The old samples were not individually marked. I can arrange identified blanks for a new manual study with the setter and quality.

A study I saw used two convenient reference faces for projection, depending on who held the gauge. We almost chased the chuck over it. Agree the reference and measurement method before collecting the new rows.

Quality has specified one reference face and a method. The setter is defining the approved seating observations; I am not adding robot loading to that study.

Can you record any chips or damaged ends found too? I mean what the setter observes, not just a box saying the area was checked.

And record any clean-up or reseating between readings. If the second measurement follows an unrecorded intervention, you'll be back to guessing why it changed.

Are you retaining the blanks until quality finishes the comparison? A surprising row is much easier to investigate when its part has not joined the next batch.

Yes, retained by identifier. The first results show length explains only part of the projection variation. Some seating observations differ too; no single cause established.

Was the reference-face method used consistently in those results? Interested because that was the mundane failure in the study I mentioned, and it wasted a lot of clever discussion.

Quality confirmed the recorded method was consistent. They are reviewing the seating cases with the setter before defining another study. Robot sequence design remains waiting for that result.

Thanks for explaining the length result. I would like to know what they find in the seating cases too.