No, it clears on a timer; the application can acknowledge later, but the PLC does not wait for that acknowledgement.

Ask them to walk one request through the two programs and state who changes each signal. Include a slow response and a stopped application, since extending the pulse alone leaves those cases unclear.

I have requested that sequence and asked the integration lead to own the shared description rather than keeping separate assumptions in two documents.

Do you have an observed request and application-read trace for a missed case? That would help the authors discuss the timing actually present instead of only their intended sequence.

Only the PLC trace so far; application-read timing is being requested, so I cannot yet demonstrate exactly which read missed the request.

Thanks. Has the shared clearing sequence been agreed yet?

Normal ownership is drafted, but clearing after timeout and restart remains unagreed; I am keeping the interface review open.