Our UR5e scratch folder has shifting bright bands labelled as damage. The number visible in every image belongs to the holder, not the trim plate. I cannot link several labels to a physical inspection decision.
Separate the images whose physical plate identity can be established from those that remain unlinked. Then have quality relate actual inspected locations to the identified plates. A common holder number should not become a guessed sample history merely because it is the only legible label.
Two plate links established from the original capture notes. Other old examples remain unlinked. I've removed those from the working references without inventing new sample numbers.
We had a stores description that followed several different physical pieces around and looked wonderfully specific on the paperwork. Your holder number sounds similar: useful information, just attached to the wrong level of the problem. Keep it as the holder identity rather than throwing that bit away too.
Get the appearance method stated with quality's decisions. A moving bright band may be presentation-related while another faint mark on that same plate is a real reject.
Keep the unknown images marked unknown in the archive, so the next helpful person does not restore them to the lesson folder because the original photographs look sharp
Quality accepted one linked plate under the defined method. The other has a faint rejected line near an edge. Neither physical decision follows the moving bright band. Both actual plates are retained for presentation checks.
Check those plates again after repeated handling as needed, not only before it starts; if the comparison adds a mark midway, the earlier accepted decision should not silently describe the altered surface.
Who is doing the repeated seating, Beth? I'd want the ordinary loader involved, especially if a good-looking view depends on knowing which corner to nudge first.
Loader and vision technician together. The holder permits two seated orientations; one view hides the rejected edge behind its lip. Same plate, very different usefulness.
Then that hidden-edge view cannot establish inspection coverage for the job. Keep it in the comparison as the failed candidate rather than choosing it because the accepted surface looks calmer there.
Not yet. The underside mark is invisible when seated. Tooling is adding a visible orientation reference to the proposed arrangement; the camera-only instruction was rejected at the walkthrough.
Good that the walkthrough caught it. I wouldn't make the loader responsible for remembering a hidden mark simply because the people who built the example can picture it perfectly.
Keep the chosen presentation tied to the complete holder arrangement and the physical defect locations when its review is completed. The visible orientation reference helps loading, but it does not by itself establish that all required surfaces remain observable.
Revised reference fitted and the loading comparison repeated. Both loaders selected the intended orientation without prompting; the retained rejected line stayed visible. Quality confirmed the plates' surface condition after the check. Unknown old images remain excluded.
Representative accepted finishes are in the next set, with their identities and judgements established. Tuning is still open. The holder and reference errors are corrected; we're not calling a two-plate comparison the whole inspection validation.