Seems like demining is done manually - slow and dangerous. I'm surprised we can't build the tech to identify mines and extract them safely.
The problem in the Falklands is that demining machines are very destructive of habitat. A demining machine effectively ploughs the land - a useful side-effect if you're clearing agricultural land, but a serious problem if you're clearing a nature reserve.
http://www.gichd.org/resources/publications/detail/publicati...
Another thing that makes demining really tedious is the amount of scrap metal in the earth - and if it's former farm land that's been mined it's bound to have loads of metal in it, mostly remains of agricultural equipment, and on the minestrips of former Yugoslavia you often enough hit archeological artifacts during mine-clearing.
If you wanted to do automated demining, you'd need:
1) scanner technology that can differentiate between scrap metal and mines, as well as penetrate at least 1-2m into the ground
2) robots on 2 legs and with 2 arms - basically, humanoid robots. Anything other is only usable on flat fields.
3) robots must be able to conquer vast amounts of vegetation. For example, the Yugoslavian wars were 2 decades ago - many of the minestrips that are still not cleared up are in the backside of the country and heavily overgrown by everything from grass to massive trees now. (Source: am half Croatian, seen some minefields)
4) robots with high-quality visual processing. The most devious mines are those that have been hidden in / masqueraded as stuff like old cans, soccer balls or other stuff that children pick up, play with it and boooom. Any robot which has to do demining in such areas must be able to detect and avoid these.
5) cheap robots. A single robot with above capabilities will run well into the 8-figures at the moment (esp. due to #2), a human is cheaper (especially as there's protective armor available that helps to prevent or soften severe injuries).
While a tank or an AK47 might work in this environment as intended, a robot with fine gears and other relatively delicate moving parts will have trouble with the sand... and the corrosive sea water environment. It will require an extensive amount of maintenance.
We see 4 legged animals in forests and on mountains, seeming to donjust fine navigating the terrain (and often doing better than humans).
Hm. Could you use light bots to search out and mark the mines, and send in diggers only once the area has been thoroughly inspected? Of course that just makes everything even more complicated...
How much of normal mine clearance is searching vs. removing? Is there any advantage to robots that mark mines for manual removal, or is that already the "easy" part?
> 1) scanner technology that can differentiate between scrap metal and mines, as well as penetrate at least 1-2m into the ground
Forgive me if this is naive, but do mines actually get buried 2m deep? I would think that would be unfeasibly difficult. Or are we thinking of shallower mines that were buried by natural soil movement?
> 2) robots on 2 legs and with 2 arms - basically, humanoid robots. Anything other is only usable on flat fields.
It could conceivably be helpful to use a robot that can cover moderately rugged terrain. You'd still need to use manual methods on the rougher patches, but it might at least reduce the size of the problem.
I realize it's probably unfeasibly expensive no matter what, but it's interesting to think about the problem.