional spacecraft.
Robustness is also key. "You can't upset a MEMS switch with a single cosmic ray impact, but you can upset a semiconductor switch by doing that," Robinson explained. This means that MEMS are much more resistant to radiation damage and so would be "able to operate comfortably in a very high-radiation environment such as the Van Allen belts." The Van Allen belts occupy a region of space between LEO and the crowded geosynchronous orbit. Potentially a valuable piece of orbital real estate, it has remained relatively unexploited because of the practical and economic difficulties of shielding conventional satellites from high doses of radiation.
Then there's MEMS resistance to high levels of vibration and the shock of, say, a rocket launch. George referred to a demonstration by Honeywell Corp., of Morristown, N.J., of an imaging array of MEMS-based bolometers (bolometers are used to detect infrared and microwave radiation). The array was tested on the Earth's surface by being shot from a howitzer. Though subjected to accelerations estimated at 20 000 times gravitational acceleration, the bolometers worked perfectly afterward.
The reason MEMS devices stand up to these kinds of forces so much better than conventionally sized devices is clear: the mass of each moving component is extremely small, so that potentially damaging internal forces between components are also very small.