From pilot to production.
Use a pilot to learn what needs to change before extending automation to more work.
Keep the learning visible
A pilot should leave behind more than a demonstration. Record what was tested, what worked, what required an intervention, and what remains uncertain. Include feedback from the people who operated and maintained it.
Compare the outcome with the original question. If the scope changed along the way, explain why and record which parts of the original goal have not yet been tested.
Check what changes at a larger scale
Another shift may have different experience. Another line may handle different parts. More connected machines may change who needs access to records and who responds to exceptions. List these differences before repeating the installation.
Treat the next rollout as a defined extension, with its own acceptance criteria and owner. Reuse what is understood; make new assumptions explicit.
Make ownership durable
Put operating instructions, configuration records, and support contacts where the team can find them. Agree how changes are requested and recorded after handover.
Review the expanded system at an agreed interval. Continuous improvement works best when observations have somewhere to go and someone is responsible for acting on them.