Estonian farmers and food producers are testing artificial intelligence in everyday operations, with recent reporting highlighting applications ranging from administrative work to grain drying and automated quality control. Researchers say the decisive question is whether these tools deliver reliable savings under real operating conditions.
In a report published on 18 October, public broadcaster ERR described how Airika Verrev, a production manager and consultant at Pajo Ranch, uses AI to prepare spreadsheets and analyse information. She also uses it to locate relevant legal provisions, while checking the answers it provides.
Marten Madissoo, an agricultural-engineering researcher at the Estonian University of Life Sciences, emphasised that farmers still need expertise to assess automated recommendations. He identified logistics, storage, processing and field analysis as potential applications, particularly where farms already hold several years of records. ERR
The discussion extends beyond general-purpose chatbots. At the EPA agricultural exhibition in Tartu on 14 October, businesses and university representatives examined specific projects. An account published by the university on 16 October focused on the quality of data, employee involvement and the need to define measurable benefits before investing.
IntelliDry presented a machine-learning system intended to improve grain-dryer operation. By estimating grain moisture, it allows an operator to monitor drying remotely while handling other tasks. More precise control could reduce excessive drying and energy consumption, although the company said further work was needed to assess the investment’s profitability. Metsanduse ja inseneeria instituut
Food producer Nõo Lihatööstus described trials of machine vision for checking packaging, labels and the readability of printed dates. Such repetitive inspections can demand sustained attention from employees, making them a practical area for testing automation.
Participants also stressed a limitation that applies across these projects: a model performing well on test data is not yet a dependable production system. Businesses must establish whether it works consistently with their equipment, staff and changing operating conditions. Metsanduse ja inseneeria instituut
More autonomous agricultural machinery remains at an earlier stage. ERR reported that the university’s field robot is still being tested, with safety and operating boundaries among the issues requiring attention. Its existence should therefore not be read as evidence that fully autonomous farming is already routine. ERR
The AIRE artificial-intelligence and robotics centre is inviting companies to submit demonstration-project ideas by 30 November. The approach allows businesses to test a specific application before committing to a larger investment—an important step where promising technology has yet to demonstrate a commercial return.
