CHINZY / FDE and robotics acceptance fieldbook / 06
False alerts and missed events
Replay labeled samples. Count alerts on normal samples separately from events that triggered no alert.
Test setup
Fix the route, checkpoints and task in recorded data or a controlled test area. Score arrival separately from usable evidence. Retain timestamps and samples requiring human review.
Contrasting examples
Figures and IDs are illustrative fixtures, not product results or contract terms. Save inputs, versions and conditions before repeating an unexpected result.
| Example input | Expected observation |
|---|---|
truth=normal; alert=yes | false_alert=1; missed=0 |
truth=event; alert=no | false_alert=0; missed=1 |
Recording the result
Keep a separate row for each example; one pass cannot stand in for the other. Enter the actual output, evidence location, environment version and review decision. Leave unresolved results open.
Download test worksheet (CSV) ↓