AI Applications for Ale
As we saw in Raiders of the Lost Ale, beer brewing has an extremely long history. In fact, the latest evidence indicates it predates our records of history.
But while brewing beer is, in essence, a low-tech process, high tech is applied now to achieve a truly personalized beverage experience. The promise of getting the beer that suits your personal preferences for body, bitterness, etc. is what underlies the business model of IntelligentX.
As explained in the video below, “IntelligentX produces the world’s first beer brewed by artificial intelligence, which improves itself from customer feedback.”
Instead of someone reading the feedback and tweaking according to their understanding of what the customer wants, the brewers “use a complex machine learning algorithm to determine what consumers like about our beers, then brew new versions which are more finely tuned to people’s tastes.”
Three steps to personalized results
IntelligentX’s site presents it as a three step process for the customer:
1. Create your profile upon joining, choose the type of beer you prefer from the four varieties we currently brew – Pale AI, Black AI, Golden AI or Amber AI.
2. Train our algorithm. Let the fun begin! Use our platform to tell us what appeals to your palate and what doesn't. Too hoppy? Not intense enough? You tell us, we'll get tweaking!
3. Enjoy beer brewed for you. It’s now time to sip and savor. And if there’s room for improvement, let us know…your next box awaits!
IntelligentXt then asks: “What beer are you?”
In the video below, Rob McInerney, the founder & CEO of Intelligent Layer and co-founder of IntelligentX Brewing, explains the use of AI in improving everyday products.
The tech trend in beer
This approach of applying AI in the form of machine learning to deliver customized beer made headlines in 2016, and it may have inspired others breweries to consider how machine learning can be applied to improving their own products.
Among those is Kirin Holdings, the second-largest brewery in Japan As reported in the Nikkei Asian Review, in 2017, it partnered with the Mitsubishi Research Institute to work a test program of AI-enabled brewing. The idea was to use machine learning to “determine the desired flavor, aroma, color and alcohol content, then produce the corresponding recipe.”
This is data-driven brewing at its best, considering the company is sitting on two whole decades worth of data to analyze. “It takes a professional brewer a decade or more to hone such skills,” according to the Asian Review.
Brewed in the USA with AI
So AI has been applied to brewing beer in the UK and in Japan, but what about in the United States? It wasn’t going to be left behind in the high tech application.
As reported in Food & Wine, in 2018, "Charlottesville’s Champion Brewing teamed up with the nearby machine learning company Metis Machine to brew their new ML IPA—a computer’s vision of what should essentially be the ideal IPA.”
This was designed to be a very scientific endeavor, and Champion was very selective about the data it used to come up with the recipe of the quintessential IPA. They approached the project as a data scientist would.
Hunter Smith, owner of Champion Brewing Company was quoted in the article, saying, “We provided the parameters on which IPAs are judged at the Great American Beer Festival (SRM, ABV, IBU) and matched that range with the 10-best-selling IPAs nationally, as well as the 10 worst selling IPAs at a local retailer and Metis came up with the results.”
He was proud of being the first to take that approach to data-driven brewing.
AI also solves beer production problems
While machine learning is good at working with data to predict what would improve a beer’s flavor, or at least make it more appealing to a particular type of taste, it’s usefulness is not limited to the brewery. It can also be used to solve problems that lead to losses of product.
As the video above explains, Joe Vogelbacher and Eric Flanigan, co-founders of Sugar Creek Brewery in Charlotte, NC needed to find a way to address their problem of waste that added up to tens of thousands of dollars in losses each month.
When you have foam on your glass of beer, you may find yourself with somewhat less than a full pint, but when you have a foam problem in production, the losses can be huge. In the case of Sugar Creek, the losses attributed to the foam issue added up to over $30,000 a month.
Sugar Creek Brewery decided to look for a high tech solution, which they found in a joint effort between IBM's artificial intelligence and IoT platform combined with Bosch's IoT sensors. The solution led to no more losses and better quality.
As they explained in AI and IoT Help Perfect the Brew at Sugar Creek Brewing Company, “the AI and IoT technology tells my team about many aspects of the beer, which are critical to efficiently creating a quality product. Parameters such as fill time, temperature, pH, gravity, pressure, carbonation, and level are all fed to the IoT cloud for analysis. This data can inform new processes or refine existing ones to ensure our beer meets the high expectations of our consumers.”
Before they implemented the technology solution, “imbalances in pressure and temperature would create foam and waste beer,” as the brew proceeded “from tank to tank in our bottling lin.” That resulted in inconsistencies in bottle fill levels, which meant they had to remove a significant number of bottles from the line, which amounted to a great deal of waste.
“Data collected and analyzed via the IBM Watson/Bosch interface identified an issue causing excessive foaming in the bottle.” That alone amounted to saving over $10,000 a month on the wasted product from bottles that didn’t have the right amount in them.
But the advantage is not limited to the ability to immediately spot a bottle that is not properly filled before it proceeds down the line. They now also “have more controlled, precise fermentations which leads to a better flavor in the bottle.”
Better flavor is something anyone can appreciate. So even those who have issues with AI's impact on society and manufacturing would likely drink to these improvements in brewing.