Asia Brewers Network

Resistance Is Futile: Artificial Intelligence And Beer


In yet more potentially nightmarish news to round off 2020, artificial intelligence continues its slow yet steady assimilation of the beer industry.

From brewing and logistics to marketing and sales, breweries large and small have begun to integrate AI technology into every possible aspect of their business.

The range of applications stretches beyond from building customer feedback loops to developing better recipes and maximising brewery efficiency.

One of the earliest claims of integrating AI technology into brewing was made it 2016 by the IntelligentX Brewing Company.

Generating extensive media coverage at the time, IntelligentX used AI to gather and process customer feedback at scale – a process that co-founder Dr Rob McInerney explained during a TEDx talk a year later.

The brewery’s process was surprisingly simple, contrary to many densely technical AI applications.

Drinkers would visit a Facebook Page listed on bottles to give feedback on their beers, where a chatbot would ask them ten questions.

The chatbot in turn processed the accumulated review data and sent analytics to human brewers in order to improve and adjust different aspects of their beer recipes.

Asian Brewing’s AI Adoption

At a much larger scale, some of the world’s biggest brewers have applied AI technology to tackle genuine business challenges in recent years.

Japan has been one of Asia’s early leaders in applying AI technology to the beer industry.

Japanese brewer Kirin worked with Mitsubishi Research Institute in 2017 to apply machine learning to analyse two decades of brewing data, claiming they were using algorithms and sensors to take over roles traditionally taken by us humans to craft beer recipes.

While this may seem unnerving and potentially a strategy to reduce brewery headcounts, the counter-argument to such programs is that they’re able to boost profitably by accelerating product development and improving existing beers.

While there has been limited reference to the project since then (besides a favourable reference to incorporating AI into manufacturing processes in Kirin’s 2018 Annual Report), this could also be out of fear of revealing trade secrets or having their strategy copied by competitors.

Asahi, another well-known Japanese brewery, has claimed to have deployed artificial intelligence in digital marketing initiatives.

The brewer used predictive analytics to drive a higher return on ad spend (commonly referred to in the advertising industry as ROAS), specifically to identify and engage customers more receptive to their marketing messaging on messaging app LINE.

A Global Shift: Artificial Intelligence in Brewing

Beyond Asia, there are a range of other stories about the use of artificial intelligence and its related fields in major craft breweries.

American brewery Deschutes deployed machine learning technology and predictive analytics in their brewing process.

The objective of their program was to more efficiently manage fermentation progress and free up tank capacity, which was reportedly achieved 36 hours per fermentation.

Carlsberg launched the ‘Beer Fingerprinting Project’ to better measure the aromas and flavours in their beer in 2018.

Working with Microsoft, Aarhus University, and Technical University of Denmark, Carlsberg fused together with the use of specialist sensor hardware,  he three-year project’s objective is to ‘map a flavour fingerprint for each sample and reduce the time it takes to research taste combinations and processes by up to a third, to help the company get more distinct beers to market faster.’[1]

Neural Networks For Home Brewing

Homebrewers aren’t immune to the lure of incorporating AI into their brewing processes, too.

More recently, a keen homebrewer and professional engineer at NVIDIA Eric Boucher deployed a blend of his employer’s technology, open-source software and a neural network to generate a range of beer recipes.

Neural networks are a form of machine learning where a program learns to recognise patterns from data.

Eric mined recipe data on ales from homebrewing websites and then fed over a hundred recipes into Textgernn, software that trains neural networks using text.

The AI-generated a series of recipes, with the engineer ultimately selecting and brewing a blonde ale (‘Full Nerd #1’) with an unusual hop bill (Warrior, Cascade, and Amarillo). 

Prepare To Be Assimilated

While homebrewing on a 20-litre system is certainly a different beast to commercial-grade operations, the possibilities AI presents to our industry seem immense.

The existing commercial applications of the technology in areas like recipe development, improving brewery efficiency and marketing are likely to see more widespread adoption in coming years amongst early adopters and larger commercial brewers.

Machine learning, neural networks and other related technologies may present moral dilemmas in our industry, too. If machines drive everything from recipe development and brewing through to marketing and managing consumer feedback, are we still an industry driven by craft – or bits, bytes and steel?


Article by:

Oliver Woods

Oliver Woods

Oliver is a marketing strategist by trade and a craft beer enthusiast by choice. He is the co-founder of Kakilang Brewing Company, a nomadic nano-brewery and lives, works and drinks in Saigon (Ho Chi Minh City), Vietnam.

Share this article