show index hide index
The advent of artificial intelligence is revolutionizing a multitude of fields, including whisky tasting. By challenging human olfactory and taste abilities, AI, thanks to sophisticated algorithms, has proven exceptionally accurate in distinguishing not only the origin but also the dominant aromas of different whiskies. Systems developed by the Fraunhofer Institute in Dresden demonstrate unprecedented efficiency in analyzing chemical compounds and combining these findings with precise olfactory descriptions. With an impressive success rate, they outperform human experts in the field.
In a world where technology continues to advance, artificial intelligence (AI) has now taken the reins of whisky identification. By outperforming human experts thanks to advanced algorithms, it is able to detect the origins of a whisky as well as its most subtle notes. German researchers have developed a study that demonstrates this feat, with an impressive success rate. This article takes you deep into these findings and their implications for the whisky industry.
AI and whisky origin recognition
Researchers at the Fraunhofer Institute in Dresden have developed algorithms that can recognize with 94% accuracy whether a whisky is of American or Scottishorigin. This impressive result surpasses the ability of human experts subjected to the same test. By training the AI on a dataset comprising 390 distinct molecules commonly found in whiskies, this work has made it possible to perfect this origin identification procedure.
These algorithms use several analytical techniques, such as gas chromatography and mass spectrometry, in order to dissect the chemical compounds in the samples and reproduce them into a unique aromatic profile. This set of techniques makes it possible to determine the provenance of the whiskey with a scientific approach that now rivals the human sense of smell.
AI excels at detecting dominant flavors
Artificial intelligence does not stop at identifying the origin. It has also shown itself capable of detecting dominant flavors in each whiskey sample. By analyzing whiskeys like Talisker, Glenmorangie, and other equally famous varieties, the algorithms managed to identify with great precision the five most present notes. For example, AI detected that American whiskeys often contain characteristic caramel notes, while Scottish whiskeys reveal fruity nuances such as apple and notes of phenol, which give them a slightly peaty taste.
The key to success lies in the education of the algorithms, where they were subjected to various samples. Using descriptive grids associating terms like “floral”, “woody” or “fragrant”, the AI learned to match chemical data to human perceptions, thus achieving expertise comparable to that of the best tasters.
Practical applications in the whiskey industry
The implications of these advances are multiple for the whiskey industry. On the one hand, AI could be used to detect counterfeit whiskeys, thus guaranteeing the authenticity of the products on the market. On the other hand, it would facilitate the quality control, making it possible to verify that the final aroma of the product corresponds to the producers’ expectations.
Furthermore, when developing new whiskies, AI could play a crucial role. Creating a whisky requires time and expertise, and algorithms could save significant time and money. In short, this technology offers solutions to optimize both accuracy and efficiency in the production and marketing of whiskies.
The future of AI in tastings
The success of artificial intelligence in the whisky field foreshadows other applications in the alcoholic beverage industry. In the past, it has already demonstrated its ability to identify Bordeaux grands crus and successfully recognize different Belgian beers. The future could well see the introduction of AI in various other products, transforming the way we consume and enjoy our favorite beverages.
