One loose panel on a bench. No stack trial shown.

I would compare manual preparation as one option. If someone can present separated panels at another station, that may reduce the automatic separation work, though their time and space need to be included.

That doesn't remove the need to know one panel arrived. A person can put two down too. I'd still ask how the proposed system detects a doubtful delivery.

Agreed about the check. I was suggesting another way to feed the tool, not counting a person's preparation as proof of one panel.

Our sample sheets had interleaves and a little oil between some layers. No idea whether yours do, Hana, but I'd send stack conditions as well as panel dimensions. The lovely loose piece misses all that.

Support deserves its own trial as well, since a panel that stays attached can still bend enough to arrive badly positioned in the receiving fixture.

Include the receiving supports in the demonstration, so release is checked where your panel will actually land.

Zara, our stack does have interleaves; I missed them in my description. Caleb, no cup-access detail yet. I have asked for the stack and receiving fixture to be included.

Who removes those interleaves in the proposed process? It may be a separate task from separating panels, and the person doing it needs an agreed place to work.

I'd also have the integrator show the response to a double-panel indication or uncertain reading. Actually, show the recovery workflow too, not just a lamp in a demonstration.

Don't assume a double-sheet sensor can simply be bolted wherever there is spare room. Ask the supplier to validate its proposed check with these panels and surfaces. I don't know enough about your coating to pick a method from here.

I'd ask about worn cups and the cleaning routine while they're designing it. Those affect who gets called when the same stack stops behaving, even if the new-tool test goes well.

Sarah, manual interleaf removal at a separate preparation position is now one priced option. Stack trials and the single-panel check are still to be defined. Thanks for making that hidden task explicit.

Will you return with the support result when a representative transfer is tested, Hana, particularly whether the panel stays suitable for that receiving fixture?