Reinventing tree inventory

Arboair's calculated ROI per million m³sk per year is 1.6–2 MEUR in the base case, 2.4–3 MEUR in the upside case and 3.6–4 MEUR in the stretch scenario. The benefit falls into five categories: more efficient harvesting (0.4–0.6 MEUR), more efficient transport (0.2–0.3 MEUR), improved assortment yield from pulpwood to sawlogs (0.3–0.6 MEUR), time savings in planning and production (0.2–0.4 MEUR) and reduced industrial downtime (0.4–1 MEUR). Sources are data from Skogforsk, the Swedish Forest Agency and customer data.

Arboair's stand data allows timber buyers to screen large areas and identify harvest-ready stands without physically visiting each site. The inspector can prioritise field visits based on digital stand data – species, volume, damage – and approach the landowner with a fact-based offer before a competitor has even started screening.

Yes. Arboair's AI analysis automatically identifies dead and damaged trees, windthrow and stem breakage with a precision, speed and scale that is not possible with manual inventory. The analysis is performed at individual tree level, which means the share of damaged trees per stand can be quantified and ranked for prioritised action.

Yes. Arboair's AI models calculate volume per individual tree from image data captured by drone. For each tree we identify tree ID, species, height, crown diameter, stem diameter (DBH) and volume in m³fub, and in some cases CO₂ and biomass. These data points are automatically aggregated to stand level. Validation against real harvester data from harvest sites shows an average volume deviation of -3.34 per cent, and the average price deviation was -0.26 per cent. Validation against full calipering shows a deviation in diameter at breast height of -0.03 cm across roughly 3,000 measured trees. This applies to standing forest. The analysis works equally well for standing forest as for damaged stands. In the case of forest damage – for example storm, stem breakage or spruce bark beetle attack – Arboair's AI automatically identifies and estimates the volume of dead and damaged trees at individual level, providing a direct basis for prioritising action and calculating timber value.

Analysis of a damaged area can be carried out in days rather than weeks. A manager who traditionally spends 75 working days manually surveying 7,500 hectares per year can complete the same review in roughly 20 days with Arboair's digital analysis. That time includes walking out to validate important areas. For storm damage, where fast action determines timber value, this delivers a direct economic advantage.

Since founding in 2019, Arboair has analysed more than 600 million trees across 13 countries. Customers include Södra, Mellanskog, Holmen, Norra Skog and Boxholms Skogar.

Yes. Arboair's platform and AI models are adapted to the full breadth of forest actors – from private forest owners with smaller properties to professional managers and large forest companies with hundreds of thousands of hectares. For private forest owners, Arboair offers two ways to get started: either instructions for collecting image data yourself with a consumer drone, or a complete assignment where Arboair handles the entire data collection. The result is the same – tree-specific data on species, height, volume and damage delivered as a finished decision basis. For professional forest actors such as forest companies, timber buyers and managers, the platform is built for large-scale and recurring inventory of large areas and many properties, with the ability to integrate data directly into existing planning systems such as Skogshubben, pcSkog, VSOP and Esri applications.