AI Forecasts Greenhouse Yields in Russia
Neural models are helping growers predict harvests and optimize crop management in real time.
In Russia’s Ulyanovsk region, AO Teplichnoye is using neural network models to forecast greenhouse yields. The system matches plant data with actual growing conditions to generate accurate predictions of future output.
The solution was presented at the first AI forum of the association Teplicy Rossii (Greenhouses of Russia), held with the participation of Sberbank. The event brought together agricultural executives and regional governor Alexey Russkikh.
Cameras Handle Quality Control
Beyond yield forecasting, AI is being applied across other operations. Greenhouse facilities are using computer vision during packaging: cameras identify produce and automatically sort it into categories. Since the beginning of 2026, the operation has harvested around 1,700 tons of vegetables, all of which have been sorted automatically.
A Digital Twin for Agronomists
The company is now developing a digital greenhouse model. The system will track plant conditions over time and provide agronomists with recommendations on how to adjust care. These insights will be based on real-time data, including irrigation schedules, watering volumes, temperature, and other parameters.
Governor Alexey Russkikh said the approach will be extended to other regions and proposed establishing data-sharing between farms.
AO Teplichnoye is also considering creating a unified digital environment for greenhouse operators.
