An estimated 100 million land mines sit just beneath the earth's surface, waiting for something to come along and supply the 10 to 20 pounds of pressure required for detonation. A large animal could do it; so could a falling branch or a rock set loose by erosion. Often it's a human being who does it; 25,000 civilians are killed or maimed annually by these remnants of wars past--sometimes decades past. Beyond the human cost, the mines render valuable land unusable in impoverished regions such as Angola, Cambodia, and the former Yugoslavia.
Generally, government- and UN-trained civilian groups armed with face shields, metal detectors, and pointed sticks attempt to remove these mines. The process is slow. Experts say that 20 new mines go in the ground for each one that is removed, and without some sort of breakthrough, catching up seems unlikely. But things may change. Professor of physics Paul Horowitz '65, Ph.D. '70, recently attended the Landmine Brainstorming Workshop at MIT, where he met with a number of deminers. "They come begging to you, 'Is there anything you can think of to help us?'" he says, "and the answer is, 'Yeah!'"
Horowitz is lead author of "New Technological Approaches to Humanitarian Demining." The study, produced by a group of academic scientists and sponsored by the Pentagon's Defense Advanced Research Projects Agency, details several potential solutions to the demining problem. The challenge is not so much one of detection as of discrimination. Metal detectors locate mines, but they also locate every other object containing a fraction of a gram of metal. In an area that was once a battlefield, the ratio of suspect objects to actual mines can be as great as 1,000 to 1.
In current practice, the deminer lies flat on the ground and uses a pointed stick or prod to scrape away the dirt around a suspect object. Depending on the density of the soil, this can take anywhere from five to 20 minutes. If the object is a mine, it is detonated by explosives; if not, the deminer moves on to the next suspect. "If you're willing to use a prod," Horowitz asks, "why not prod with something that will tell you more?"
The goal is to produce techniques that will reduce discrimination time to less than one minute per object. Horowitz believes the technology exists to do this--but it has to be applied. "This is not building a rocket," he says. "This is simple stuff."
One idea employs a prod designed to test for a mine's "acoustic signature." This takes advantage of the fact that a mine--containing plastic, metal, explosives, and air spaces--has very different acoustic properties from the surrounding soil. The specialized prod sends a series of vibrations at varied sonic frequencies and then records responding vibrations from any suspicious object. This information produces an acoustic signature that can be compared with the frequency responses of known mines. One of Horowitz's students, Jonathan Wolff '97, has built a prototype that shows promising results in identifying a mine by type less than 10 seconds after the prod touches it. Other physics-based techniques in Horowitz's study use nuclear quadrupole resonance and x-ray backscatter technologies.
Perhaps even more promising is the field of vapor sensing. Trained dogs have been quite successful in detecting mines. No one is exactly sure what part of the mine the dog detects, but Horowitz says that "dog handlers will follow their dogs through a minefield without injury." But several factors can reduce canine effectiveness: weather, soil type, and how recently the mines were planted. Efforts are therefore underway to come up with an "artificial nose."
One intriguing proposal for further demining research involves genetically altered fruit flies. A fruit fly is capable of detecting even a few molecules of a pheromone (sex attractant). Horowitz's study suggests it would be possible to breed mutant fruit flies that are attracted to TNT, the explosive used in 80 percent of land mines. Released in quantity in a minefield (and phosphorescently tagged), the flies would cluster around buried mines, providing an obvious visual clue. At the end of the day, TNT would be used to recapture the flies. One advantage to such a method, Horowitz says, would be that "there'd be nothing mechanical to repair."
The study notes that, even after careful demining, a residue often remains: a very small percentage of mines cannot be cleared safely. It therefore recommends the use of plentiful "indigenous ruminants" (mainly sheep and goats) to locate these last few mines "sacrificially" (wounded animals "would be slaughtered and usefully consumed"). Animals could also replace the current "validation" procedure, in which deminers, arms joined, form a row and walk across a cleared area to show local residents that it is safe. Animal-rights groups may dislike this solution, but the study notes that failure to complete validation may result in tragedy for a child playing soccer in a newly cleared public area.
~ Matt O'Keefe
