Artificial intelligence is making waves in the world of whisky, as researchers have successfully used the technology to analyse and identify key aroma profiles and origins of various whiskies. Conducted by Germany’s Fraunhofer Institute for Process Engineering and Packaging, the study explored the molecular composition of 16 US and Scottish whiskies, including renowned brands like Jack Daniel’s, Maker’s Mark, Laphroaig, and Talisker.
The findings, published in Communications Chemistry, reveal that AI systems can offer greater consistency and precision in evaluating whisky aromas, even surpassing human expert panels in certain aspects.
The research team trained AI algorithms using chemical and aroma profiles compiled by an 11-member panel of whisky experts. The AI was then tasked with identifying the five most prominent aroma notes in the samples and distinguishing between US and Scottish whiskies.
The system achieved an impressive 90% accuracy in pinpointing the whiskies’ origins, though researchers noted that this success rate may drop when applied to samples outside the trained dataset.
Dr Andreas Grasskamp, the study’s lead researcher, highlighted the AI’s ability to complement human expertise. Speaking to The Guardian, he said, “This technology is not meant to replace human panels but to provide a consistent and objective method of analysis.”
Vikram Damodaran, Chief Innovation Officer, DIAGEO India told Business Today, "AI has well made its inroads into the spirits industry, and we are excited to explore its potential further. Developing flavour profiles and notes, traditionally an intensive and long-drawn process spanning months, has now become faster and more accurate with the help of AI. This empowers Master Blenders and Taste Masters to experiment with greater precision and creativity."
The study identified specific compounds that distinguish US and Scottish whiskies. For example: • US Whiskies: Menthol and citronellol, compounds linked to caramel-like notes. • Scottish Whiskies: Methyl decanoate and heptanoic acid, which contribute to smoky and medicinal aromas.
The use of AI in whisky analysis could have significant implications for quality control, product consistency, and even fraud detection. The technology’s ability to analyse molecular compositions with precision could streamline processes traditionally reliant on subjective human assessments.
While human expertise remains integral to whisky-making, AI’s introduction offers a new layer of reliability and objectivity, potentially revolutionising the industry’s approach to quality assurance.
"Through initiatives like ‘What’s Your Whisky’ and FlavourPrint (AI-powered flavour profilers), we have harnessed technology to help consumers map their unique flavour preferences, fostering deeper sensory appreciation while offering personalised recommendations from our portfolio. AI’s applications extend far beyond consumer engagement. On the technical front, machine learning systems can now detect subtle changes in aromas during production, enhancing both precision and efficiency in whisky-making," Damodaran added.