The evaluation of canine training aids over time on working dog performance



Journal Title

Journal ISSN

Volume Title



Canines (K9) are the front line of defense in detecting narcotics by police and military working units across the nation. The way canines learn how to detect narcotics is by using small amounts of actual narcotics known as training aids or with training aid mimics. These mimics are synthetically made and replicate the targeted narcotic scent (pseudo). Even though there’re a number of research studies exploring canine detection in terms of olfaction capabilities, identification of target odor volatiles, to name a few, there are no existing studies evaluating the lifespan of narcotic canine training aids, defined as the length of time in usage during training. In this study, two types of training aids are be evaluated: real narcotic training aids (cocaine and heroin), that range in age from fresh to up to 10 years of age and commercially bought pseudo narcotic scents (cocaine and heroin). The main objective is to perform an instrumental evaluation in terms of the volatile odors being emitted for any given sample in relation to their respective age to see how their lifespan affects K9 detection. To do so, the evaluation period included: freshly acquired, 2 week, 4 week, 6 week, and up to 12 week old in order to observe the dissipation rate of chemical vapors in a controlled setting. Headspace analysis of all training aids was collected with Solid Phase Microextraction (SPME) and analyzed with Gas Chromatography-Mass Spectrometry (GC-MS) to obtain a chemical odor profile. Simultaneously, field testing with four local police department narcotic detection canines teams (all male Belgian Malinois breed of various ages) occurred to observe K9 alert performance with respect to training aid age. The hypothesis is that over time, training aid vapors decrease in concentration and provide less of the chemical odor signature for the canines to detect, which thereby decreases alert performance. The findings include primary Volatile Organic Compounds(VOC) (80% in quality abundance and frequency occurrence in narcotic types) for heroin: ethanone,1-(2-hydrophenyl) -, phenol, and undecane, 5-methyl-. Primary VOC’s for cocaine included: ethanol, 2-butoxy-, benzene, 1,3-dimethyl-, p-xylene. Using multi-factorial ANOVA modeling there was no statistical difference between the amount of these target VOCs in either heroin or cocaine with respect to their age or narcotic training aid type. Although no consistent chemical odor profile was found across the different types of training aids tested, canines were able to alert with great accuracy, obtaining an average of 96% Positive Predictive Value (PPV) in alerts for heroin and 100% PPV for cocaine. Average Negative Predictive Values (NPV) for heroin yielded 91% and 94% for cocaine. The variation of VOC odor vapor highlights how canines, no matter the age or narcotic type, are able to learn a distinct odor and follow its odor pattern as it changes over time to identify it. This directly impacts biological detection of narcotics by presenting scientifically based evidence on the chemical odor profiles of various drug types and their effect on efficient and reliable canine detection in the field. This research project attends to the need of standardizing forensic canine detection practices within the United States national defense at home and abroad.



Narcotic odor, Solid phase microextraction-gas chromatography/mass spectrometry (SPME-GC/MS), K9 training aid