Search Results
6/27/2025, 11:11:36 PM
>>508901930
Depending on the economy we're something like two years away from fully automatic harvesting robots, they already have some for crops easier to pick but they've been slapping AI systems on robotic arms and teaching them to go pick fruit one by one right on the tree for apples, tomatoes and stuff.
So far it's expensive and slow as shit, still leaves a third of the harvest hanging there so you still need some wagies to finish the job, and it's all made by nerds who never picked a single fruit so nobody wants to be the dumbass who buys the first generation, but with the job market being what it is and possible economies of scale if everyone in California starts investing at once because it's either this or letting next year's avocados rot in the field, unless they suddenly change their minds and import a billion indians to do it you should see drones pop up in a few seasons.
the main hurdle isn't the science (real-time motion planning, SLAM and image processing are far from trivial but it's doable given enough time, processing power and engineers), it's how much we really need that shit economically.
Depending on the economy we're something like two years away from fully automatic harvesting robots, they already have some for crops easier to pick but they've been slapping AI systems on robotic arms and teaching them to go pick fruit one by one right on the tree for apples, tomatoes and stuff.
So far it's expensive and slow as shit, still leaves a third of the harvest hanging there so you still need some wagies to finish the job, and it's all made by nerds who never picked a single fruit so nobody wants to be the dumbass who buys the first generation, but with the job market being what it is and possible economies of scale if everyone in California starts investing at once because it's either this or letting next year's avocados rot in the field, unless they suddenly change their minds and import a billion indians to do it you should see drones pop up in a few seasons.
the main hurdle isn't the science (real-time motion planning, SLAM and image processing are far from trivial but it's doable given enough time, processing power and engineers), it's how much we really need that shit economically.
Page 1