The agricultural orchard sector in the EU faces labor shortages and lacks means to attract young practitioners, while most operations that need to be performed are manual and repetitive, namely pruning, thinning and harvesting. Up to this date, there exists no high-scale automatic solution able to tackle all these operations, and those from the state-of-art solutions do not focus on how to make these solutions operable for farmers.
EURECAT and KU LEUVEN are working to use robotic manipulators solutions to handle all these operations, leveraging XR technologies to lower the entry barrier for current farmers and draw attention on the sector from younger people. Therefore, each task will have an specific gripper designed to mimic human dexterity while being suitable for everal types of orchard. Robot autonomy will be governed by a flexible architecture to generate adaptive and reactive behaviours, considering visual input from cameras to capture the information from the scene, processed to provide key features for the task, e.g. fruit location, ripeness and size for harvesting and thinning, or branch orientation and size for pruning. Using XR, farmers can get live data and visuals from the process, such that they can also supervise the robot operation and even provide feedback to adjust their performance to particular requirements.
Additionally, AI-based techniques will be used to capture human execution for autonomous operation, leveraging on the advantage from XR to provide realistic and egocentric simulations of environments and immersive experiences from anywhere. Hence, farmers could provide demonstrations of different tasks on virtual scenarios that mimic real ones, such that decision-making models can be generated and used by robots during operation to improve their performance. Moreover, these models could be even used to capture expert knowledge and teach younger practitioners perform the tasks within the same virtual environments.
In practice, with this solution farmers can deploy the robotic system to support or autonomously perform manual tasks, while supervising and adapting operations through intuitive XR interfaces, that can be also used to preserve agronomic know-how and teach young practitioners. The main costs relate to initial investment, integration, and training, which are reduced by reusable platforms and user-friendly XR tools, and key benefits include lower labor dependency, higher productivity and quality, safer working conditions, and increased attractiveness of the orchard farming sector.