快猫短视频

Although James Trevelyan is based at the University of Western Australia, he does most of his de-mining work in Asia, notably Pakistan, where he has family. He has his work cut out: 70 of the world鈥檚 185 nations are affected by landmines, 30 of them seriously. An engineer by training, Trevelyan specialises in robotics and still runs a research laboratory at the university. But when he tried applying robotics to mine clearance, he quickly realised that new technology was not what was needed. The way forward, he decided, was to learn more about the people who had to clear the mines.

Your wife was born in Pakistan. How has that family connection helped you with research in that region?

There are two things about this that are relevant. One is that I鈥檓 a Muslim: I had to become a Muslim to marry her. The other is that I have, through her and her family, lots of connections. These things have given me an understanding of the culture, which has proved invaluable in understanding problems.

Could you give me an example?

One of the main issues with Afghan de-miners is that they will not lie down when they are investigating suspected mines, and this puts them at greater risk of injury if an explosion occurs. Expatriates who were managing the de-mining programmes gave me a whole range of opinions about why this was, including that it had something to do with their religion, or that Afghans did not like to expose their backs, or that it was just sheer bloody-mindedness. I knew enough about the religion to know that wasn鈥檛 the answer, and I also knew enough about Afghans to know that fear of exposing their backs was no explanation.

So what was it?

After a while it became beautifully clear and simple. De-mining in Afghanistan is a high-status occupation. It is difficult to become a de-miner. They don鈥檛 get paid much by our standards, but it is a huge amount by theirs. And they wear a smart uniform. A de-miner will not lie down and get his uniform dirty any more than a businessman in a Western country would lie down in his suit. So they squat.

What do you do about that?

Accept that de-miners are going to squat, and provide them with equipment that will protect them in that position.

How did that go down with the UN and other organisations running the de-mining programmes?

It was not accepted at the time, and I don鈥檛 think it鈥檚 accepted to this day. De-mining programmes are run by people who come in for between six months and two years, and it takes them that long to realise they鈥檙e not going to solve that problem. Then they are replaced by another generation of people who think they鈥檙e going to teach the Afghans to lie on the ground.

What is de-mining work like?

Very boring, physically demanding and hazardous. Underline hazardous. It can be very safe to work in minefields, but one wrong step and you lose your leg. It can also be tedious. The routine consists of clearing vegetation, marking the ground, scanning the ground with a metal detector, then investigating and in most cases removing every single piece of metal. De-miners can go for several months without finding a mine, but in that time they will unearth tens of thousands of pieces of metal. Some minefields have a lot of mines, many don鈥檛 have any: often the mines are removed illegally.

How did you get into de-mining?

My wife, who is involved in the Red Cross and is also a researcher in international politics, became aware of the landmine problem and tried to twist my arm into doing something about it. After a year of this she was at a conference in Melbourne in 1995 and got hold of Lieutenant General John Sanderson, who at the time was the army chief of staff. She told him: 鈥淚鈥檝e been trying to persuade my husband to do something about landmines,鈥� and he said, 鈥淕ood, we really do need people to look at that problem.鈥� When the chief of army staff tells you that you really should do something, you can鈥檛 refuse.

How did you start 鈥� with robotics?

The first year was a learning exercise. I was an engineer, and up to then I had worked mainly on robots. I鈥檇 designed robots to shear sheep, robots you could control over the Internet, and so on. I thought, how could you use robots in mine clearance? The main problem was the lack of a sensor that could accurately detect mines. Robots are just as delicate as people, and it would be important to fit them with sensors to ensure they don鈥檛 walk on mines. I thought of a machine suspended on cables above a minefield that would pound the ground to get rid of mines. It was very simple-minded and totally impractical. After six months, I realised that the sensing problem was not going to be solved in the short term. And that has proved correct. Here we are, seven years later, and there are still no sensors that can reliably tell us that it鈥檚 safe to take the next footstep in a minefield.

And that forced you to take a different approach?

I decided to look at the issue from a more fundamental point of view 鈥� from the point of view of those doing the de-mining. Through my family connections I met many of the people involved and realised that there were a host of seemingly simple problems that nobody was looking at.

Such as?

