CHINZY / FDE and robotics acceptance fieldbook / 01
Scanned part identifiers
Use part IDs differing only in zero and O. Keep the crop; require a source check if the scan is ambiguous.
Test setup
Freeze the corpus and mark the expected passages. Ask a reviewer to trace each answer to the original, recording missing evidence, wrong citations and correct abstentions separately.
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 |
|---|---|
clear=A0-17; PO=A0-17 | exact_match |
blurred=A?-17; candidates=A0-17,AO-17 | review_original |
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) ↓