Start with what each observed condition means and who is authorised to respond. An empty source pocket and an uncertain interrupted transfer may look similar from a distance, but they do not justify the same recovery decision.

I would ask operators to describe what they think the current message permits. That is slightly different from asking whether the instructions are clear; people can be quite certain about an interpretation the programmer never intended.

Also separate the saved sequence position from the block's physical location. Knowing the next instruction does not prove that the previous transfer finished.

Ada and Liam Baker, our message just says pocket empty after either case. Hazel, one operator thought it meant skip this pocket; another waited because the block might already be held.

Our sleeve lesson needed a deputy as well as clearer cases. We walked through the example call and checked they received the identifiers before arranging an installed exercise. A contact name without somebody able to interpret the call was not much help.

Who gets the call when you're unavailable, and do they have the same information?

Hassan, yes, and the deputy also read empty as permission to skip. Liam Allen, our integration lead owns the recovery design with the application assessor. They've separated known empty-source handling from uncertain transfer and agreed when we stop and escalate.

Ada and Hazel, that's in the planned supervised exercise: known empty source, interrupted transfer with location uncertain, and the escalation boundary. The deputy will take the example call instead of me.

The installed exercise caught one more wording error: our example sheet called the displayed next pocket the current pocket. Nobody resumed the uncertain case, and the deputy got the right tray identity, but we're correcting that example before issuing the handover.