Case Study: An Evaluation of Detection Dog Generalization to a Large Quantity of an Unknown Explosive in the Field

Odour generalisation and detection dog training

Detection dogs are required to search for and alert to specific odours of interest, such as drugs, cadavers, disease markers and explosives. However, the odour released from different samples of the same target substance will vary for a number of reasons, including the production method, evaporation, degradation, or by being mixed with extraneous odours. Generalisation, the tendency to respond in the same manner to stimuli which are different - but similar to - a conditioned stimulus, is therefore a crucial requirement for working detection dogs. Odour is a complex modality which poses unique challenges in terms of reliably predicting generalisation, when compared with auditory or visual stimuli. The primary aim of this review is to explore recent advances in our understanding of generalisation and the factors that influence it, and to consider these in light of detection dog training methods currently used in the field. We identify potential risks associated with certain training practices, and highlight areas where research is lacking and which warrant further investigation.

Calibrating canines-a universal detector calibrant for detection dogs

Since the advent of the Universal Detector Calibrant (UDC) by scientists at Florida International University in 2013, this tool has gone largely unrecognized and under-utilized by canine scent detection practitioners. The UDC is a chemical that enables reliability testing of biological and instrumental detectors. Training a biological detector, such as a scent detection canine, to respond to a safe, non-target, and uncommon compound has significant advantages. For example, if used prior to a search, the UDC provides the handler with the ability to confirm the detection dog is ready to work without placing target odor on site (i.e., a positive control), thereby increasing handler confidence in their canine and providing documentation of credibility that can withstand legal scrutiny. This review describes the UDC, summarizes its role in canine detection science, and addresses applications for UDC within scent detection canine development, training, and testing.

Development of an automated human scent olfactometer and its use to evaluate detection dog perception of human scent

Working Dogs have shown an extraordinary ability to utilize olfaction for victim recovery efforts. Although instrumental analysis has chemically characterized odor volatiles from various human biospecimens, it remains unclear what perceptually constitutes human scent (HS) for dogs. This may be in part due to the lack of methodology and equipment to train and evaluate HS perception. The aims of this research were 1) to develop an automated human scent olfactometer (AHSO) to present HS to dogs in a controlled setting and 2) use the AHSO to evaluate dogs' response to different scented articles and individual components of HS. A human volunteer was placed in a clear acrylic chamber and using a vacuum pump and computer-controlled valves, the headspace of this chamber was carried to one of three ports in a different room. Dogs were trained to search all three ports of the olfactometer and alert to the one containing HS. In Experiment 1 and 2, the AHSO was validated by testing two dogs naïve to HS (Experiment 1) and five certified Search and Rescue (SAR) teams naïve to the apparatus (Experiment 2). All dogs showed sensitivity and specificity to HS > 95% in the apparatus. In Experiment 3, we used a spontaneous generalization paradigm to evaluate generalization from the HS chamber to different scented articles exposed to the same volunteer and to a breath sample. Dogs' response rate to the different scented articles was < 10% but exceeded 40% for the breath sample. In Experiment 4, we replicated this result by re-testing spontaneous generalization to breath and when the volunteer had breath exhausted/removed from the chamber. Dogs' response rate to breath alone was 88% and only 50% when breath was removed. Altogether, the data indicate that exhaled breath is an important and salient component of HS under these conditions.

A laboratory model of canine search vigilance decrement, I

Previous studies have found that infrequent targets can reduce dogs' vigilance. The purpose of this study was to develop a laboratory model to evaluate the effects of infrequent targets on dogs' search behavior and performance. Dogs (n = 18) were trained to detect smokeless powder in an automated olfactometer in two distinct rooms ("operational" and "training"). During baseline, the dogs received five daily sessions at a high target odor frequency (90%) in both rooms. Subsequently, the frequency of the target odor was decreased to 10% only in the "operational" room but remained at 90% in the training room. Last, the odor prevalence was returned to 90% in both rooms. All dogs showed a significant decrement in detection performance in the operational room when the target odor frequency was decreased but simultaneusly mantained high performance in the training room. This decrement was largely due to decreases in adequate search behavior. All dogs recovered performance when the odor frequency was increased again to 90%. Trial accuracy was associated with tail position, search score, latency, and duration of environmentally directed behaviors. The data show that low target odor prevalence significantly reduced search behavior and performance and that there are behaviors that can be used by handlers to assess their dog's search state.