Setter and quality have agreed the reference face and identified the retained blanks. They are measuring overall length before repeating the seating comparison.

Include the end that meets the stop. A raised edge there can make the same blank seat differently even with the reference face and jaw arrangement settled.

Can the second setter identify that face from the sheet, Callum? I mean without the first setter pointing to it.

Yes, Omar. We added a labelled view and the second setter used it correctly. The first repeat results mostly follow overall-length differences, but one blank still gives changing readings.

Keep repeating that identified blank, not another handful. Does it have an observable end condition different from the ordinary ones?

Also keep the setup identity on the sheet. If the next shift changes the jaw arrangement, they need to know they are starting a different comparison rather than extending this one.

Quality found a raised edge on the inconsistent blank's stop end. Setter can show it catching before full seating. The other blanks seat repeatably and their exposed lengths follow their measured overall lengths.

Have quality decide that blank's disposition before it returns to the stock. The new loading estimate should not treat its rejected end condition as ordinary seating variation.

Was that blank outside the incoming requirement, Callum?

Omar's question matters. Finding an awkward edge explains the observation, but quality still has to compare it with the actual requirement. If accepted stock includes it, the planned process needs to handle that reality.

Quality rejected that edge against our incoming requirement. It is separated from usable stock. The ordinary overall-length range is accepted, so we are carrying that range into the loading study.

Show the rejected example in the receiving instruction, with the call to quality. Do not make the loader responsible for inventing an acceptance limit from a photograph.

Agreed. My earlier wording got ahead of the actual disposition; Callum has answered it now.

Thanks for checking it, rather than calling every awkward blank scrap.

Comparison finished with the same identified blanks and both setters. Conforming blanks repeat from the agreed reference; rejected blank stays separate. The sheet now includes overall length, setup and end condition. Automatic loading remains a later assessment, but this measurement question is answered.

Useful result. Keep both parts of it in the estimate: real length variation and the separate rejected-end finding.