Le 29/04/2025

Fruit Harvesting Robots: Adapting Orchards and Expectations

Growers and manufacturers discuss redesigning tree fruit groves for automation and the benefits of specialised versus vs. multi-function robots

As technology advances, agricultural robots are poised to play a transformative role in the tree fruit industry. At World FIRA 2025, growers and agricultural robot developers discussed what’s working, what could be improved and whether orchards should be changed to suit robots — or the other way round. They also debated the pros and cons of having a robot that does one job well versus one that can perform multiple functions.

René Koerhuis from Future Farming moderated the discussion. Panelists included:

The 47-minute roundtable is available to view in English and Français on the FIRA Forum YouTube channel. The key takeaways included:

  • Designing orchards and groves to be robot-friendly
  • Labour cost savings associated with fruit-picking robots
  • The importance of rigorous testing
  • A need for robust after-sales support from manufacturers.

Below are a few highlights from their discussion.

Adapting Groves and Orchards for Mechanical Harvesting

Efraim Albecht Neto from Fundecitrus (Brazil) described changing the characteristics of his farm to accommodate fruit-picking robots. “In Brazil we have some very tall trees and people need stairs to harvest the fruit. I have worked with the Agromilora to transition from 3D orchards with big trees to 2D orchards that are more like a wall.”

Marine Louragant from Centre Technique Interprofessionnel des Fruits et Légumes said the espalier-style “2D” trees are something her organisation has also started implementing. But noted change will come slowly. “We are not going to be able to change the orchards overnight because growers don’t want to pull out crops they’ve had for 10, 15 years,” Louragant said.

Iganasi Iglesias and Saio Landonio, both on the technology side, advocated for adapting citrus groves and orchards to be more suitable for mechanical harvesting. Iglesias cited planar canopy architecture as the superior choice.

“In a planar system, accessibility is increased by 80% and it allows for high precision. If we need labour, for pruning or whatever, it is 40% more efficient in a planar canopy,” Iglesias said. “Also, the light distribution is better and food quality increases. I am convinced that the future is adapting the architecture of the trees to the mechanisation of robotics.”

Landonio agreed. “Just as we adapted cultivation for tractor mechanisation, we should adapt the canopy for robotisation. Also, for precision farming, mapping the trees is easier if the tree is 2D and not 3D.” He suggested using robots to pick 70-80% of the harvest and manual labour to complete the remainder — or forgo manual picking if the cost doesn’t make economic sense.

Multifunction vs. Specialised Agricultural Robots

The panel also explored whether it’s better to have agricultural robots designed for a specific task or one that can perform multiple operations. Neto favoured a single machine capable of various jobs.

Louragant countered that an all-in-one solution could be more expensive and less efficient than specialised robots performing tasks in parallel. “For the grower, it may seem ideal to have one machine that does everything...but if there are multiple tasks that must be carried out at the same time, this can create a problem. Also, the cost to develop a machine that can do everything could mean that robot will cost a lot of money.”

Iglesias was in favour of a multi-tasking robot but emphasised fruit harvesting robots should be the priority. “With apples, pears and peaches, labour is 50% of the cost of production...we have the experience, robotic harvest is possible,” Iglesias said.

“Another important task is winter pruning [but] you can imagine how difficult that will be with a robot,” Iglesias added. “For me, the focus of the next decade will be harvest first and then winter pruning.”

How Fruit Harvesting Robot Manufacturers Can Improve

Koerhuis shifted the discussion to the panelists’ experience with agricultural robots — expectations versus reality. Both Louragant and Neto described issues around performance due to lack of real-world testing.

“Some [manufacturers] go too fast,” Louragant said. “They present growers with solutions that are still prototypes that haven’t been tested or finalised.”

Neto echoed Louragant’s comments. “It’s different when you have the robot operating in a real field compared to a situation where you can control everything,” Neto said.

Louragant and Neto also identified a need for more robust training and support, especially in the growers’ native language. “Five or six years ago, we tested an autonomous tractor, but everything was in English. When you have a grower or a technician that only speaks French, that makes it difficult.”

Louragant noted things have improved in recent years as the agricultural robotics industry matures.


View the Entire Roundtable Discussion

The 47-minute discussion is available to view in English or Français. Visit the FIRA Forum YouTube channel to watch this event and others from World FIRA 2025.


More Roundtable Recaps from World FIRA 2025

Catégories : #Alimentation
Auteur
  • Megan Denny
    GOFAR : Freelance Copywriter