It is generally understood that dogs are more sensitive vapour detectors than any artificial equipment. That said, dogs (and rats) may be less reliable than machines and it is difficult to determine which scent animals are tuned in on. The low detection threshold of dogs and rats makes it difficult to measure detection capabilities for these animals. However, recent developments in artificial scent detection technology may be approaching the limits of dogs and rats, giving the potential for independent assessment of the capabilities of these animals under some circumstances. This is important if we want to establish detection benchmarks for dogs, rats and chemical detectors.
The GICHD has, with help from Sandia, undertaken a small project where the aim was to compare the explosive odour threshold for rats and dogs. Although we were not able to establish the various thresholds scientifically, we were able to loosely compare the detection levels for dogs and rats when it comes to detecting explosive vapour and other substance from landmines. The results showed that capabilities could extend down to theoretical limits of single molecules per inhalation. However, not all individual dogs and rats showed this capability, and it was not possible to estimate of reliability at these low levels. These initial tests also showed that animals often find alternative cues, other than the chemical vapour, to achieve the reward and confound the test (or training) objectives. More tests are required to determine detection threshold for dogs and rats with some degree of reliability.
A simple yet robust vapour sensing method should be developed both for testing purposes and as a training aid in free running and REST applications for dogs and rats. The method needs to include both a variable vapour source and a testing regime that eliminates the use of secondary cues for reward by the dogs. The vapour sources can also be developed for use as a training aid to enhance the capabilities of the dog handler/trainer. Testing of vapour sensing threshold reliability could be complementary to performance tests of field dogs in the future. This would enhance the evaluation of missed mine incidents for causes related to insufficient vapour or reduced vapour sensing performance of the dog.
