SMART MACHINES ON THE FARM
Who will plant and pick the crops in the years ahead? On farms all over the world, there are simply not enough workers, and experts at the Food and Agriculture Organization (FAO) warn that the gap between the jobs available and the people willing to fill them will keep growing. There is, however, some good news. Smart machines guided by artificial intelligence (AI) could help farmers produce as much food as before with far fewer hands, and some of these machines are already at work in the fields.(lines 1–6)
The story of computers on farms goes back about 30 years. In the 1990s, a handful of large farms in Australia fitted their tractors with satellite navigation so that the machines could follow straight lines across huge fields. The early systems were so costly that only the richest farmers could afford them. As prices dropped, the technology spread, and the tractors gradually learned to steer themselves. On most big farms in developed countries, self-driving tractors have now replaced tractors driven by people, and the fields are often plowed and planted through the night.(lines 7–12)
The maps that these tractors make as they cross a field have turned out to be useful in another way. Researchers recently combined them with millions of plant photos to train AI programs to tell young vegetables apart from weeds. Small robots that run on this AI now travel along the rows and burn every weed they spot with a narrow laser beam, so no chemicals are needed. "It takes a fraction of a second," explains agricultural writer Daniel Morris. "Still, farmers walk through the rows afterwards to check that nothing has been missed."(lines 13–18)
Another use of AI on farms is in drones, small flying machines with cameras. Many farmers already send drones up for simple jobs, such as counting the fruit trees in an orchard. More advanced drones can be programmed with information about a particular field. Each morning, they fly the same route and compare the new pictures with earlier ones, noticing changes that a person might easily miss, such as leaves turning yellow. When a drone finds signs of disease, a warning appears on the farmer's phone within minutes, giving the farmer time to treat the plants before the disease spreads.(lines 19–25)
What do farmers think of all this? According to several surveys, most of them are happy to hand simple jobs over to drones, and a large majority have a positive view of weeding robots. Most of those who have tried advanced drones describe the warnings as useful. On one question, though, almost all farmers agree: the decision about when the crops are ready to be picked must be made by people, not by machines.(lines 26–30)
Does this mean that farms will one day run without people? Morris doubts it. "Ripe peaches and berries are so soft that a machine can easily damage them," he points out. "And when a sudden storm or a new pest appears, only an experienced farmer knows what to do." Even so, he expects AI to change farming deeply. "Machines will take over many of the hardest jobs in the fields. Farmers will then be able to grow more food and spend more time planning for the future."(lines 31–35)