LouisBrown0920 · 7 Dec 2025, 09:44 UTC
Have tooling review how to prevent reversal physically, and have controls define how the installed insert identity agrees with the selected recipe. A closed signal alone does not identify either family.
6 pointsTooling can review keying. Controls had assumed the operator chose the right family first.
21 pointsStores can make the identities legible too, but I'd keep that alongside the design review. A label on a bag will not tell the system what somebody fitted after throwing the bag away.
4 pointsInclude removal and replacement in the state review. We've had to keep an uncertain handling state through restart elsewhere; here an earlier correct insert reading must not quietly survive a changed insert as though nothing happened.
4 pointsLeo, that case was missing. I've added insert removal after a valid check to the review.
15 pointsAlso include unreadable identity and a readable but wrong family. The proposed response needs to remain clear when the system cannot establish agreement, not only when a known correct insert is fitted.
6 pointsWho will maintain the family and revision mapping, Louis? If stores receives a replacement insert later, somebody needs to say whether the existing recipe association still applies.
20 pointsTooling owns insert revisions. Controls owns the mapping, with tooling approval for changes.
18 pointsThat gives the test cases an owner. Have they agreed the uncertain-state behaviour too, or is the current proposal still only a way to read an identifier?
16 pointsStill a proposal. Wrong family, reversal and unreadable identity remain review cases, not passed tests.
6 pointsInclude those limits in the handover draft. The operator should not infer that a new readable label has made the planned automatic check ready for use.
14 pointsThis discussion is closed to new replies after six months without activity. Last activity: .