Which bracket failures belong in our FR5 fixture study?

Yes, ordinary brackets too. The ugly ones cannot tell the whole story.

Use all routes for the batch count. A smaller identified sample can support detailed fixture work, but keep its selection explicit. Those are different amounts of work for whoever has to collect the observations.

Bruno, hole spacing outside requirement after release, passing after correction. Tess, we have identities and dispositions for the whole batch already; the missing work is observing fit-up and the fixture contacts on current parts.

And include the correction person's time, unless you want that job to disappear from the estimate again.

The incoming fit-up check needs to name the features and how the part is supported. A note saying it looked aligned in the fixture will not give the designer a useful before-reading.

Were the released measurements taken with the same inspection support before and after correction?

I would ask the process specialist to choose the sample, not just whoever has a free morning. They may need the less obvious cases too.

Liam, with quality involved. The person choosing samples needs to know what actually passed.

Jess, same inspection support and revision H for both released checks. Anika, correction remains in our cost sheet. Stephanie, the incoming check is defined separately; we are not copying the final values into it.

Robin, could you start with the people already doing that correction, then let process and quality choose the comparison from their work? They may know which awkward fit-up never makes the final sheet.

That is useful input, Yasmin, but I would not let recollection select the entire sample. The unnoticed first-time passes still need a route into the study. Otherwise the spreadsheet improves while the selection bias stays put.

Robin, have you reconciled how many passed first time, how many needed correction and how many remain rejected? That can be done from the identities you already have.

Not yet. I have the rows, but some bracket identities occur twice because of the second inspection. I will count the parts through their routes rather than total the pass rows.

Tess, fair point. I meant include their observations, not let the best-remembered disasters become the whole sample.

Count reconciled with quality: eighteen first-time passes, five corrected passes, one rejected bracket, twenty-four parts. Process and quality have chosen the current comparison using that route breakdown, including ordinary passes. Observation still to do.

Thanks for answering the support question. Will the correction technician be available during that observation, or are their explanations still coming second hand?