Why farm robots range from tiny to enormous
Agricultural robotics spans an unusually wide range of physical scale, from small experimental weeding robots that creep between rows of vegetables to massive autonomous tractors that can cost as much as a house and cover dozens of acres in a day. This spread exists because farms themselves vary enormously in what they grow, how much land they work, and what economic pressures they face. A large commodity grain operation values throughput above almost everything else, since profit margins per acre are thin and the crop itself is low-value per unit, so automation there tends toward scaling up existing equipment, autonomous versions of large tractors and combines that do the same job faster and with less labor, rather than toward fine per-plant precision.
A smaller specialty farm growing high-value crops, like berries or leafy greens, faces the opposite set of tradeoffs. The crop is worth much more per unit, labor for delicate tasks like weeding or picking is often expensive and hard to find, and the growing area per farm is smaller, so a compact, slower, more precise robot can make economic sense even though it covers far less ground per hour than a big tractor would. The robot doesn't need to be fast, it needs to be gentle and accurate, and the farm doesn't need to cover a thousand acres, it needs a machine sized to a few dozen.