Skip to Content

Finding Land Mines Faster

Andrew Heafitz, an MIT instructor, is working to improve the field of humanitarian demining.
January 24, 2006

“Humanitarian” demining – the process of reclaiming land that has become unusable because of residual antipersonnel mines – is a relatively new field, which began in earnest in the late-1980s. For the past four years, Andrew Heafitz, an instructor in the MIT course Design for Demining, has been creating devices to improve this labor-intensive, hazardous, and underfunded task.

A tinkerer and an inventor by nature – he got his first patent as a junior in high school and won a prestigious inventor’s award in 2002 while an undergraduate at MIT – Heafitz’s passion is designing and building low-cost tools that make locating land mines faster and safer, and that can be manufactured and operated by local people in places like Afghanistan and Zimbabwe.

Recently, Technology Review spoke with Heafitz about his work.

Technology Review: Before discussing your own projects, could you speak a little about the current situation with humanitarian demining? A decade ago, it looked pretty grim, with new mines getting buried much faster than older ones were being cleared.

Andrew Heafitz: Well, all these figures are generally made up….The most recent ones I’ve seen, though, show that they’re still putting in more [mines] than they’re taking out, but it’s gotten much better. The International Campaign to Ban Landmines and the 1997 Mine Ban Treaty made a big difference.

The main technological advances have to do with more sophisticated blending of techniques and mechanization. For example, the flail, which is a bunch of rotating chains that beat the ground, was originally invented for demining. But it was damaging mines without setting them off, and making them unstable. Using a flail as a kind of weed-wacker, though, to clear away brush – which can be 50% of the problem [in demining] – is very effective.

A equipment advance is ground compensation metal detectors. The ground has a magnetic force that can throw off older detectors. Ground compensation detectors look for a differential signal instead of an absolute one.

So now one approaches the problem in layers: one might use a flail for brush-clearing, bring in dogs to find a perimeter, then do manual detecting in a much smaller area with no brush in it, which is much faster and has a better success rate…[But] there are no silver bullets in this field. The only way to really solve the problem is to make land mines stigmatized, like chemical weapons.

TR: One of your first successes was redesigning a simple probe, which the demining community has recognized as a marked improvement. Can you describe the development of that device? Did you make any money from it?

AH: A probe is used after a metal detector has found a possible mine. It’s like a large shish-kebab stick that’s poked into the ground at a shallow angle so you don’t hit the pressure plate on top of a mine. Traditionally, probes are either round or like bayonets. We found that using an oval design and twisting it when probing reduced the force needed by 30-50%. It means you don’t have to push as hard on the side of a mine, and you can go deeper in the ground. We also made it longer and added a hand shield.

When we sent it in a deminers’ e-mail list, though, we got laughed at. The demining community is really tough: If it doesn’t work, it will get them killed. But a couple of people tried it and found it was noticeably easier. Today it’s incorporated into a toolkit made by Security Devices in Zimbabwe, and people tell us they’re modifying their probes. It’s now called the MIT Profile Probe.

We give away our designs instead of patenting them. People can make their own.

[Click here to go to an MIT web page with images of mines and demining devices and processes.]

TR: Many of the things you and your students have worked on – the probe, a magnetic rake, protective gear – are low-tech. Wouldn’t devices that use autonomous or remote-sensing elements in some cases be more effective and safer?

AH: There are places for mechanized demining, but the best results are often making small changes to existing equipment that can then be built and used by anyone. Let’s say you’re in Angola and your robot’s 9-volt battery dies: Where are you going to get a new one? Or Windows crashes – who’s going to know how to reboot it? I saw a demonstration once with a sensor that used a keyboard. It got sand in it – and that was the end of the trial.

Also, there’s variation from area to area – rocks, trees, river banks. It’s not like the image of a golf course that’s been mined.

Secondly, you employ a lot of people [with handheld tools]. A deminer in Afghanistan earns $120 a month, so you can have 1,000 people working for the cost of a piece of expensive machinery.

As far as safety, a deminer will have an accident about every 30 years – it’s safer than driving a taxi in many of these places. If someone is properly trained and wears a $40 face visor and Kevlar blast vest, he’ll walk away from an accident with minor hearing damage for a couple of weeks.

One of the most important things [in demining] is the hand-over ceremony. A local community has to believe their land is clean – or it’s a waste of time. After the deminers are done, they will link arms and walk across the field – and then they start playing soccer on it. The community joins in, and it becomes a big party. Would you play soccer behind a robot?

TR: Another tool you’ve worked on is a “mag” rake? What is it?

AH: A big problem in demining is false positives: for every land mine there can be 1,000 metal fragments [in a field]….This is the most tedious, boring work you can imagine. Our goal is to use the deminer in the best possible way: not digging at fragments, but at a dangerous mine. A rake with magnets on it could pick up surface fragments, saving a lot of time. You would kneel behind it and “tickle” the ground.

We did studies that showed there are situations when a rake makes sense. Now the big question is: How do you build one that will not blow apart if there’s an accident? The last thing you want is to add more shrapnel to a mine field.

TR: Another of your current projects also addresses false positives. And it’s also a lot more complex: redesigning a military mine detector so it’s affordable for individual deminers. Can you describe that effort?

AH: Well, we’re still negotiating, but a nonprofit group, Benetech, contacted us on a project where they’ve licensed U.S. military technology for detecting nitrogen bonds in explosives. It uses what’s called Quadrupole Resonance (QR) technology that identifies explosive material using radio frequency. The device can also detect low-metal mines. They want to make a stripped-down, low-cost version of it, for around $2,000, instead of $60,000. If such a device can be made, deminers wouldn’t have to dig up all the false positives. It might speed up [humanitarian] demining by a factor of 10.

We plan to work with Benetech on incorporating their design for a lower-cost sensor, which looks feasible, into a machine that can be carried by an individual.

TR: What do you see yourself doing 10 years from now?

I really like product design – inventing things and designing products. But I don’t really want to design a better toaster….Demining is still early in its product curve; things we do can actually make a difference.

Keep Reading

Most Popular

Large language models can do jaw-dropping things. But nobody knows exactly why.

And that's a problem. Figuring it out is one of the biggest scientific puzzles of our time and a crucial step towards controlling more powerful future models.

OpenAI teases an amazing new generative video model called Sora

The firm is sharing Sora with a small group of safety testers but the rest of us will have to wait to learn more.

Google’s Gemini is now in everything. Here’s how you can try it out.

Gmail, Docs, and more will now come with Gemini baked in. But Europeans will have to wait before they can download the app.

This baby with a head camera helped teach an AI how kids learn language

A neural network trained on the experiences of a single young child managed to learn one of the core components of language: how to match words to the objects they represent.

Stay connected

Illustration by Rose Wong

Get the latest updates from
MIT Technology Review

Discover special offers, top stories, upcoming events, and more.

Thank you for submitting your email!

Explore more newsletters

It looks like something went wrong.

We’re having trouble saving your preferences. Try refreshing this page and updating them one more time. If you continue to get this message, reach out to us at customer-service@technologyreview.com with a list of newsletters you’d like to receive.