Image credit: Tevel
About ten percent of fruits fail to be harvested globally each year because of labor shortages, accounting for a portion of the total global loss from failed fruit harvesting of about $30 billion annually. Tevel Aerobotics Technologies created FAR™, fruit-harvesting flying autonomous robots, to assist seasonal laborers and orchard staff in reducing lost revenue.
Tevel’s robots use artificial intelligence and computer vision before, during, and after the fruit-picking process. Orchards can order a specified number of robots for a date and time range using Tevel’s Harvesting Management application. When the time comes, Tevel sends the specified number of robots, the vehicle base, and necessary technicians to the orchard. Four robots are tethered to the vehicle base, which is driven between orchard rows. The orchard inputs the type of fruit into the Harvesting Management application, as well as the desired harvesting specifics, including grade and quantity. Then, the robots get to work.
Automated Picking, Pruning, and Trimming
Tevel completed pilot programs at orchards in Italy and the US, and the robots have successfully harvested stem fruits including apples, peaches, and plums. The company plans to test the robots in other orchards to harvest citrus and avocados. Eventually, the robots could also prune, trim, and thin trees and bushes.
Tevel expects that FAR will reduce overall operating costs for orchards. Labor costs can make up as much as 50 percent of an orchard’s overall operating costs. Unlike human laborers, the robots can work day and night; they have low injury risk, and do not require accommodations or benefits. As labor shortages continue and prices increase, Tevel’s FAR can be appealing to orchards needing quick assistance.
Image credit: Tevel
Harvesting the Power of AI
The robots have embedded cameras, sensors, and a suction tube arm. Vision algorithms detect and categorize fruit based on ripeness, size, and other factors. This ensures that only good, ripe fruit is harvested by the robots. When the robots detect ripe fruit, as specified by the orchard, they use their suction tube arm to harvest and deposit fruit into the vehicle.
Vision algorithms also help the robots to detect foliage and other objects, including other robots and human laborers, in their surroundings, so they can avoid getting stuck or damaged during harvesting. Maneuver algorithms help the robots plan and execute their trajectory path, and balancing algorithms stabilize the robots when an external force is applied, possibly by foliage or other objects.
Tevel’s Harvesting Management application and the robots use perception algorithms to help with fruit tracking and data integration. These algorithms work with optimization algorithms that use orchard data to create a cost-effective and efficient harvesting plan. Orchard operators can use the Harvesting Management application, available on mobile devices, to view real-time updates about the harvest progress, as well as harvest data such as cost, time, and percent completion.
The global population is expected to increase by two billion people by 2050, requiring agricultural space for fruit growing and harvesting to double by 2050. At the same time, agricultural labor is projected to be reduced by 50 percent by 2050.
Working with these statistics is a challenge. Scholars agree that agriculture production must increase about 70 percent by 2050 to meet the increased global demand amid such challenges. Tevel Aerobotic Technologies hopes that its flying autonomous robot (FAR) can do just that.
- Learn more about FAR™ by Tevel.