ANN ARBOR, MICH. — Harvard physicist Paul Horowitz and a former student play an odd game of hide-and-seek, burying objects in the Cambridge, Mass., dirt and finding them again with sound waves. The exercise tests what Professor Horowitz believes will be a cheap, portable way to detect land mines.
The device consists of a baton-sized probe, a laptop computer, and a set of headphones. Mimicking a bat's sonar, it sends out 20 acoustic "pings" a second that reflect off what the tip touches. The setup can check the signals it receives against a library of acoustic signatures in the computer, telling mine from matter in milliseconds with a synthesized voice that advises "danger, land mine" or "harmless object."
Removing land mines makes nobody's top 10 jobs list. It's a slow process, combining the perils of a high-wire act with the accuracy of beachcombing.
An estimated 100 million unexploded land mines stud the globe, and experts say it would take 10,000 years to demine Cambodia alone, using present methods. But an infusion of money for humanitarian demining research from the US government has started to pay dividends.
Several technologies, including Horo-witz's - which he is building with the help of the Bosack-Kruger Foundation in Redmond, Va. - appear ready for service. They promise devices that may dramatically cut the time it takes to find and neutralize land mines and make the process much safer.
Brute force is the demining method of choice for the military, which can drive a specially outfitted vehicle across a minefield to clear a swath double-time. But for peasants in Cambodia, Mozambique, or Laos - who sweep with a metal detector, then scrape and clear dirt by hand to uncover the explosive - neutralizing just one mine involves enormously intense labor.
Able to detect plastic mines
Usually the metal detector has pinpointed a tin can or the shrapnel omnipresent in war zones. Sometimes the mine explodes, inflicting devastating, occasionally fatal injuries. Add the fact that most new mines are cased in plastic, and thus invisible to standard metal detectors, and the old way simply cannot cope with the mines now in the ground, let alone the 20 being laid for each one destroyed, says Kosta Tsipis, head of the Humanitarian Demining Project at the Massachusetts Institute of Technology in Cambridge, Mass.
Mr. Tsipis is particularly optimistic about a method developed by MIT and JENTEK Sensors, a corporate affiliate of the school, that uses a modernized metal detector called a meandering winding magnetometer, or MWM. Like the old version, the MWM generates an electromagnetic field, and can sense disturbances in that field caused by other metallic objects.
By aligning the detected fields, an MWM probe can give the operator a rough idea of everything from the object's size, shape, and depth in the earth to its material makeup in a matter of seconds.
MWMs could be used with an air compressor, to blow dirt off the mine, and a smothering, explosive foam to safely detonate it. Tsipis estimates that the combination would cost a few thousand dollars - about the same as a premium metal detector - but would be 10 to 20 times faster. Researchers at MIT are now developing a similar device to find plastic mines, he says.
David Summers, an engineer at the University of Missouri, Rolla, has been working on a different tool that detects and defuses both plastic and metal mines using high-pressure water jets. The remote-controlled device has 12 nozzles, each capable of shooting water 8 inches into soil in 1/10th of a second. "We listen [through earphones] to the noise as the jet goes down, and the nozzle can then point to where the mine is, based on the changes in the noise when the jet hits the target," Summers says. A processor does some heavy lifting to filter the sounds for the eavesdroppers.
Like Horowitz's system, Summers's relies on a computer to distinguish a pipe or a tree root from the 700-odd types of mines.
"The problem is that in some parts of the world a pipe is a bomb and so much of the time, they [minelayers] are making them themselves, so you can't rely on finding a known signature," he says. When Summers's machine finds a mine, it switches to a second head that can liquefy the soil and suck it up. The operator - sitting 500 feet away and watching safely on a TV screen - employs a third head to sever fuse from charge with a sand-water mixture.
Cost a major factor
At about $20,000 for a refined prototype, Summers's device needs to be cheaper before it becomes standard in poor countries, where $150 a month is not an uncommon wage for a deminer. But the most serious drawback is its need for water. While significantly quicker than manual demining, its 300-gallon-an-hour thirst makes it a poor option for arid regions.
As a field deminer who has set up clearing programs in Cambodia and Bosnia, Georges Focsaneanu has surprisingly little enthusiasm for efforts of engineers in the laboratory. "No one has come up with a magic bullet yet," says Mr. Focsaneanu, who is from Ottawa. One reason, he says, is that new methods don't yet find mined areas more effectively than the current way - which often involves an accidental detonation.
Focsaneanu thinks the research dollars would be better spent training local deminers, and believes no extant technology beats creep-and-probe in ensuring that mines get unearthed. "You leave behind one land mine on a piece of ground and that whole area is mined" in the minds of people living there, he says.