Both cover-family inserts fit the carrier and either looks plausible backwards, which is a generous interpretation of interchangeable; what should our changeover review tie to the selected recipe before automated use?
Separate the family and orientation questions, because the right family can still be fitted the wrong way round; what does the current station actually detect?
The selected recipe also needs a documented relationship to the installed insert identity. Preventing reversal mechanically would solve only one part. Include removal, replacement and unreadable identity in the proposed sequence before choosing an implementation.
Lin, the recipe currently remembers the last selection, which would be another hole even after the insert gets an identity check; I've added that to the review.
Alex, has the designer proposed anything for the reversed fit yet? Interested whether mechanical changes can remove that choice without making replacement awkward
When a proposal exists, test every permitted family against each recipe, including rejection cases. A demonstration with the right insert and right recipe would leave the original concern largely untouched.