Agricultural Living Labs depend on observations and measurements collected under real farming conditions. Some measurements, however, can be time-consuming to collect and analyse manually.
CAgriLab is exploring how image analysis and machine learning can complement these methods.
By June 2026, HAMK had developed prototype algorithms addressing three agricultural measurement tasks: estimating grass and clover ratios, estimating above-ground forage biomass in multi-species systems, and detecting cow activity.
The work builds on earlier development involving image-based approaches to soil structure, biodiversity and vegetation assessment.
Turning field observations into usable data
These tools demonstrate how AI could help Living Labs generate useful measurements from imagery and other field data.
The resulting information could ultimately form part of the wider digital record associated with a Living Lab, allowing measurements, experiments and outcomes to be brought together within CAgriLab.
The work also provides practical examples of how computational tools can be combined with agricultural datasets within the wider CAgriLab environment.