For many months now we have been playing games to encourage Venn to work with her nose, and we are continuing to develop those skills. As mentioned previously, with our acceptance testing still outstanding, we can do no more as far as SARDA is concerned, but with time on our hands, little in the way of alternatives to keep a “busy” dog occupied, and heaps of enthusiasm, we have been trying to ensure the best start when the world turns right-way-up again when we are finally able to start moving forwards.
If all goes to plan, then Venn will work as a trailing dog. This type of dog is able to follow the deposits of human scent left behind as a person passes through the environment. Whilst it is nice to imagine the dog picking up the scent and following it to actually find the bewildered missing person lost in the woods, these dogs also play a key role in determining the “direction of travel” so that other resources such as air-scenting dogs, helicopters and search teams already in the area can be quickly focussed into a smaller, more accurate search pattern. The latter is much less glamorous for the trailing dog-team, but equally important in terms of finding missing persons.
A trailing dog has to answer two questions: “Was the missing person here?” and “Where did they go?”
“Was the missing person here?” is a complex challenge for the dog. In order to answer it, the dog must be able to take and memorise a sample of the missing person’s scent, find a match for it amongst the background scent-noise of every other human who has been in the area, and communicate the fact that it has found a match to the handler. This process is called “scent discrimination” and the communication is called an “indication” and this is where training starts.
For us, scent discrimination started not with human scent (probably considered a high risk activity just at the moment anyway) but with smelly items that are out of context in the grounds. At the moment we are using cloves, cardamom, garlic, ginger, cinnamon, cumin, coffee and vanilla as our panel.
The process starts with scent-boxes: old takeaway cartons with the scent added to the lid. In the first instance it would typically be one scented and two unmarked decoys spread out in a line perpendicular to the wind. To learn the game, the dog is encouraged to sniff a sample of the scent, and from downwind is then taken to investigate the boxes. Any sign of interest in the correct box is encouraged and rewarded. Conventionally a reward would be contained within each box, to be given when the dog selects the correct one. However, given Venn’s surprising lack of food motivation, this initially resulted in much wasted food, so we opted to treat over the correct box from an early stage. Either way the end result is that the dog associates finding the marked box and getting a reward for her troubles.
Once the dog is expecting a reward for getting it right, the indication can be developed, either by asking, or by seeing what she does naturally to speed up the delivery of the reward. We were quite lucky in that Venn naturally sat at this point. She started by pairing this with a silent bark, but over time this has developed to a sit and full-fledged bark.
Once the find and indication is bombproof, the game can be made more challenging, first by trying the same single scent set-up with different scents, and then by adding actual scents to the two decoy boxes.
And then repeat in different settings, in different conditions, and with different scents until it is all second nature.
Despite her red harness being part of the pantomime of repeated patterns that tell Venn that she is working (along with my fluro jacket, and the unrolling of the trailing line), all of this work is done off the collar. This will be important further down the line.
Commentaires