I wonder about affordability, but I am very pro-farmbot. Far more than self-driving cars. If anything, farmbots could provide a lot of very helpful groundwork

I wonder about affordability, but I am very pro-farmbot. Far more than self-driving cars. If anything, farmbots could provide a lot of very helpful groundwork in a lower-stakes environment which the self-driving cars people could use. My only real concern is if farmbots could be cheap enough to absorb a large portion of the -cide market. Most herbicides, for example, could be replaced by something which plucks. Insecticide money could, partially, go into bots that could pick-and-crush unwanted insects aboveground. We would still have grubs. Fungicide expenditure could be slashed: detect infection on a plant, spray it and its neighbors, maybe a few more downwind. Your bot, named He Who Walks Behind the Rows, trundles along slowly, observing each plant. Per acre, it would have to travel roughly three and a third miles. At a half-mile-per-hour pace, thirteen hours a day, you could swing two acres a day per bot. Let's say your bot eats your production cost in half, about four hundred dollars per acre. If a bot lasts only a season, well, it has to cost under a grand to be worth it, and I haven't even touched power costs yet. So it will need to survive multiple growing seasons. I can see Overlord stations every few acres, maybe grabbing some solar power or charging up off a grid, storing it so that it can be dispersed to the bots at night. They could serve as data collection points, centimeter-level GPS correction servers, and so on, as well. I could see the temptation to work irrigation into the bot somehow but water is heavy and I imagine something entirely separate to provide water. I'm making a lot of assumptions, like "each plant ought to be looked at each day" and the standard thirty-inch rows, etc.