Living Labs can produce large amounts of information from field experiments, soil measurements, farm-management activities and other research. Finding the relevant information can be difficult when users need to understand where particular datasets are stored or how they have been structured.
Natural-language search offers a different approach: allowing users to search using ordinary language.
By July 2026, CAgriLab had developed a natural-language search prototype, alongside prototype functionality for uploading and visualising Digital Twins. The project's wallet-free publishing and data-harmonisation tools had also reached production status.
Making Living Lab evidence easier to find
The longer-term opportunity is particularly relevant to regenerative agriculture.
The effect of a regenerative practice can vary according to soil, climate, crop, management history and other local conditions. Finding evidence therefore involves more than searching for the name of a practice: researchers need to understand the context in which an experiment took place.
CAgriLab is exploring how natural-language interfaces can make this growing body of Living Lab information easier to navigate.
Instead of manually searching individual datasets, a future user might ask a question such as: "What experiments have tested cover crops under conditions similar to mine?"
The prototype represents an important step towards making CAgriLab's underlying data and knowledge accessible without requiring users to understand the technical infrastructure behind it.