Such as the lack of low-cost, comfortable protective equipment for de-miners to wear, and simple hand tools that are comfortable and effective. Above all, there was almost no public information available about what minefields actually look like. It has amazed me how many researchers imagine that a minefield is like the nice, neat strip of moist grass just outside their laboratory.

You suggested that most research is still being directed at technologies that will help people detect individual mines. Is this misplaced?

Definitely. One of the key things that has motivated work on high-tech detectors is this notion that de-miners spend 20 minutes digging up a target only to find it is a metal scrap 鈥� and couldn鈥檛 they have a device that would tell them in a few seconds? Our research showed that it takes on average 45 seconds for a de-miner to investigate a metal fragment. It鈥檚 difficult for any foreseeable technology to do much better. In most countries, 50 to 70 per cent of the time you are clearing vegetation. Detection technologies still require you to clear the vegetation. So you can see that their overall impact on efficiency is going to be negligible.

What should researchers be concentrating on instead?

It is much cheaper to locate the areas where there are no mines and leave the mined areas to be dealt with later.

How do you do that?

You can use dogs to detect traces of the explosive. The technology is called REST (Remote Explosive Scent Tracing), and although it is still being developed, it鈥檚 already being used in places. You suck dust from a minefield through filters with a vacuum cleaner, then get the dogs to sniff the filters. Another idea is to look for traces of explosive in groundwater. TNT and other explosives used in mines are soluble in water. These kinds of techniques are generally used to monitor pollution. Whether they can be applied to mines I don鈥檛 think has been seriously investigated.

But I have a more radical approach for dealing with minefields in agricultural areas: why clear the mines? Anti-personnel mines do little damage to agricultural machinery, so it should be possible to adapt agricultural machinery and carry on agriculture in minefields after making sure there are no anti-vehicle mines.

How do you protect those farming the minefields?

We know how to protect people in vehicles: use a quarter-inch armour plate around the cab, and thick polycarbonate windows. There are many solutions to this problem, and some of them are already being used on de-mining vehicles.

How did you arrive at that idea?

I was concerned at the length of time that will pass before large areas of land are cleared of mines. Look at Croatia and Bosnia, where there is about 12,000 square kilometres of contaminated land. The number of square kilometres being cleared each year, at a budget of tens of millions of dollars, you can almost count on your fingers. That land is not going to be cleared in the foreseeable future using the sort of techniques we use now.

And the cost of restoring farmland to full production increases the longer it鈥檚 left out of production. If you go to Cambodia you find that large areas of valuable farmland suspected of being mined are being overgrown by immense thickets of bamboo which can only be removed by machinery. You鈥檒l see large drainage systems that don鈥檛 work properly because they are waiting for teams to come in and de-mine the access paths. In many countries, people are farming minefields anyway. Some 80 per cent of the mine clearance in Cambodia is being done by untrained villagers who have run out of patience waiting for de-mining teams to clear their lands. No records are being kept and no one knows how well it鈥檚 being done. I am simply arguing for some modest investment to improve on this.

One of the worst-affected countries for landmines is Afghanistan. What impact has the US bombing had on de-mining efforts there?

One of the immediate effects of the bombing was to blow up ammunition dumps, which scattered vast quantities of unexploded ammunition across the streets of the major cities 鈥� ankle-deep in places. Clearing that has been the immediate priority, and it has taken months of work. The other areas of immediate concern are cluster-bomb strike areas and removing mines from houses. All that has created an enormous demand on funds. And things have got worse. The security situation is so poor that there is very little work that can be done outside the urban areas, where the real mine problems are. You can鈥檛 send de-mining teams out unless they are going to be secure and aren鈥檛 going to have their vehicles and equipment stolen. Practically every usable vehicle was destroyed by US forces, so even a battered old truck being used by a de-mining team is a very attractive target.

Has your work on de-mining changed your perception of how research in general should be done?

Yes it has. My colleagues in industry say they spend most of their time dealing with human behaviour. Human behaviour is probably the greatest source of uncertainty in engineering, and de-mining is no different. How do you know the de-miners have cleared all the land they said they cleared? My research has always had a cross-disciplinary thread to it. To build a good sheep-shearing robot I had to understand sheep, shearers and the skill of shearing, as well as the technology involved in building a robot.