The shared tray station has needed repeated pickup changes for polymer housings. Tray looks unchanged, but the held housing looks less square. Before we blame FR10 drift, what observation would separate a seating problem from something happening during the grip?
Compare the housing seated before contact with the held position, using a safe external view or the approved inspection arrangement. Retain the program edits and as-found tooling identities too. A stable-looking tray outline does not establish that its inserts or contact faces are unchanged.
Current clip starts after lift, so it cannot show which stage introduces the lean. Maintenance has saved the edits and will inspect the tray inserts and finger contacts. No finding yet. I have stopped describing the tray as checked just because it looks still.
Use the actual affected housings in that review if they were retained and identified. An unlabelled fresh handful might show normal handling without explaining the observation that prompted the pickup edits.