Our UR5e inspection screen adds the last bracket again after reconnect, although the PLC completion was already high. The developer suggests remembering the previous bit. Doesn't that memory disappear on an application restart?
We had a counter behave after reconnect and fail after the application restarted. The remembered bit only lived in memory. Ask what identifies the completed attempt and whether the application has a durable record that it already counted that attempt.
There is an attempt reference beside the PLC result. Our database stores it, but reconnect goes straight to incrementing the displayed total without checking that record.
Then the developer has something stronger than the bit to use. Reconcile the retained result with its stored attempt before counting it. Keep the raw observation available, including whether the application has already handled it.
No, Amara. The screen keeps its own saved total. Developer is replacing that with a total derived from the result records and checking repeat delivery of the same attempt.
Offline test now keeps the count unchanged for an already-recorded attempt, after reconnect and restart. A genuinely new result still adds one. Interface recovery checks are not finished (we haven't changed any PLC clearing rules).
Have the handover show how an unknown retained attempt appears too. The known duplicate is fixed in your test; a result absent from local history still needs reconciliation rather than an automatic guess.