Just Say No! — Science and Training of Detector Dogs

Angel Landrau and Dr. Stewart Hilliard

This will be a two-day crash course designed to introduce detector dog operators (both amateur and professional) with the basic theoretical and practical knowledge they need. Dr. Hilliard will lead on theory, and Angel Landrau (MSgt US Air Force, Ret; previous MPC Teams Training Advisor for the US Army Special Operations Command, and current MWD consignment evaluator for the 341st Training Squadron) will bring his completely original approach to practical work. The course is organized around already-trained dogs—evaluating their accuracy and specificity on operations, identifying their weaknesses, and on-the-spot interventions and training plans to address those weaknesses. The orientation of this course is unconventional: In addition to the creating ”tuned” control of the dog’s alert and indication (or CoB and final response) by trained odor, we emphasize methods to test for and eliminate false responding controlled by distractor odors, masking odors, and other stimuli. In order to be accurate, operational detector dogs have to correctly say “No, my trained odor is not here” far more often than they have to correctly say “Yes, my odor is here.” This is why our motto is “Teach your dog to just say No!” and we will demonstrate procedures to assess how good your dog is at saying No and also improve this critical ability. Other unconventional ideas we introduce will include: a) Rewarding indication responses with random rather than continuous reinforcement; b) assessing the reliability of the dog’s alert versus the indication (which one provides the handler with the best evidence?); c) teaching the dog to be a specific discriminator rather than a non-specific “feature detector;” and d) isolating search training (long-duration sniffing and searching) from alert/indication training, so that handlers can adjust the number of hides they place in training to reflect operational conditions, rather than placing hides– far too frequently– just to reward their dog for search behavior. This seminar will be geared for K9 operators from law enforcement and military, but also welcomes trainers and handlers from private industry or sport training, depending on the venue.

Program Overview:

We will focus on core knowledge that trainers must have in order to understand detector dogs, and produce accurate and reliable real-world K9 detection performance. We will welcome handlers of any kind of detector dog—narcotic, explosive, sport, or conservation dogs, etc. This event will be more practically-oriented than standard Kynology events, perhaps ⅓ lecture and theory and ⅔ practical work and training. Number handler-d0g teams will be limited to eight, obersvers wil be limited to the size of the venue. Dr. Hilliard’s lectures will provide essential knowledge on the following topics:

1. Pavlovian Conditioning: Essentials of stimulus — Stimulus learning, and what exactly is odor recognition?
2. Instrumental Conditioning: Essentials of response — Consequence learning and practical applications to detection.
3. Alert vs indication — How do you know when your dog has encountered target odor? What is your best evidence on operations and for the courts?
4. Signal detection theory — The dog as a signal detector, and the effect of asymmetric reinforcement schedules for hits vs correct rejections
5. Stimulus control of detection responding — Generalization and discrimination, blocking, overshadowing, latent inhibition, sensory pre-conditioning, and relative validity– Why detection training goes right when it goes right, and why it goes wrong when it is a dumpster fire
6. Control by target (trained) odor — “Tuning” the detector dog’s generalization curves for accurate detection of real-world targets 
7. Control by non-target and other stimuli — Minimizing the influence of odors and other stimuli, especially the odors of delivery devices and other big &%$#ing technical problems, to prevent misses and false responses
8. Handling options for improving search behavior, improving accuracy when detecting, localizing, and indicating trained odor, and decreasing false responding.
9. Handling options for improving motivation and establishing multiple “levels” of motivation for different detection tasks