Revolutionising the berry industry with robotics and AI

First data collection.

Sweden’s wild berry tradition is facing sustainability challenges due to its reliance on labour-intensive methods and foreign workers, contributing to high carbon emissions. In the "BerryAI" project the researchers from the Robotics and Artificial Intelligence group at Luleå University of Technology together with local SMEs Elva Sustainable Systems, Burliden Lantbruk, and Norrskensbär are developing a cutting-edge solution, using machine learning and sensor technologies to revolutionize berry picking, reducing carbon footprints, and ensuring long-term sustainability in the industry.

About

Our research is focused on developing a cutting-edge AI-based solution that will have potential to enable autonomous wild berry picking, addressing key challenges like labor shortages, environmental impact, and operational efficiency in agriculture.

Benefits for SMEs

We will develop a comprehensive and well-annotated dataset featuring the four major wild berry species: bilberries, lingonberries, cloudberries, and crowberries. This dataset is critical for training machine learning models and enabling small and medium-sized enterprises (SMEs) to automate berry picking. By overcoming challenges related to data bias and the lack of existing comprehensive datasets, we aim to ensure that the dataset accurately represents various environmental conditions.

We will create an AI perception framework powered by advanced convolutional neural networks (CNNs). This framework will detect and classify berries with high precision, significantly improving the efficiency and accuracy of berry picking operations. In doing so, it will reduce labor costs, dependence on seasonal workers, and the carbon footprint associated with transporting foreign labor.

The project's goal

We will develop a scalable hardware prototype for data collection and field validation, allowing SMEs to adopt these technologies in real-world agricultural applications. This prototype will serve as proof of concept, laying the foundation for future advancements in automated and sustainable berry picking solutions that can be implemented across Scandinavia and beyond.