On May 7-8 2012 Mindful Behaviors hosted a Steve White seminar on tracking and scent detection. We all enjoyed learning from Steve whose knowledge of canine scent work and animal learning theory is very impressive, he is also a phenomenal speaker. I’m already looking forward to inviting him over again.
The first day we had a workshop on Hydration Intensified Tracking Training (HITT), a method developed by Steve for teaching tracking on hard surfaces and under challenging conditions. The whole method is based on error-less learning where you keep changing criteria one at a time only after your dog has been at least 80% successful at the previous level. I had been experimenting with this with Forbes previously but lately had become a little lazy about it. Unfortunately Forbes was sick during the seminar so I couldn’t take him with me but I got to borrow the amazing Sojo from Oracle Aussies for the two days.
During the practical part of the workshop, the dogs worked on nose down behavior on asphalt. Below are two movie clips of Sojo enjoying the work (thanks Lesly Tailor for being the handler of my BlackBerry), this was his first tracking experience ever. Tracking is all about letting the dog figure out where the reinforcement is coming from by using pure classical conditioning: food is only available on the track. Of course they may wander off but once they figure out what pays off, it is amazing to watch how they are glued to the track:
When doing HITT with Forbes, the challenge I have been having is how to introduce the articles. His article indication is a retrieve, I have been practicing it like crazy and it works very well outside tracking. But while tracking, Forbes finds smelling the ground much more reinforcing than the articles, therefore, I had been scratching my head how to get around this. On vegetative surfaces his article indication is much better. In Steve’s technique you put the article at the end of the track by a gate or a fence. This way the dog cannot continue tracking without indicating the article and they will not get any reinforcement ahead because they cannot continue. After the seminar I practiced this at the Home Depot parking lot after their closing hours (looking like a complete moron hopping over fences :-)!), after a few training sessions the indication started to improve. Below is a movie of Forbes tracking with an article at the end, the location had no fencing so I tried it with a curb. Clearly, the indication is not as strong as is desirable but he did it. We’ll go back to more solid barriers.
The second day of the seminar was called Scentsational! which was all about training dogs to detect specific scents, be it drugs, money, essential oils, the odor really is not the point but the technique how you teach it. Steve trains everything by back-chaining, which is the way to do it if you want solid behaviors. The sequence of events is: search-locate-report. You start with the last component of the chain, which is the report. This way the dog is always moving towards a more familiar behavior and this keeps the behavior chain together. Simple behaviors to use as the report behavior are a sit or a down: as soon as the dog finds the scent in question they do a sit/down to tell you that they found it. Here is Steve’s movie link on how to train the whole sequence by using back-chaining:
At the seminar we had six pre-selected demo dogs doing the practical work. Sojo and I were one of the demo dog teams who kept practicing the foundation behaviors in the background while the seminar went on. The scent of the day was clove oil donated by Kate Mutzke (thank you, I wish you could’ve made it to the seminar!). Demo dogs worked on a robust nose targeting on a target stick containing the scent (operant conditioning) or eating treats from a container containing the scent (classical conditioning). Here is Sojo and myself practicing targeting of the scent:
At the end of the day when the demo dogs had had a lot of exposure to the target scent, we tested if they would pay any attention to the scent somewhere else in the room. The time of exposure that the dogs had had to the scent was altogether very short but some of them did start to display gravitation to the scent hidden in the room. Here is KPA CTP Stacey Shaw’s dog Target doing a great job:
Previously I had been experimenting with scent detection by using tea as the target odor. Instead of using containers that keep flipping over all the time, I got a lot of useful ideas at the seminar as to what to use, I need to make a trip to IKEA and will write another blog about this soon.