Carlsberg’s future beer created with AI

Denmark ranks as one of the world’s most innovative countries and is among the most digitized in the world. At the same time this tiny Scandinavian country has an old beer industry which is in need for innovation.

In order to strengthening the Danish position in the world beer market represented by Carlsberg, the old Danish beer company launched in November last year a sensational technology collaboration based on artificial intelligence with Microsoft, Technical University of Denmark (DTU) and Aarhus University under the name The Beer Fingerprinting Project.

The Beer Fingerprinting Project will run for three years and is funded with 4 million USD by Carlsberg and with 2,8 million USD by Innovation Fund Denmark, a fund that was established by bringing together research, technology development and innovation grants.

Hundreds shades of beer

To understand the essence of it all, it requires a simple introduction to beer production:

Beer basically consists of barley, water, yeast and hops, which together can create over 400 flavors. A person with well-developed taste buds can maximally taste 150 of them.

In addition, Carlsberg works with between 100 and 1000 yeast samples daily. All of these have a bearing on how the final beer will taste.

Each player contributes with expertise

In order to take the production one step further with the help of AI-technologies, The Beer Fingerprinting Project team members contribute with expertise.

Microsoft develop the software, which is necessary to interpret the data. In other words Microsoft supplies the computing power and the artificial intelligence. Microsoft’s AI solutions including machine learning and the digital cloud platform which enables selection and development of innovative brewer’s yeast at a much higher speed and quality. It will reduce time and cost to develop new beers and the developed sensors can cut the 8 to 24 months needed to develop a new beer by one third.

The other players in the project, Aarhus University, develops the sensors, and Technical University of Denmark DTU is responsible for data processing.

Flavors becoming algorithms

Advanced sensors are coded to detect how beer tastes, and the many flavors are converted into algorithms.

The algorithms collect data that compares, in order to create the perfect taste by measuring, for example, sweetness and thus ensuring that the beer does not taste incorrectly.

Already, Aarhus University has developed sensors and a prototype that can taste a difference in flavors between Carlsberg and Tuborg Pilsner, Wibroe and Nordic.

Advanced analytics and intelligent cloud technology are a corner stone of the project which also expected to spark innovations beyond the brewing world, as the technology can be used across other industries.