Dairy tech: Arla Foods’ new AI tool predicts how much milk 1.5m cows can produce
With the new Artifical Intelligence (AI) software, calculating milk production is reduced to “a few hours”
12 Jun 2019 --- Arla Foods has developed a new artificial intelligence (AI) tool to better predict its milk intake from farms. According to the company, 200 million kilos of milk can now be utilized more efficiently each year, through predictions of how much milk 1.5 million cows will produce in the future, elevating value chain sustainability. Prior to the tool’s use, a forecast of this scale took days to create and was carried out through manual calculations of “piles of Excel sheets,” Arla notes. The new AI software speeds up the process, which is reduced to “a few hours,” and yields a “1.4 percent greater accuracy.”
The new milk intake forecasting tool is implemented in all Arla’s markets across Europe, including Denmark, Germany, Sweden, the UK, Belgium, Luxembourg and the Netherlands.
“Arla’s in-house data scientists worked alongside our IT supplier to develop the AI system, using a 10-week sprint to present working pilot version and then 6 months to refine into a fully operational tool,” Torben Fabrin, Senior Vice President of IT at Arla, tells FoodIngredientsFirst. “In terms of shaping the future of the food and beverage industry, AI can have a key role to play. Ultimately, it’s about predictions. Whether this is forecasting where our raw ingredients need to be and when in order to reduce waste, or whether it’s helping us identify which new products and innovations will be successful based on previous learnings, these predictions can allow us to be more efficient. For us at Arla, this is part of our ongoing sustainability journey to develop a more environmentally friendly supply chain so we definitely see other opportunities for artificial intelligence around our business.”
“The better we are at predicting what our milk intake will be, the better we can plan and optimize our entire value chain, which both improves profitability for our farmer-owners and drives sustainability. The new AI tool provides us with an insight into our supply of milk that we have never had before,” says Michael Bøgh Linde Vinther, Director of Global Milk Planning at Arla.
Each year, Arla collects around 13 billion kilos of milk from some 10,300 farmer-owners across Northern Europe. The cooperative is now keeping a close eye on emerging technologies that raise the efficiency of dairy production. “More data drives better decision making,” the company emphasizes.
Arla’s AI tool factors in variables such as seasonal changes, the number of farmers converting to new milk types, the farmers’ geographical characteristics as well as how much milk they produce on a daily basis.
“We are now able to make important strategic decisions on a more informed basis. The data has become more valid as it is now formalized in a bulletproof system rather than based on individual knowledge. It’s amazing to see how this new technology is able to optimize and improve an, up until now, very time-consuming task,” says Vinther.
As an example, the company says that it is now possible to make the distinction between how much milk should be collected from farmers in North Germany and West Germany three to five months ahead of time.
“This kind of knowledge is very valuable, as it allows us to be able to plan and adjust the number of Arla trucks traveling across the country. In that way, we can reduce our costs and save the environment from unnecessary CO2 emissions,” explains Vinther.
In March, Arla launched of an ambitious target to accelerate the transition to sustainable dairy production with an intensified focus on farming processes. The main target highlighted is a goal to reduce greenhouse gas emissions by 30 percent per kilo of milk over the next decade and to work towards carbon net zero by 2050.
According to a recent analysis from the UN’s Food and Agricultural Organisation (FAO), global milk production continues to become more efficient and sustainable with a global average of 2.5kg CO2 per kilo of milk. Arla farmers contribute to this result with an average emission intensity of 1.15kg CO2 per kilo of milk, about half of the global average, notes the company.
AI is shifting the food and beverage landscape
Industry looks to AI for its big data crunching capabilities. Increasingly, such digital technologies are offering valuable utility in improving traceability, trust and ultimately shape food supply chains to withstand the predicted challenges of the future. A rapidly growing urban population and increased consumer awareness regarding eco-friendly sourcing are prompting food companies and manufacturers to employ digital solutions to ensure product safety, quality and sustainability.
In April, food waste initiative Winnow launched an AI innovation coined Winnow Vision, which promises to revolutionize food waste management in commercial kitchens to benefit businesses and the environment. Winnow Vision utilizes a camera, a set of smart scales and machine learning technology to recognize the different foods that are discarded and calculate the financial and environmental cost. This may allow foodservice businesses to adjust their food purchasing decisions and reduce spending while tackling food waste.
In the same month, the world’s first AI-developed whiskey was launched by Swedish distillery Mackmyra, in collaboration with Fourkind, a Finnish technology consultancy specializing in AI spearhead projects, and Microsoft. The companies headline this breakthrough as the first time a complex consumer product recipe has been created with machine learning, which may further benefit food and beverage producers in a broader range of applications. Through a dataset analysis, the AI can generate more than 70 million recipes that it predicts will be most popular and of the highest quality.
A 2018 CSB-System survey found that decision makers within the food and beverage sector believe that digitization will have a “huge role to play” in the future of the industry. This is despite several challenges, including the lack of employee skills and low awareness of what technological solutions are available on the market. Even with large strides being made in this forward-looking space, investment in IT also remains “significantly low,” the survey finds. As such, it may be a while before most novel technologies are seen as ubiquitous across the industry.
By Benjamin Ferrer
