Optimizing Beer Taste Through Artificial Intelligence

Researchers in Belgium utilize artificial intelligence to enhance beer flavors while emphasizing the crucial role of brewers.

Belgian beers are renowned for their diverse flavors and rich heritage, with fruity lambics and complex Trappists being among the favorites. Recently, scientists have delved into the realm of artificial intelligence to further refine these brews.

Professor Kevin Verstrepen from KU Leuven university spearheaded a study demonstrating how AI can decrypt the intricate relationships within human aroma perception. Beer, like other food products, contains a myriad of aroma molecules detected by our senses, and these compounds interact with each other, influencing how we experience each flavor component.

The research published in Nature Communications involved scrutinizing 250 commercial Belgian beers of various styles such as lagers, fruit beers, and ales. The analysis focused on factors like alcohol content, pH levels, sugar concentrations, and over 200 flavor-related compounds like esters and terpenoids.

Additionally, a tasting panel meticulously evaluated the beers, considering 50 attributes like hop flavors, sweetness, and acidity over a duration of three years. The researchers also incorporated 180,000 beer reviews from RateBeer to compare consumer sentiments with the panel's ratings, revealing significant correlations in evaluating aspects such as bitterness, sweetness, and aroma.

Verstrepen highlighted the impact of slight chemical variations on overall taste perception. Surprisingly, some traditionally off-putting elements could enhance flavor when present in lower concentrations and combined with other aroma compounds. By employing machine learning techniques, the team developed models to forecast the taste and appreciation levels of beers based on their composition.

With these models in hand, the researchers modified an existing commercial beer by introducing specific substances identified as critical predictors of overall enjoyment, like lactic acid and glycerol. The results, based on feedback from the tasting panel, showcased enhanced ratings across various categories for both alcoholic and non-alcoholic beers.

Despite the potential of AI in refining beer flavors, the researchers emphasized that the expertise of brewers remains indispensable in bringing these optimizations to fruition. While the AI models can propose chemical alterations to enhance beer quality, it is ultimately the brewers who translate these recommendations into actual recipes and brewing processes.