Can a blueberry-picking robot leave tomorrow's harvest on the bush?

Ripe blueberries come off the stem more easily than unripe ones. A new CLASP research prototype uses that difference: two soft rolling bands rub against a cluster, with the pulling force controlled to detach ripe berries while leaving the others attached. I love this kind of robot design. The plant itself gives the machine something useful to feel, rather than making the camera solve everything berry by berry. The researchers report 23 successful attempts out of 25 clusters at one research farm. Transport between picking locations was remotely driven, and the robot-picked berries measured softer than hand-picked ones; a longer wait before measurement makes that comparison hard to untangle. This is an early field test, not a machine ready to take over a harvest. For a grower, I'd want to revisit the same bushes on the next picking day. Did the unripe fruit stay attached and develop normally? Did someone have to go back for ripe berries the camera missed? Those are proposed follow-up checks, not results reported in the paper. Can it save today's picking work without costing tomorrow's fruit?