Thanks Robin.
An interesting article that shows MAG following others (such as TETRATEC in
Syria) by also training the dogs to find the ‘triggers’ rather than the
explosives. But it is journalism so it says nothing new and does get a little
confused. It has long been accepted that a close relationship between handler
and dog is essential for dog motivation, so their training must include time for
that relationship to be built. Training should include training the dogs to run
in a search pattern than ensures ground coverage (something no rat can begin to
do). And long ago in Jordan, I saw NPA dogs exercising after work in a swimming
pool – the poor animals do have a hard time in excessive heat.
The plastic mine they showed appears to have been mass produced and have a
large explosive fill (not 15kg!) so would be a fairly easy target for a good dog
to find after the mine has been in place for a while. With unconventional
improvised hazards, the initiation systems may include microwire crush switches,
nylon, rubber or plastics that cannot be detected with a metal detector and they
may surround the main explosive at a distance – so making a safe approach to
detect the actual explosives impossible. The mine pictured had a pressure plate
on top with a fuze mechanism beneath that contained some metal, so could
presumably be detected using a good metal detector. But it may have had other
initiation systems as well (in ISIS patterned minefields, it is common for some
random mines to have secondary initiation systems designed to kill those sent to
clear them). So the dog has to be trained to find the initiation system and
avoid triggering it rather than just find the explosive charge. In urban search,
the finding of improvised initiation systems that may be concealed in
rubble/rubbish can make the well trained dog the ONLY search method that is
practical. This is a time when a search system that is not 100% reliable really
is better than nothing. But I would still QA with a second dog repeating any
search to increase my confidence that the first dog was “tuned in” today. Both
dogs have to be trained using real examples of the actual items they are looking
for, of course.
It it is a dogs’ ability to locate initiation mechanisms that do not
include obvious volatiles that is most interesting. It seems that you can train
a dog to locate any apparently inert item in a given environment. This is
interesting to me because most research into mechanised explosive detection has
relied on locating molecules (parts per million/billion) of the explosive charge
on the assumption that this is what dogs detect. It certainly can be - but the
dog is detecting far more than that when it finds either a factory sealed
munition or a microwire crush switch. In the earlies we called it detecting a
“bouquet” rather than any particular flower and I think that is still a useful
term.
Regards,
Andy
Sent: Wednesday, July 28, 2021 11:17 PM
Subject: [HD] Canine ‘super-detectors’: the dogs working to free
Iraq from Isis mines | Global development | The Guardian