Our UR5e startup sends the bracket request a second time

It has the original request ID. The developer is disabling resend for that uncertain state.

Does the request ID reach the receiving application, or is it only a local log label? That decides whether the lookup can really find this attempt.

Toby, it reaches the receiver. Cecilia, the bracket is identified and held with the inspection lead.

Give the reconciliation work an owner and time, otherwise everyone will agree to hold it and nobody will have a job to finish

For the teaching example, show known never-sent and sent-but-uncertain side by side. The second one should not look like a broken version of the first. I would have the covering person explain the difference before showing the authors' answer.

Include a real never-sent request in the software test too. Removing every route to submission would stop the duplicate neatly and leave you with a very quiet inspection station.

Has the original lookup returned anything yet?

And keep the bracket's current disposition visible beside the lookup. A receipt alone should not make the held part appear accepted in the operator view.

Amara, receipt found, no result returned yet. Inspection lead owns reconciliation and the bracket remains held.

We had a queue of uncertain samples gradually renamed new work because nobody wanted the awkward list on screen. Same physical samples, cleaner-looking queue. I'd keep yours visible with the owner, even once the software fix makes new work behave properly.

Oscar Ali, leave the receipt as receipt found until its outcome is reconciled. The query has answered whether it was received; you still need the result and its link to this bracket.

Hazel, that's a useful example to discuss with the covering person. Also show a genuinely reconciled item leaving the hold list, so the lesson isn't that uncertain work must sit there forever.

Original result recovered and matched. Quality accepted that bracket and recorded its release from hold. Software checks continue.

Thanks for closing the physical bracket question too. Does the repaired startup preserve that result when reopened, rather than recreating the old pending entry?

Yes. Known result, uncertain send, missing receipt and genuine never-sent cases passed offline, including reopening.

Run those through the installed application with the maintainer before calling the startup repair complete. Include an unavailable lookup: it should preserve uncertainty, not quietly return never sent.

Installed checks passed, including unavailable lookup. Covering operator followed the uncertain hold and recognised the reconciled result without prompts.