United News of Bangladesh

Researchers have developed an AI-powered robot that can harvest tomatoes more efficiently by deciding the best way to pick each fruit. The robot not only identifies tomatoes but also determines the easiest way to harvest them. This comes as farm labour shortages are pushing the agriculture sector towards greater use of automation, especially in harvesting. Tomatoes are difficult for machines to handle because they grow in clusters. A robot must carefully pick ripe ones without damaging unripe fruits nearby, which requires accuracy and smart decision-making. To address this, Assistant Professor Takuya Fujinaga of Osaka Metropolitan University’s Graduate School of Engineering created a system that helps robots judge how easy a tomato is to pick before attempting to harvest it. The system uses image recognition and data analysis to select the best angle for picking. It examines factors such as the position of the tomato, its stems, and whether it is hidden behind leaves or other parts of the plant. Unlike traditional systems that only detect fruit, this method introduces “harvest-ease estimation,” which measures how likely a robot is to successfully pick a tomato. During testing, the robot achieved an 81 percent success rate. In some cases, it adjusted its approach and successfully picked tomatoes from the side after failing to pick them from the front. The study highlights that several factors affect robotic harvesting, including how tomatoes grow, the position of stems, and visual obstacles like leaves. Researchers say this technology could lead to better coordination between humans and robots in farming. Robots may handle easier tasks, while humans focus on more difficult harvesting. The findings were published in the journal ‘Smart Agricultural Technology’. Source: Science daily