Can't decide which pocket recovery should offer

The placement routine returning. No separate pocket check. Our programmer agrees that completed is too strong a label for that alone.

Yes, I'd be uncomfortable asking an operator to trust that label.

We've sketched an interrupted-transfer view. It shows the requested pocket, last saved stage and a warning that the part location needs reconciling. No automatic next-pocket choice after uncertainty.

Can the operator actually see that pocket from where they use the screen?

Oh, that would ruin your afternoon. Were the trays reversible too?

I'd rather pay for a clearer tray than another screen full of state names. People still have to identify the pocket you're asking them to inspect.

The back row is hard to see from the screen, Maya. Harish's numbering point helps too. Our sketch had row letters that never appeared on the tray. We've asked for the physical layout and recovery view to be reviewed together.

Who handles an interrupted tray if the operator cannot establish where the block ended up?

Maya, yes, ours fitted either way, which was the other half of it; we changed the locating arrangement with the fixture designer, then had to redo the training pictures.

Our shift lead takes over uncertain cases under the recovery procedure being written. We are not asking operators to reach into the station to see a hidden pocket.

Thanks, Hazel. That was the awkward bit I was trying to get at.

Have the shift lead try the draft too, because ours found the confusing tray view faster than any of us who wrote it.

I'd still challenge how often this needs a shift lead. A procedure can be correct and still make every interrupted tray expensive to deal with.

We walked through the draft with both shifts. A clearly completed placement can be reconciled using the agreed checks. Unknown location stays held for the shift lead. We also found the screen retained a pocket highlight from the previous tray.

That's the sort of detail I'd want in the costing trial. How long does the ordinary, unambiguous case take now?

Does changing trays invalidate the saved recovery context, or can that old highlight return with the previous tray's data?

Please test that one; a convincing wrong highlight is worse than no highlight.

Tray identity wasn't attached to the highlight. Programmer is fixing that and adding the tray-swap interruption test. Aisha, we haven't timed the revised recovery yet. I won't put the old clean-run figure against it.