What molecular detection-dog training is

What a “molecular” dog is

A dog's sense of smell is vastly more sensitive than ours and can pick up an odour even at very low concentrations. We call it “molecular” because it works at the level of the odour molecules that make up a target's scent — a pest, its larvae, its pheromones.

Trained correctly, the dog tells that odour apart from thousands of others and actively searches for it, marking precisely where it finds it.

Imprinting on the target odour

The first step is imprinting: the target odour is presented to the dog in a controlled way and a positive association is built. From then on, for the dog, that odour becomes something worth searching for.

Positive reinforcement

Training is based on play and reward: when the dog finds and marks the correct odour, it gets its reward. No coercion — the dog works because, to it, the search is a game.

The alert (the “marking”) is a clear, repeatable behaviour — for example sitting or freezing on the spot — so the handler is in no doubt about what the dog has found.

Generalisation and “proofing”

To be reliable in the field, the dog learns to recognise the odour across different contexts, amounts and environmental conditions (generalisation) and, above all, NOT to alert on similar odours or distractors (proofing, or discrimination).

This is the stage that cuts false positives and makes the dog's work usable to support plant-health decisions.

From the training ground to the real field

In the field the dog-handler pair is what counts: the dog searches and alerts, the handler reads the alerts and turns them into useful data — georeferenced points, maps, reports. Training and verification continue throughout the dog's working life to keep reliability high.

The method we use too

This is exactly the approach we work with: starting from the target odour — in our case the pest's pheromones — we train the dog to locate it in the soil and on the plant, supporting early detection in plant-health emergencies.