Andy Smith
MIne-action specialist

Free running EDD/MDD and accidents

(The terms EDD and MDD are used as synonyms below.)

QUESTION: A UN advisor told me that free running EDD are almost as good as dogs on the long/short leash, which is the system preferred by most EDD handlers today. He claimed that EDD have never missed a mine and that there have been no accidents related to EDD when used after mechanical preparation or as primary clearance. Can you give me any data on this issue? I have no idea, but it sounds like he can't be right.

First, dogs are usually used where people think there is a low probability of finding anything, so the chances of them missing something that is not there are slim. However, your instinct is right. There are at least 22 accidents in the DDAS where a dog missed a mine. Nine different demining NGOs and commercial companies were involved.

22 is not that many - but the DDAS has fewer than half the accidents that have been investigated and recorded. Also, that 22 involve missed mines that were found while the deminers were still there and an accident report was generated. Many more are found after the deminers have packed up - and many are found without them exploding and causing an accident. These are not in the database because they are rarely investigated and even when they are, I have not been allowed to gather those records.

Anyone who says that free running dogs are "almost as good" is ignoring the absolute nature of mine clearance. Land is clear or it is not. It cannot be almost clear. I have seen free running dogs search thoroughly - but I have also seen them wander about and not search large areas. I also know that they have missed large (easily visible) mines. To my mind, a single EDD search can never give sufficient confidence to declare land clear. Two appropriately trained and motivated dogs must be used one after the other and they must be used in a way that guarantees every bit of land has been searched. Using two dogs means that the perfrmance of each is QA "tested" by the other and there is a good chance of knowing when a dog is not working well. EDDs are best used as described in detail here. This involves two dogs being used and the handler stepping on all of the land AFTER the second dog has completed its search. To prevent the second dog simply following the scent of the first, the way they are run must be varied - but they should always be on a lead and work in a prescribed area. This opinion is informed by the fact that there is no record of an accident after using that method in the DDAS.

That said, the UN advisor was also partly right. With the right dog, right handler, and right minefield on the right day a free-running drill MIGHT work excellently. But it would be impossible for me as supervisor or QA to tell whether that was happening with sufficient confidence to release the land as "Cleared" to IMAS. With the drills described in Chapter 8: Mine Detecting Dogs of the Generic SOPS, I can tell when the dog is not covering the ground and I know that the handler has stepped on the entire area so he has a compelling reason to make sure that the dog is working well.
As supervisor or QA person, I do not have to be able to read a dog like an EDD handler - but I do need to have full confidence that the system used will find all mines/ERW and that it is being applied thoroughly. With free running dogs, I cannot have confidence of either.

The Generic SOPs are derived from best practice in several EDD/MDD programmes and have been widely approved by EDD people I respect.

See also: Comment and Questions