Spanish cellar trials begin with 48 AI-sorted wine barrels

AI-assisted barrel classification

Researchers at the University of Valladolid have sent 48 wine barrels sorted with the help of artificial intelligence into cellar trials as part of a Spanish project. The initiative focuses on the oxygen permeability of the barrels’ wood. By placing the material into categories based on this property, the researchers hope to make barrel ageing more predictable for wineries.

The available information does not explain what data, sensors or models the artificial intelligence system uses to perform the classification. It also does not identify how many wineries are involved, which wines will be placed in the barrels or how long the trials will run. The cellar phase can therefore be understood primarily as a practical test of how the predefined barrel categories behave under real ageing conditions, although the detailed experimental design has not been disclosed.

Oxygen and the ageing process

The project is based on the idea that classifying wood by oxygen permeability may help wineries anticipate the course of barrel ageing. However, the published summary does not specify which measurements will be used to assess the wines or how large a difference researchers expect to find between the various barrel groups.

Testing 48 barrels gives the researchers an opportunity to compare the AI-supported classification with observations made in the cellar. For now, there is no information on whether the trial includes control barrels, repeated tests or direct comparisons using the same wine across different barrel categories. No conclusions can yet be drawn about whether oxygen-permeability classes correspond to measurable or consistent differences during ageing.

Potential value for wineries

If the initial wood classification proves to have a reliable relationship with ageing behaviour, it could help wineries make barrel selection and ageing plans more predictable. Such information could potentially support decisions about matching barrels with intended cellar strategies. The source material, however, provides no trial results and does not address the method’s accuracy, cost, commercial availability or suitability for broader adoption.

The movement of the 48 barrels into cellar testing marks a practical stage for the research, but its value will depend on evidence that has not yet been reported. Details about the timetable for results are also unavailable.

The original report was published by Bing News – Wine AI.