Some notes only identify the batch. One route card does name a particular bracket.

I would start with that card, not average the batch notes into a story about it. We found a retained sample from a later run sitting beside earlier inspection sheets once. Perfectly real records, wrong comparison.

Does aligned mean a measured result on that card, or a visual setup observation? Those could be quite different kinds of evidence.

Visual observation. There is a separate angle reading, with no measurement method written beside it.

Then don't price the new fixture as reproducing that angle. You haven't established what it measures yet.

And who is allowing for the correction labour in the proposal?

Jane's point doesn't require throwing the reading away; it can stay in the review as an observation whose method still needs explaining, rather than a target somebody should reproduce

Can the original inspector identify that method while they're still available? It might be easier to ask about one particular card than hand them the entire folder and ask what used to happen.

They identified two setups. Restrained reading used the fixture face. Released inspection used the drawing datum arrangement.

That explains why subtracting the two numbers would be a dubious shortcut. Do you have enough recorded geometry to relate the setups, or only the readings?

Ask the measurement owner to establish that relationship. Or, if the information is missing, state that the historical change cannot be calculated from these notes alone.

Only readings survive. Quality won't translate them into a release change. That comparison stays unavailable.

You can still report the released nonconformance against its documented requirement, if quality confirms that record; unavailable change isn't the same as no evidence of a problem

So can the new estimate carry a stated allowance pending the trial?

Liam, why are we back to a price before the trial scope is defined? That's how the missing work becomes someone's free afternoon.

Not a guessed fixture price. An explicit uncertain allowance, with the study priced separately. Should've said that.

The study can specify the missing evidence now: identified incoming pieces and fit-up, documented fixture condition, then released measurements under the agreed method. The welding and measurement specialists need to set that scope. None of that requires pretending the old figures answer it.

Agreed sample plan requested from both specialists. Cost uncertainty remains explicit. No live welding changes made during this records review.

Does the plan keep any later correction and reinspection identifiable too? You don't want the next retained sample to lose the same part of its history.