Detection Dogs – The Smell of Success

Shane Clarke, inventor of the Dogwall, has trained Cocker Spaniels to become an elite force in the detection of pests. 

Termite detection dogs have been around for over 20 years. However, with a fully trained dog costing around $25,000, not many have tried this unique business tool. Even fewer companies have fully embraced dogs as part of their termite business, in part due to the variability in quality of the dogs and their variable detection skills.

However, what if the cost of a termite detection dog was significantly lower? What if the success rate in detecting termites and termite damage was close to 100%? What if the dogs could detect more than just termites? What if you could even train the dogs yourself? All this may soon be possible thanks to a new dog training system developed by Shane Clarke from Pestforce – the Dogwall.

Mr Clarke is a leading figure in the Sydney pest control industry and has been using termite detection dogs for nearly 25 years. Pestforce has developed a trusted reputation in the northern suburbs of Sydney based on their high levels of expertise and service. The use of termite detection dogs not only provides a point of difference to their competition, but also adds significant value to their business. “We use termite detection dogs to help find termite activity and damage, in termite baiting and detecting nests in trees, it really has changed the way we do termite work.”

Over the years Mr Clarke has been working with Steve Austin, perhaps the leading dog trainer in Australia. In observing his techniques, Mr Clarke thought there could be a better, quicker way to train detection dogs. After a few brainstorming sessions, the Dogwall concept came into existence and development began. The finished article has now been operating for nearly six months, with great results.

The Dogwall

The Dogwall works on the principle that dogs like to chase and retrieve. This innate behaviour is utilised by training the dogs to chase and retrieve tennis balls. Smells being used in the training process are hidden in drainpipes behind the wall. When the dog detects the target smell and indicates (for example by sitting or barking), a tennis ball drops down the drainpipe and the dog is allowed to play with the ball. This reward provides the positive reinforcement for a correct detection.

The Dogwall, with 16 potential odour sites

An important element of the training is the automation of the system and the use of the Dogwall app. The beauty of this system is that it separates the trainer from the training stimuli. The dog does not associate the handler with the training and does not use handler odour cues or behaviours to help make a correct decision. Apart from the fact that this training method provides a faster, truer learning curve, once the dog has completed the training program, it allows the dog to work with any handler, even a complete stranger.

The pipework and automation behind the Dogwall, for housing odours and delivering tennis balls

What can the dogs detect?

As we all know, a dog’s sense of smell is far more sensitive than that of a human. Mr Clarke draws this analogy. “If humans had visual skills similar to the dog’s odour sensitivity, we would be able to spot a single red grain of sand on Bondi beach! For example, the sense of smell is such that dogs are able to detect termite damage that may be many years old, as it still holds traces of termite odour.”

By tapping into this sensitivity the right dog can be trained to a variety of stimuli. The specificity of the stimulus depends on the detection work the dogs will be required to carry out. For example, you could train a dog to a particular termite species or even just to one colony of a particular termite species, but that would not be very useful. For termite detection, it is necessary to include a variety of termite species from a number of different colonies, as well as termite mudding and termite damaged timber to ensure the dog is correctly trained.

An important part of the training is to provide distractions, to make sure they will still indicate to the correct smell under different conditions such as at night, during rain and with loud music playing. It may also be necessary to train the dog to exclude a particular smell. “We had a dog that was particularly food driven and would become distracted by food sources in the house. So we trained the dog to detect termites in the presence of food, including dog food. As a result of this ‘exclusion’ training, the dog could successfully be used in houses with food sources present.”

The beauty of this training technique is that you are able to train a quality dog to a variety of stimuli, not just termites – bed bugs, rodents, fleas and more. Biosecurity Queensland is already using dogs in the detection of fire ants. However, when training a dog to multiple stimuli you do need to consider whether the combinations may be suitable or provide a conflict. For example, training a dog to termite and rodent smells may give rise to a lot of ‘false positives’ when using the dog in termite detection (when you are not really interested in rodent detection). However, a combination of termite and bed bug detection may be suitable.

In fact, the use of detection dogs is really only limited by your imagination, with potential uses in agricultural and environmental fields in the detection of pests, fungi and weeds. Mr Austin used the Dogwall to train a dog to detect Hawkweed, an invasive weed in the Snowy Mountains. On the first training run when the dog was released in an area previously cleared of the weed by humans, the dog found 20 plants still present!

Dog breeding program

Although the Dogwall is an excellent training tool, it is still important to have dogs that are willing to learn and good at detection – this comes down to breeding. Mr Clarke uses Cocker Spaniels, as they naturally want to fetch. At the start of the program, only about 10% of puppies were successfully trained into working dogs. To improve this Mr Clarke set up a breeding program. By using Cocker Spaniels that have high energy and a strong chase and retrieve drive, the success rate has gone up to 40% and indeed the latest litter achieved an 80% success rate.

You can train your own dogs

The beauty of this system is that it can allow businesses and organisations to train their own dogs. Mr Clarke intends to sell modules of the Dogwall (you will need around six modules/drainpipes), along with access to the app and training videos. Not only does the app provide separation between the dog and trainer, it actually contains a training guide, taking the dog and trainer through the training processes and recording performance. Once the dog has passed through the three training levels for a particular stimulus, they are ready to go to work.

Not only will this dramatically reduce the cost of a detection dog, but it also gives the opportunity to re-train dogs to new stimuli as business needs change. Re-training of a quality dog can happen in a matter of minutes (depending on the stimuli).

Mr Clarke recalls the time he lost his glasses in a paddock. “I simply trained the dog using a spare pair of glasses, it took five minutes. A minute later I had my lost glasses back!”

Mr Clarke firmly believes detection dogs have the potential to change the way the industry does pest control, but as with the switch to any new technology the first step can always be the hardest. However, with the Dogwall, Mr Clarke believes bringing detection dogs into a business has become a whole lot easier, allowing companies to tap into the significant benefits a detection dog can provide.

More research on termite detection.

More information on termite inspections.

Choose Your Country or Region

Asia Pacific