Metal detectors can’t distinguish a landmine (top left) from junk, but the new radar-based Rascan device creates images (bottom row) that could at least double detection rates
Never let it be said that rubbish collectors aren't some of the bravest people on Earth. I don't mean the nice people who empty your bin, but the experts who scour the forgotten fields of past conflicts, first with sniffer dogs, then metal detectors, and finally with long-bladed knives, stretched out on their stomachs while nervously probing the ground ahead. Their usual name is minesweepers, but their profession would be more accurately categorised as waste removal. The UN estimates that there are more than 100m landmines in the world, which kill some 5,000 people every year. Yet minesweepers spend 90 per cent of their time digging up pieces of harmless junk on the off-chance it could cause an explosion. "You get a signal from your metal detector, and you don't know what it is," says Phil Halford, who works for the Mines Advisory Group (MAG). He has spent his working life clearing explosives everywhere from Afghanistan to Sudan. "It could be a piece of tin, a piece of scrap, whatever. The biggest problem is false-positive detection." But now an international group of scientists has come up with a solution. Called Rascan, their device uses radar to not only spot a buried object, but to pick out a rough image. It could help minesweepers across the world skip over rubbish and focus on their real job. Colin Windsor, a retired British nuclear physicist working on the project, estimates that this could at least double the detection rate. Radar of a more primitive form is already used in the field. It relies on the same long-wavelength radiation used in microwave ovens. These waves travel through dense materials mostly unheeded until they come across an object with starkly different properties to the surrounding medium, at which point they bounce back in the direction of the transmitter. In this way, radar systems searching for mines send out a pulse into the ground and, if they detect its return, assume the worst and give a warning buzz. These radar systems complement metal detectors because they can spot mines made of plastic as well as metal, but they are no less prone to false positives. Rascan, on the other hand, should have a far higher success rate because it sends out a continuous stream of microwaves rather than a pulse. When the stream reflects off buried objects, parts of it (unlike a pulse) reflect at different points depending on the object's shape. The result is a returning stream of microwaves that are out of sync with one another. By detecting these differences, Rascan can build up an image of the object, and hopefully show whether it is a mine or, say, a drinks can. Windsor hopes Rascan will come to replace normal ground-penetrating radar, which he says never took off in minesweeping. "Normal radar is expensive—several times the cost of [our system]—and you need some computational cleverness before you can interpret it," he explains. "It's not so clear cut." The journey from the lab to the field is long, however. The first step for Windsor and his colleagues is to develop a suitable vehicle for Rascan, so it can trundle over a minefield before human beings have to venture across it. Researchers at the University of Florence are now developing a robot that can do just that. But the biggest question is whether minesweepers can learn to trust the system. It's one thing ruling out false positives, but they need to be sure Rascan doesn't rule out true positives either. "We have to be able to say to communities, 'This is safe,'" says Kate Wiggins, a spokesperson for MAG. "Ultimately, the only way to make sure is to go through an area with hands." And in any case, before Windsor's group gets to field tests they need to find funding: at least £100,000 for the initial stage. Since they are looking to develop Rascan as a humanitarian tool—that is, one for protecting inhabitants of past conflict zones—they would prefer their sponsors to be philanthropists rather than partisan institutions. Yet the main interest so far has come from the military and, with the scientists spread across the US, Britain, Italy, Russia and Japan, this has caused problems. "The military of any one country gets a little nervous," says Tim Bechtel, a group member at Franklin and Marshall College in Pennsylvania. "It becomes a sticking point, because the US military doesn't want any of its money going to Russia. We've run into this problem with Italian colleagues, too." For now, then, Rascan's developers confront a stubborn irony: although every year landmines kill thousands of people and more than $600m (£380m) in international aid is spent on trying to remove them, the search for even modest funding is proving to be a minefield in itself.
Never let it be said that rubbish collectors aren't some of the bravest people on Earth. I don't mean the nice people who empty your bin, but the experts who scour the forgotten fields of past conflicts, first with sniffer dogs, then metal detectors, and finally with long-bladed knives, stretched out on their stomachs while nervously probing the ground ahead. Their usual name is minesweepers, but their profession would be more accurately categorised as waste removal. The UN estimates that there are more than 100m landmines in the world, which kill some 5,000 people every year. Yet minesweepers spend 90 per cent of their time digging up pieces of harmless junk on the off-chance it could cause an explosion. "You get a signal from your metal detector, and you don't know what it is," says Phil Halford, who works for the Mines Advisory Group (MAG). He has spent his working life clearing explosives everywhere from Afghanistan to Sudan. "It could be a piece of tin, a piece of scrap, whatever. The biggest problem is false-positive detection." But now an international group of scientists has come up with a solution. Called Rascan, their device uses radar to not only spot a buried object, but to pick out a rough image. It could help minesweepers across the world skip over rubbish and focus on their real job. Colin Windsor, a retired British nuclear physicist working on the project, estimates that this could at least double the detection rate. Radar of a more primitive form is already used in the field. It relies on the same long-wavelength radiation used in microwave ovens. These waves travel through dense materials mostly unheeded until they come across an object with starkly different properties to the surrounding medium, at which point they bounce back in the direction of the transmitter. In this way, radar systems searching for mines send out a pulse into the ground and, if they detect its return, assume the worst and give a warning buzz. These radar systems complement metal detectors because they can spot mines made of plastic as well as metal, but they are no less prone to false positives. Rascan, on the other hand, should have a far higher success rate because it sends out a continuous stream of microwaves rather than a pulse. When the stream reflects off buried objects, parts of it (unlike a pulse) reflect at different points depending on the object's shape. The result is a returning stream of microwaves that are out of sync with one another. By detecting these differences, Rascan can build up an image of the object, and hopefully show whether it is a mine or, say, a drinks can. Windsor hopes Rascan will come to replace normal ground-penetrating radar, which he says never took off in minesweeping. "Normal radar is expensive—several times the cost of [our system]—and you need some computational cleverness before you can interpret it," he explains. "It's not so clear cut." The journey from the lab to the field is long, however. The first step for Windsor and his colleagues is to develop a suitable vehicle for Rascan, so it can trundle over a minefield before human beings have to venture across it. Researchers at the University of Florence are now developing a robot that can do just that. But the biggest question is whether minesweepers can learn to trust the system. It's one thing ruling out false positives, but they need to be sure Rascan doesn't rule out true positives either. "We have to be able to say to communities, 'This is safe,'" says Kate Wiggins, a spokesperson for MAG. "Ultimately, the only way to make sure is to go through an area with hands." And in any case, before Windsor's group gets to field tests they need to find funding: at least £100,000 for the initial stage. Since they are looking to develop Rascan as a humanitarian tool—that is, one for protecting inhabitants of past conflict zones—they would prefer their sponsors to be philanthropists rather than partisan institutions. Yet the main interest so far has come from the military and, with the scientists spread across the US, Britain, Italy, Russia and Japan, this has caused problems. "The military of any one country gets a little nervous," says Tim Bechtel, a group member at Franklin and Marshall College in Pennsylvania. "It becomes a sticking point, because the US military doesn't want any of its money going to Russia. We've run into this problem with Italian colleagues, too." For now, then, Rascan's developers confront a stubborn irony: although every year landmines kill thousands of people and more than $600m (£380m) in international aid is spent on trying to remove them, the search for even modest funding is proving to be a minefield in itself.
