![\"A](\"https:\/\/images.newscientist.com\/wp-content\/uploads\/2024\/12\/19105415\/SEI_233737707.jpg\")
<\/div>Colour, smell, taste and chemical constituents can all be used to distinguish whiskies<\/p>\n
Jane Barlow\/PA Images\/Alamy<\/p>\n<\/div>\n<\/figcaption><\/figure>\n<\/p>\n
Artificial intelligence can tell Scotch whisky from American whiskey and identify its strongest constituent aromas more reliably than human experts \u2013 by using data rather than tasting the drinks.<\/p>\n
Andreas Grasskamp<\/a> at the Fraunhofer Institute for Process Engineering and Packaging IVV in Germany and his colleagues trained an AI molecular odour prediction algorithm called OWSum on descriptions of different whiskies.<\/p>\n Then, in a study involving 16 samples \u2013 nine types of Scotch whisky and seven types of American bourbon or whiskey \u2013 they tasked OWSum with telling drinks from the two nations apart based on keyword descriptions of their flavours, such as flowery, fruity, woody or smoky. Using these alone, the AI could tell which country a drink came from with almost 94 per cent accuracy.<\/p>\n Because the complex aroma of these spirits is determined by the absence or presence of many chemical compounds, the researchers also fed the AI a reference dataset of 390 molecules commonly found in whiskies. When they gave the AI data from gas chromatography\u2013mass spectrometry showing which molecules were present in the sample spirits, it boosted OWSum\u2019s ability to differentiate American from Scotch drams to 100 per cent.<\/p>\n Compounds such as menthol and citronellol were a dead giveaway for American whiskey, while the presence of methyl decanoate and heptanoic acid pointed to Scotch.<\/p>\n <\/p>\n The researchers also tested both OWSum and a neural network on their ability to predict the top five odour keywords based on the chemical contents of a whisky. On a score from 1 for perfect accuracy to 0 for consistent inaccuracy, OWSum achieved 0.72. The neural network achieved 0.78 and human whisky expert test participants achieved only 0.57.<\/p>\n \u201c[The results] underline the fact that it\u2019s a complicated task for humans, but it\u2019s also a complicated task for machines \u2013 but machines are more consistent than humans,\u201d says team member Satnam Singh<\/a>, also at the Fraunhofer Institute. \u201cBut that\u2019s not to say that humans are not needed: we do need them to train our machines, at least, right now.\u201d\u00a0<\/strong><\/p>\n Neither model takes into account the concentration of molecules, only their absence or presence, which is something the researchers hope to rectify, and which may yield even greater accuracy.<\/p>\n Grasskamp says such AI tools could be used for quality control in distilleries, or to help develop new whiskies, as well as detecting fraudulent ones. But they could also be used for \u201canything that smells\u201d, such as other food and drink production or in the chemical industry.<\/p>\n Topics:<\/p>\n<\/section><\/div>\n