WUR-researcher: ‘Selective harvesting is the biggest challenge for robots’
The greatest labor savings in agricultural robotics can be achieved by automating selective harvesting, according to Erik Pekkeriet from Wageningen UR. However, this also poses the biggest technical challenge.
Pekkeriet leads the Vision and Robotics program team at Wageningen UR. He is confident that robots will continue advancing. “A lot is already possible, and things are moving in the right direction,” he says about agricultural robotics in general. “It depends somewhat on the type of farm. In fruit cultivation, for instance, there’s still much to be done in terms of robotics, while simpler tasks such as soil preparation and weed control in arable farming are already progressing well.” However, it’s clear that when things become complex, someone usually needs to stay nearby or be able to take over the tractor remotely. This is especially true in cases with many obstacles, mud or weeds, stones in the field, or with tall crops and orchards.
Autonomous bulk harvesting is technically feasible, though it requires investment
Technical challenges in selective harvesting
In selective harvesting, the goal is to pick only ripe products—ripe strawberries or cauliflower heads of the correct size. “Autonomous bulk harvesting is technically feasible, though it requires investment. But selective harvesting is not yet practical, even though this is where the most labor is needed,” says Pekkeriet.
Future of robotics
Robotics is progressing in areas where automation is technically easier. Pekkeriet believes this development will continue in the coming years. He outlines two primary drivers for robotics: labor savings and sustainability, which often overlap. For example, reducing pesticide use goes hand-in-hand with increased labor for weeding. “If we want to further improve sustainability, the required labor time will skyrocket,” explains Pekkeriet. “I believe that’s where robotics can make a real difference.”
Ease of use and data collection
Pekkeriet emphasizes that systems still need to become more user-friendly for farmers to fully adopt them. If a farmer constantly needs to return to the field to reposition the robot, it remains labor-intensive. “From a business perspective, it’s not practical. Ideally, you’d want to control it from your desk.”
Additional issues include robots that frequently stop, safety concerns, the labor-intensive process of setting up new fields, and creating the necessary task maps, which requires many steps. “Every new situation, such as crop stage, a different crop type, new tools, or changing weather conditions, requires new adjustments,” Pekkeriet notes. “This leads to malfunctions and learning curves.”
Future role of robots
Pekkeriet sees another value robots will add in the future: they collect various types of data. In a time when growers aim to shift from strict regulation of inputs to goal-oriented approaches, they will need reliable, tamper-proof data. “Automated systems will play a significant role here, making you a trustworthy partner.”
