| Patent holder | Nfa Forestry Automation Ab |
| Patent link | View patent |
| Category | Harvesting & Extraction › Remote sensing & mapping |
Improved forest harvester
A forest harvester configured to autonomously or semi-autonomously determine the optimal sequence of trees to be felled and processed represents a significant step toward fully automated timber harvesting, combining onboard sensing, spatial computing, and AI-driven planning within the machine itself. Conventional harvesting sequence decisions are made in real time by the operator, who assesses terrain, tree size, proximity, strip road layout, and forwarding logistics to determine the most productive and safe order of operations. Automating this process requires the harvester to perceive its immediate environment — typically through a combination of LiDAR, stereo cameras, and GPS — to build a local map of standing trees, classify them by species and diameter, assess felling direction safety, and compute a spatially optimised cutting sequence that minimises machine travel, reduces soil disturbance, and maximises timber recovery. Path planning algorithms, including variants of the Travelling Salesman Problem and terrain-aware motion planning, are employed to determine crane reach sequences and strip road positioning. Research in this space has been active in Scandinavia, where companies such as John Deere Forestry, Komatsu Forest, and Ponsse have invested in harvester automation, and in Australia where steep-terrain and multi-species native forest operations present additional complexity. The technology intersects with broader autonomous vehicle development and is often framed as a component of a larger autonomous forestry system in which harvesting sequence optimisation is integrated with forwarding route planning, road scheduling, and mill intake coordination. Realising this capability in full offers benefits including reduced operator fatigue, more consistent adherence to silvicultural prescriptions, and improved productivity on complex terrain where human decision-making is most cognitively demanding.
Supports faster, more efficient felling with moderate risk reduction through reduced machine wear and fewer stand-damage errors.
| Field | Value |
|---|---|
| Company | Nfa Forestry Automation Ab |
| Type | Patent |
| Year | 21/1/25 |
| Status | Pending |
| FWPA RD&E | 4.4 |
| Inventors | Ahrnbom Martin, Svensson Lars, Gillsjo David |
| Patent No. | WO 2025/155235 A1 |