hero image

Artificial intelligence has unlocked the mystery of wine origins, as a University of Geneva team, in collaboration with the University of Bordeaux, employed AI to identify unique chemical markers in red wines from seven renowned Bordeaux estates with absolute precision, shedding light on each wine’s distinctive origin.

AI could help solve counterfeit issues in wine industry

This study shows potential in developing tools to address counterfeit issues and provide predictive insights for decision-making in the wine industry.

Wine composition, influenced by factors like grape composition, variety, viticultural practices, and soil nature, is a complex blend of thousands of molecules. The taste of a wine can be greatly affected by even minor variations in these elements. The intricacies involved in identifying a wine’s origin based solely on sensory characteristics pose a challenge due to these complexities.

UNIGE professor, Alexandre Pouget, said that Wine experts have been trying to solve this puzzle for a long time, using methods that are often unreliable and complicated. Pouget added that the complexity of wine blends and the limitations of the techniques used make it are like searching for a needle in a haystack.

Michael Schartner, the primary author of the study, said that By utilizing a method called dimensionality reduction, researchers were able to analyze each wine’s comprehensive chromatograms, which can consist of thousands of data points, and condense them into two X and Y coordinates. This approach eliminates extraneous variables and incorporates the entirety of each chromatogram, including background noise.

Each estate has a unique chemical signature

Researchers utilized dimensionality reduction to streamline data by eliminating unnecessary variables. The process revealed distinct clusters or “clouds” of points representing wines from the same estate, organized based on chemical similarities. These clusters effectively showcased each estate’s unique chemical fingerprint, providing valuable insights into wine characterization.

St├ęphanie Marchand, co-author of the study, through this analysis, it was discovered that each wine estate possesses a unique chemical signature. Additionally, the study found that three wines were clustered on the right side and four on the left side, which aligns with the geographical location of the estates along the Garonne River banks.