Commentary by Roger Martin Martin writes about research at KU for Kansas media and in a Web zine. Contact Martin at rmartin@kucr.ukans.edu or (785) 864-7239. Martin also can be heard on KANU 91.5 FM public radio Archives More articles by Roger Martin Subscribe now to receive KU News by email

Land mine research has taken back seat to homeland security

by Roger Martin

Media folks have attention deficit disorder.

We're forever fixating on some new threat to health, then dropping it. We can skip from sickle cell anemia to bubble boys in the blink of an eye. We also tend to jump around a lot when it comes to affronts to human dignity.

Of course it's not just the media. The public can be just as fickle.

In the mid-1990s, the big thing was land mines, remember? They're sure on the back burner these days, even though it's estimated that 60 million to 70 million unexploded mines are still out there, scattered across 70 nations.

Jim Stiles, an associate professor of electrical engineering at KU, has spent five years and about half a million dollars in Department of Defense money trying to figure out how to use radar to find these mines. Back when land mines were a leading affront to human dignity, the Army funded several groups to work on land-mine detection using different techniques.

The approach Stiles and his group took was to develop radar that will detect symmetrical objects buried near the surface.

That makes sense when you think about it. As Stiles points out, people tend to make objects that are symmetrical, while nature tends toward irregular shapes.

Trouble is, Stiles says, his radar doesn't work well in extremely rocky soil where rocks and mines can cluster together to create deceptively irregular shapes.

A mine hunter doesn't want to hear stuff like that.

But other Army-funded research into mine detection also has limits, Stiles says. For example, chemical sensors that sniff out explosives get confused in battlefields that are saturated with chemical residues.

In yet another approach, acoustic detectors send a shiver of sound across a minefield. The mines vibrate the soil above them, and a laser detects this so that a computer can map the mines.

But the approach is slow, and if you speed it up you get more false positives.

The Army is most enthusiastic, Stiles says, about a sensing approach that's much like the magnetic resonance imaging used to detect tumors in the human body.

The advantage, he says, is that these MRI-like sensors can tell you not only that an object has the right size and depth to be a mine, but also that it's made up of explosives. The best solution, Stiles says, would be a de-mining instrument armed with several sensors. But it would be expensive, hard to maintain and challenging to operate.

Though the problem isn't fully solved, federal research funding is being reallocated from mine disclosure to new research areas, Stiles says.

If you've heard the words "homeland security," you can guess where trends in spending are headed.

That's right. It's suddenly a lot more important to detect explosives hidden on a human being than hidden in the ground.

Research into such matters no doubt will be urgent work until, inevitably, some fresh affront to human dignity emerges.