ALBUQUERQUE,N.M.On Sept. 11, 2001, a think tank at Sandia National Laboratories started assessing its advanced technologies against the threat of terrorism. In response to that study, Sandia recently funded a program to counter terrorist threats with microelectromechanical systems (MEMS) not only to sense dangerous chemical, biological and nuclear agents, but also to identify and track the danger itself the terrorists.
"There is one viewpoint that asserts our best leverage is to deploy MEMS sensors to detect chemical, biological and nuclear threats at the locations where they might be used, like airports. But a second viewpoint asserts we have to consider our experience in Afghanistan where we aggressively went after the terrorists themselves how could microlevel sensors help there?" said team leader and Sandia director Dave Nokes.
Sandia's research and development office has earmarked $2.5 million in 2002 for its three-year-long study aimed at countering terrorism, using advanced MEMS and related sensor technologies. By capitalizing on Sandia's expertise in MEMS sensors, power sources, computing, robotics and systems integration, the Nokes team aims to quickly put together a real-world test bed that measures the specific effectiveness of Sandia's different MEMS sensors against real-world terrorist threats.
"Our goal is to fight the current war on terrorism, rather than develop technologies for a war we might fight five years from now," Nokes said. "We will succeed or fail on the shoulders of our EEs they are the ones that will make these sensor systems work or not, plus software and communications we believe that smart sensors communicating among themselves will be critical," Nokes said.
Nokes is calling in an all-star external advisory panel to help him decide how to split the program's monies between offensive and defensive technologies.
Defensive sensor nets will likely be deployed regardless of Nokes' decision. On the battlefield, small wireless MEMS sensors "smart dust" will be dropped from the air to monitor enemy activities.
On the home front, airports are already deploying better technologies to detect explosives. MEMS sensors could make an important contribution to those systems by detecting chemical, biological and nuclear agents. When they don't prevent terrorism, defensive measures also seek to minimize damage, such as new building codes that will require sensors to automatically trigger the pumping of poisoned air out of buildings.
A hundred years ago, fire terrorized people and "loss of life was enormous," said Gerry Yonas, vice president and senior scientist at Sandia's Advanced Concept Group. "So how did we learn to live with unremitting terror? We did it by making investments in technology building codes make sure that we have automatic sprinklers and use fire-resistant materials." Fire still kills about 4,000 U.S. residents each year and 89,000 firemen are injured, he said, "but we're not terrified by fire anymore."
The fire threat is still real, but measures to prevent it or quickly respond when it does happen have taken away the terror factor, he pointed out. "Likewise, we want to restore the sense of safety and security that Americans had before Sept. 11, by turning terrorism into a psychosocial problem similar to fire."
The real challenge to defensive sensor nets, Nokes said, is not in distributing units but in harnessing them to prevent terrorism. Monitoring a few expensive sensor systems to prevent terrorism, such as baggage-inspection machines, is simple because security personnel are standing by to intervene. But remotely monitoring hundreds of inexpensive, distributed MEMS sensors, and responding appropriately, involves system integration expertise that does not yet exist. It also lacks appropriate infrastructure like "anti-terrorist" building codes.
"We need to learn how to mix together our expensive sensor installations with these new networks of inexpensive distributed MEMS sensors to make these ideas work," Nokes said.
Mitigating terrorism with defensive systems that monitor hazardous materials in public places will be difficult, Nokes said, requiring an infrastructure similar to local firehouses. Real people, not computer systems, respond to fires. Likewise, if defensive sensor nets are to be effective at mitigating terrorism, then security and medical teams will have to be constantly standing by to counter real terrorists when they decide to act.
Offensive sensor nets, by contrast, could identify and track individual terrorists, so that security teams could pick where and when to arrest them hopefully, before the terrorists enter public places with their weapons drawn. The ideal system, Nokes said, would integrate the existing infrastructure of ubiquitous surveillance from cameras in public places to satellite imagery with MEMS sensors.
In an ideal scenario, for instance, a face-recognition program running in a drive-up automated teller machine might identify a terrorist, not only triggering a paper trail of that person's transactions, but also tracking the suspect's physical movements with smart dust or even a MEMS "bug" in the suspect's transaction receipt.
The real-world test
"We can build many different types of new sensors with our MEMS technologies, such as optical sensors with adjustable diffraction gratings, but there is a lot of folklore about which kinds will work the best. What our program will do is actually build and characterize their effectiveness in real-world tests," Nokes said.
Sandia has MEMS sensors on the drawing board that specialize in materials detection, and chemical detection is most sophisticated right now, Nokes said. For instance, Sandia's "programmable diffraction grating" uses MEMS to make minute mechanical adjustments that create a matching template for any chemical so that a single, inexpensive chip can perform optical correlation spectroscopy on any known or future chemical agent.