The limitations include: machinery damages the fruit, and the results are therefore great for juicing and processed food, which don't pay well; and not good for retail-type, which is where growers make their profits.
nobody to sponsor university challenges to make a sufficiently gentle applebot. the current stuff, what, just kind of rakes through the trees? artificial intelligence is prolly good enough to make a true picker now, should anybody care to.
oh, here's another idea, trained chimpanzees. they'll work for bananas.