I have put fixture design and manufacture into separate estimate lines. Operator time is harder for me to estimate because we haven't agreed how much setup the supplier will do.

Ask for that boundary explicitly. Setup can mean unpacked equipment or a demonstrated process, depending on who is saying it.

Their quote says installation assistance, with no task list. I'll ask which fixture, tooling and program tasks they own before assigning our internal hours.

Who would take over the pilot if your colleague's usual operator is absent? On our teaching jobs the apparent saving often depended on one person being available to explain every interruption.

Include handover in the trial, yes. Otherwise you're pricing a demonstration staffed by whoever built it.

For a first estimate, list preparation, supervised running, clearing a normal interruption through the approved procedure, and changeover as separate activities; people can estimate a task much more easily than a mysterious block called support.

And include the time spent returning the bench to manual use. The work still exists when the pilot stops.

I've added those activities to the draft. We have one operator and one maintenance contact proposed, but neither has reviewed the hours. I won't present those numbers as agreed staffing.

What are you comparing it against? I can run a small batch quickly if somebody else brings every part and clears the bench afterwards. Manual baseline needs the same boundaries.

Use a whole order as the comparison unit. The nicest transfer in the middle is only part of the bill.

Would the operator help record that manual order before the supplier visit? It gives them a proper say in what counts as work, and you get a baseline everyone can recognise.

That is going into the proposal. One block variant, whole-order timing, named setup tasks and a limit on staff hours. We still need the supplier's task list and my colleague's agreement before this becomes an approved pilot.

Include what happens when the hours run out without a convincing result. A small pilot should be allowed to end as inconclusive.