Why hasn't AI reached day-to-day operations in forestry?

Reluctance is the usual answer, and two others fit the evidence better. The first is timing: a daily decision can only use records that are digital by the time it is made, and in some operations the day's loads, scale weights and site notes arrive later. The second is fragmentation: where the records are digital, they are often split across a landowner, a logger, a hauler and a mill, each with its own system. Which one binds differs between operations, and neither has been measured across the industry. A business can measure both for its own records in a week, and the answer says where AI can start.

What the usual answer says

The lists that rank for this question name barriers: cultural resistance, a shortage of skills, cost, fragmented data, poor connectivity. Each is real. In a 2025 study of 20 forestry professionals, published in Forest Policy and Economics, the lead researcher told the university's news office that a key concern was "the risk of training AI using some agencies' poor-quality or biased data and then trusting its flawed outputs for important land management or policy analysis". Lists like this say what worries people. They don't say what a business can check.

Timing

Operational decisions depend on what is happening now. A 2025 paper in the Journal of Forestry by Kilgore, Blinn and Snyder, on agency foresters administering timber sales as winter conditions change, describes "spending more time interacting with loggers and more frequent site visits to active logging operations". Those decisions move with the conditions on site, and the records behind them have to keep up.

Connectivity is one reason they may not. The OECD's December 2025 report on small-firm AI adoption found that at the end of 2024, fixed download speeds in metropolitan areas "were, on average, 44% higher than in regions far from urban centres". The gap in mobile speeds widened from 5 to 45 Mbps between 2019 and 2024. Those are regional averages of speed. They don't measure coverage at a landing, but they point the same way: "These gaps in accessibility in rural areas can hold back both digital innovation and adoption".

Fragmentation

Much operational data is already digital: harvester production files, truck GPS, electronic scale tickets. If AI still hasn't reached daily decisions where those exist, timing is not the constraint. A 2024 review of wood supply chains by Palander and colleagues appeared in Current Forestry Reports, and we read its abstract. It notes that "Due to ongoing outsourcing, the wood procurement chains and the wood supply chains were identified", and that "several different data collection technologies can be implemented" by each organisation. It concludes that "modeling time-related and sequential measures" is what makes forest logistics work. Many parties, each with its own tools, means the records a daily decision needs may be digital and still not in one place.

Measuring both

Take the records an operation depends on daily, such as the load, the scale weight, the site visit note and the machine hours. For one ordinary week, write down two things about each. The first is the lag: the time from the event to the moment the record is digital and usable. The second is the owner: whose system holds it, and whether it shares a key with the others.

Records with short lags and a shared key are where operational AI can start now. Records with long lags need capturing faster first. They can be read on arrival, entered offline on a phone, or taken from the machine's own data. Records that are timely but sit in another party's system need an agreement and a key before any AI can join them.

What the week's table usually shows

We would expect the table to split. A few records will be timely and shared, and those are the place to start. One or two will account for most of the lag, and they are often the ones written by hand or keyed weekly. Others will be timely but held by a contractor or a mill. The fix differs for each, which is why the table is worth a week before any AI project.

When it doesn't apply

Planning, finance and reporting use records days or weeks after the event, so lag matters less there. Fragmentation still does, because a margin or a cost per tonne needs records from more than one party. And some field decisions depend mostly on what the forester sees on site, where AI adds little however fast the records arrive.

Quarri for forest management is built around how a forest operation runs, from the cruise to the settled account.

Quarri is an AI-native data platform for the timber supply chain. It connects buying, production, sales and inventory for forest management, sawmill, wood products and pulp, paper and packaging operators.