Deep Forestry has raised €3 million, about SEK 33 million, to scale autonomous drones that fly below forest canopies and build 3D inventories at individual-tree level. The Uppsala company, spun out of Uppsala University in 2018, is backed in the round by Fairpoint Capital, with participation from Superorganism, Spatial Capital and First Gate Invest.

The details come from Dagens Industri, which reported that Deep Forestry’s drones collect high-precision 3D forest data and can inventory volume, diameter, biomass and carbon more accurately than traditional methods. The company calls it a step toward a fully comprehensive precision inventory at the level of individual trees.

Forestry rarely gets the startup glamour treatment, which is strange when you consider the numbers. It is land, carbon, biodiversity, industrial supply and political identity all standing upright in rows. The forest is both asset and argument.

Remote sensing has improved quickly, but forests are vertically complicated. Satellites and aircraft can do a lot from above, yet the canopy hides the structure underneath. Deep Forestry’s bet is that flying below the canopy gives owners and managers a different class of data, closer to what field crews measure but faster and more repeatable.

The unexpected observation is that the drone is not the main character. The inventory is. Better measurement changes how owners plan harvests, estimate carbon, protect biodiversity and decide which stands need attention. A drone without trusted analytics is a flying camera. A drone with defensible tree-level data becomes part of the balance sheet.

Fairpoint partner Hadar Cars described the company as operating at the intersection of robotics, AI and sustainable resource management in a market that is only beginning to digitize at scale. That phrase can sound like investor wallpaper, but here it fits. Forestry is enormous and still operationally analog in too many places.

The carbon angle is not just branding. As companies and countries lean harder on nature-based carbon claims, the quality of forest measurement becomes a credibility problem. Tree diameter, volume, biomass and growth estimates feed into financial and environmental decisions. Bad data can create false confidence, and false confidence has a long shelf life.

Deep Forestry’s official materials emphasize autonomous forest drones and instant AI analysis. If the company can make data collection routine, it could help forest owners move from periodic sampling to more continuous intelligence. Not every tree needs a personality. It does need a number.

There are operational challenges. Under-canopy flight is hard. Branches, light, GPS limitations, weather and terrain all conspire against autonomy. The product must work in the forest that exists, not the forest in a robotics video. That is why the round is interesting: it funds the grind between prototype and repeatable field service.

The more companies rely on forests for carbon claims, sourcing decisions and biodiversity commitments, the more forest data starts to look like financial infrastructure. A measurement error is not only a technical issue. It can distort valuations, harvest planning, insurance assumptions and climate reporting. That is why tree-level inventory is more than a nicer map.

Deep Forestry’s drones could make measurement more frequent and granular. Instead of relying on sparse field plots or high-level remote sensing alone, owners may be able to build a richer picture of what is actually growing, dying or changing beneath the canopy. The phrase beneath the canopy sounds poetic. It is also where much of the uncertainty lives.

This matters in the Nordics because forests are not a niche resource. They are industrial feedstock, climate policy instrument, recreational space and political battleground. Better data will not resolve every conflict, but it can make arguments less foggy.

Under-canopy drones face a navigation environment that seems designed to embarrass robotics demos. GPS can be unreliable. Branches are irregular. Light changes quickly. Wind behaves differently below the canopy. Moisture, cold and dust all take their turns. A system that works on a calm test day may struggle in commercial conditions.

That is why Deep Forestry’s university spinout background is important but not sufficient. The company must convert research into repeatable operations. Customers will not pay for elegant autonomy if the field team still needs to babysit every flight. The product has to become boring enough to schedule.

The good news is that forestry customers understand physical difficulty. They are used to machines, weather and long planning cycles. A drone system does not have to feel like consumer electronics. It has to be reliable, useful and worth the operational change.

Superorganism’s participation is a clue about where forest technology is heading. Biodiversity is becoming an investment theme because companies and governments are starting to measure nature risk with more seriousness. Forest inventory tools can support that shift if they capture more than timber volume.

Individual-tree data could eventually help identify habitat structures, species composition, growth patterns and disturbance. Some of that may require additional sensors and models. Some will require patient validation. But the direction is clear: forests are being measured for more reasons than harvest planning.

The caution is that better measurement can support both conservation and extraction. Technology does not choose the policy goal. Owners, regulators and markets do. Deep Forestry’s commercial opportunity will depend partly on helping customers prove better stewardship, not only better yields.

The raw output of a drone flight can be impressive, but forest owners do not buy awe. They buy better decisions. Which stand should be harvested? Where is disease spreading? How much biomass is present? Which areas should be protected? How confident is the carbon estimate? Deep Forestry’s product has to answer those questions, not merely produce beautiful 3D scans.

That means the analytics layer may become as important as the drone platform. Customers will want integration with planning systems, reports that auditors can understand and data that field teams trust. The company will need to translate robotics into forestry language. Every vertical AI and robotics startup eventually learns this: domain translation is product work.

The advantage is that once a customer trusts the data, switching may be painful. Historical forest data becomes more valuable over time because change is the point. A single scan is useful. A series of scans becomes intelligence.

Forestry customers can be technologically sophisticated, but they are also pragmatic. They will ask whether a drone inventory changes a decision enough to justify the cost. They will compare it with field crews, aerial data and existing planning tools. Deep Forestry has to show not only better data, but a better total workflow.

That may mean selling first to customers with the most painful measurement problems: carbon projects, biodiversity-sensitive land, complex stands, or owners managing large portfolios where manual sampling is too slow. A universal forest data platform is a tempting story, but wedge markets usually win first.

The company’s reported 2024 revenue suggests it is already beyond the pure concept stage. That matters. Robotics startups can spend years perfecting machines without proving customers will pay. Deep Forestry now has capital to deepen the product, but its commercial proof will come from repeat buyers who treat the data as necessary, not experimental.

Deep Forestry sits in a different corner of climate tech than carbon dashboards or energy marketplaces. It is grounded, literally, in moss, trunks, machinery and landowner decisions. That makes it less scalable in the pure software sense, but potentially more defensible if the data quality is hard to copy.

The Nordic region has a special relationship with forests. It also has industrial forestry, biodiversity pressure and growing demand for credible carbon measurement. A company that can make the hidden structure of forests visible gets pulled into all three conversations.

The next stage will likely be about repeatability across forest types and customer needs. Sweden alone contains enough variation to challenge a neat robotics story. Commercial forestry, conservation areas and carbon projects may each demand different outputs, even if the drone hardware is similar.

If Deep Forestry can make those outputs credible, it becomes part of a larger shift from estimating nature to instrumenting it. That phrase sounds uncomfortable because nature should not be reduced to a dataset. Yet better measurement may be one of the few ways to make economic systems take forests seriously before they are gone.

That would be a meaningful Nordic export: not another dashboard pretending to solve climate, but a machine that goes into the woods and returns with measurements people can act on. Sometimes progress sounds like propellers under pine branches.

The €3 million round will not transform forestry by itself. It gives Deep Forestry room to prove that under-canopy autonomy can become a commercial tool rather than a research curiosity. That’s a worthy test. The future of forest data may be flying at shoulder height, not orbiting above us. Original report