Spotting dog clues during deliveries

Context

Safety content is only useful if it changes what someone notices before the risky moment.

Delivery partners operate in changing environments where a dog may be visible, hidden, loose, behind a fence, or only suggested by environmental clues. A passive page can tell someone to be careful. It cannot tell us whether they can recognize the signal and choose a safer move when the situation is ambiguous.

Problem

The design problem was to move from information exposure to perceptual and judgment practice. The learner needed to scan a scene, identify meaningful clues, decide how much risk was present, and choose the next action before receiving feedback.

Design decision

Practice the noticing, not just the policy.

The scenario structure gives the learner a realistic delivery context first. Information is not front-loaded as a lecture. The learner has to inspect the situation, form an interpretation, and act. Feedback then connects the choice back to the safety principle.

Framework

This project also served as a pilot for a broader interactive-learning approach: use short realistic situations, ask the learner to make a meaningful decision, provide feedback at the moment of judgment, and keep explanatory content available as support rather than making it the entire experience.

Evidence

The evidence shown publicly is the design logic and working scenario structure. I do not publish internal safety metrics or imply a causal performance result that cannot be supported publicly.

Reflection

People cannot act on a clue they never learned to notice.

That is why safety learning should include perception and judgment, not just policy recall. The most useful practice often begins one step earlier than the final behavior.