Researchers at Oregon State University are observing human fruit pickers in an attempt to copy their movements with robot fingers. The technology could relieve people of some of the arduous labor of fruit harvesting.  “It is likely that robots will require workers for redundant tasks such as harvesting and pruning,” George Kantor, a professor at the Robotics Institute at Carnegie Mellon University, who was not involved in the research, told Lifewire in an email interview. “However, there will always be a need for human managers to make decisions about the best way to utilize resources and balance risk.”

Picking Challenges

About 70% of fresh produce growers and manufacturers had difficulty recruiting the seasonal staff they needed in 2021. But harvesting fruit efficiently is a skill humans have honed for millennia but has proven challenging to teach to robots.  “Speed, reliability, and cost are the main drivers,” Kantor said. “There is also a need to grasp the fruit without damaging it, though that is not nearly as challenging as the first three. A human harvests 1-2 apples per second. Farmers have a very low tolerance for equipment failures in the field. It is relatively easy to do a small-scale proof-of-concept harvesting robot, and many university researchers and startups have gotten that far. Scaling to reliable, cost-effective production is a big challenge.” But manufacturers are racing to build robots that can beat human pickers. Tevel Aerobotics Technologies is working on a flying, autonomous robot that uses artificial intelligence (AI) to pick fruit from the air. “Farmers are struggling today with recruiting fruit pickers, a situation that puts the whole industry at risk,” Tevel founder and CEO Yaniv Maor told Ag Funder News. “The situation in orchards is worse than in greenhouses for a few reasons. Fruit season in the orchard is shorter than for greenhouses, and orchards are mostly located in remote villages [where] the local labor is not available, and the imported labor is not sufficient.” 

Many Ways to Pick Fruit

Robot-picked fruit is already on the shelves of some stores in Britain. Automatons Two robots developed by Fieldwork Robotics harvest berries in Portugal. In the United States, Georgia Tech Research Institute (GTRI) has developed a robot designed to handle the task of thinning peach trees.  “Most folks are familiar with the harvesting of fruit and picking it up at the market,” Ai-Ping Hu, a GTRI senior research engineer who is leading the robot design project, said in a news release. “But there’s actually a lot more that gets done before that point in the cultivation cycle.” The Georgia robot uses a LIDAR sensing system and GPS to find its way through peach orchards and steer clear of obstacles. The LIDAR system determines distances by targeting an object with a laser and measuring the amount of time it takes for the laser beam to reflect back, while the GPS technology measures locations as specific as a fraction of an inch.  Once it’s found a suitable peach tree, the robot uses a  3D camera to determine which peaches need to be removed and grabs the peaches using a claw-like device. “There’s no robot in the world right now that can harvest or thin peaches as well as people can,” Hu said. “The technology’s not quite there yet.” Replicating the dexterity of a human hand is still a major challenge for roboticists. Hillel Chiel, a biology professor at Case Western Reserve University, told Lifewire via email that your hand, unlike current robot systems, can quickly integrate tactile feedback to actuate many different degrees of freedom. The human hand also has the ability “to use a variety of stimuli (movement past a surface, pressure, responses to force, all of which can be integrated to determine the fragility, shape or weight of an object) to rapidly and dynamically adjust grasp are critical features of what the human hand can do,” he said.  Soft grippers can conform to the shape of many different objects, and embedded soft sensors can allow for delicate grasping, Roger Quinn, a  professor of engineering at Case Western who studies robotics, told Lifewire in an email interview.  “Despite decades of some really remarkable work in developing human-like hands and similar graspers, actuation, tactile sensation, and control remain primary research problems for control of fine movements and force such that both rigid and very soft and delicate objects can be manipulated,” he added.