Forests cover more than half of Sweden’s land, making them central to the country’s biodiversity, economy, and climate goals. But monitoring and managing them remains a challenge. A recent project explored how digital twins and enhanced AI models could pave the way for smarter, faster, and more accurate forest management.
Conventional methods for forest management face many challenges: manual inventories, large volumes of images, outdated data, and extensive annotation make the process slow, costly, and labour-intensive. A Swedish project demonstrated how emerging digital tools can offer a promising alternative, providing more precise, current, and comprehensive data for decision-making while saving both time and costs.
The two-year initiative, which ran from autumn 2022 to December 2024, was led by the Swedish software company Arboair, in collaboration with Holmen and AI Sweden, and supported by Vinnova.
A digital twin of the forest
Central to this work was the creation of a digital twin of the forest, a virtual environment that allows forest managers to make data-driven decisions from their desks. This makes it easier to plan operations more accurately, such as thinning and felling, monitoring forest health, detecting damage early, and even simulating the effects of different management strategies to optimize both economic and ecological outcomes.
One of the project’s most significant outcomes was demonstrating the effectiveness of using synthetic data to train AI models, which can be useful in addressing complex real-world applications. By generating 3D forest scenes, the models learned to identify rare species, detect disease, and spot damage.
–This insight is a game-changer because it means we can develop highly accurate analysis for rare but critical occurrences, like specific types of storms or insect damage, without needing vast amounts of real-world examples, which can be difficult and expensive to collect. It significantly accelerates our ability to create robust and reliable solutions, says Arboair’s COO Jacob Hjalmarsson.
Scalable and adaptable
The potential of this technology extends well beyond Sweden. The challenges of forest management are universal, and Arboair’s solution is highly scalable, adaptable to different forest types and ecosystems worldwide.
–We see it as a key Swedish export, and a leap in Swedish AI development and research, that can inspire and enable a global shift towards more efficient and sustainable forestry, says Hjalmarsson.
Looking ahead
The project has opened many doors for Arboair’s solution. Now, the company is planning to continue the dialogue with its partners, including Holmen and AI Sweden, to build on these results and explore the next generation of AI applications in forestry.
Source: Arboair.
Interested in cleantech solutions for the agriculture and forestry sector? Explore our list of companies at swedishcleantech.com/companies
