Journal Archive > 2002 > February

Keith Albert

Keith Albert
© Mark Morelli

Doctoral student works to rid the world of land mines

Chemistry doctoral student Keith Albert is an active guy. He enjoys hiking and camping and has been known to take leisurely walks through land mine fields. That's right, land mine fields.

As a member of Robinson Professor of Chemistry David Walt's research team, Albert visited mine fields at the Fort Leonard Wood Army Base in Missouri to determine if the portable "artificial nose" detection system that the Walt lab and their collaborators had created was up to snuff.

"The army base had constructed a land mine field where we could test our equipment," Albert says. "The mines were buried without their 'triggers,' so the field was safe to walk in and to conduct experiments. But if there was a thunder and lightning storm, the field supervisors could not ensure everyone's safety. In the event of such a storm, we were told to high tail it out of the field's perimeter."

Man's best friend
The inspiration behind the artificial nose technology is man's best friend. Dogs have a heightened sense of smell and can detect, among other things, the small amounts of explosive vapor that reside on the surface above buried land mines. Dogs are very effective land mine detectors, but they are expensive to train, and there is the risk of injury.

Recognizing the need for more effective detection techniques for plastic-based land mines, the U.S. Department of Defense (DOD) launched a program aimed at creating better land mine detectors. The goal: To help rid the world of the more than 100 million land mines that are buried in more than 60 countries. Tufts, along with several other academic institutions and corporations, was given funding by the DOD's Defense Advanced Research Projects Agency (DARPA) in the late 1990s to develop more effective means of locating land mines.

"I joined the Walt lab in 1997, and at that point, a lot of work had already been done to develop the concepts of what we call the artificial nose, which is a system that discriminates between different odors or vapors," Albert says. "For instance, a human can discriminate between numerous smells. Even though we don't have specific receptors for bacon or blueberry pie odors, once your brain has been exposed to these odors, the next time you smell them, you will be able to recognize them."

Cross-disciplinary collaboration
The application of smell to chemical sensing grew out of a collaboration between the Walt lab in chemistry and the laboratory of John Kauer, professor of neuroscience at the School of Medicine, in the mid-1990s.

Albert and another graduate student, Shannon Stitzel, worked to further advance the artificial nose's capabilities, both in the lab and in the field. They wondered if it would be possible for an artificial system to detect the same explosives odors or concentration levels that the advanced olfactory system of a dog could.

Albert spent countless hours screening sensor materials and collecting data. While the odors he was trying to identify were minuscule, the effort was not. "There were many challenges that faced us during this project," Albert says. "The odor signature components that the Department of Defense told us to identify were in such ultra-low levels that one had to not only have a system with a great deal of sensitivity, but also one that was highly specific and fast. If the system didn't have these attributes, then the user could risk injury or even death."

The pitfalls that come with a faulty land mine detection system can be summed up in two phrases: False negative and false positive. A false negative occurs when a user gets a "no" reading in an area that does, in fact, contain mines. When this happens, the user would proceed through the mined area believing that was safe. A false positive occurs when the system says a mine is present when it is not. Because the user thinks there is a mine, valuable time is wasted.

Every 22 minutesÉ
Developing sophisticated land mine detection equipment is critical. Every 22 minutes, a civilian is killed or maimed by a land mine somewhere in the world. Thirty to 40 percent of land mine victims are under age 15.

The researchers knew that their work in the lab could tell them only so much. "The lab environment can be controlled. You control the humidity and temperature. You can perform your research during the day or at night," Albert says. "But when you take your equipment to another facility, and you're told that it has to operate under unknown conditions at a certain time, it's a great challenge and a great way to test your capabilities."

At Fort Leonard Wood, temperatures and humidity levels in the mid-90s replaced the lab's controlled environment. The laboratory-based system was re-created on a jogging stroller with different software and different components, including a battery pack. "Our collaborators helped us build a portable sniffer system on a jogging stroller, and that's what we took to Fort Leonard Wood," Albert says. "It was both exciting and stressful to contend with environmental factors like terrain, temperature and the ultra-low concentration levels of explosives vapors above the mines."

The difficulties presented in the field could be measured in inches. While the research team had no problem detecting explosives vapor in the lab or in the field with spiked ground samples, the buried mines at Fort Leonard Wood presented a problem. Nearly all teams on the DARPA project failed to detect buried plastic land mines at the field. "From the data obtained during the field tests, we were sure that our portable equipment was unable to detect buried land mines," Albert says. "But we definitely proved the concept that an artificial nose has the capabilities for such detections."

The science behind land mine detection is complex. It may be some time before a perfect system is developed that can detect all the different types of mines that are buried everywhere from shallow waters to the hard terrain of places like Afghanistan. Still, the effort will continue. It's a pursuit borne of science but with a humanitarian end.

"When you talk about land mines, people can comprehend it. They understand that there's a problem," says Albert, who is scheduled to receive his Ph.D. in chemistry in May. "But the future looks bright for land mine detection research, and I'm proud to say that I was a part of it."

Bob Bochnak covers the Graduate School of Arts & Sciences for the Tufts Journal.