MILKING DATA FOR A SMARTER CHEESE FACTORY
ANY WAY YOU SLICE IT, ANALYTICS CAN OPTIMIZE
CHEESE PRODUCTION - AND REDUCE FOOD WASTE
MILKING DATA FOR A SMARTER CHEESE FACTORY
ANY WAY YOU SLICE IT, ANALYTICS CAN OPTIMIZE CHEESE PRODUCTION - AND REDUCE FOOD WASTE
cheese-sidebar-top-left

Colin Nugteren and his team at Dutch analytics firm Notilyze already knew plenty about using real-time data to optimize the manufacturing process.

But their expertise took a tasty twist during the 2023 SAS Hackathon, when a Dutch cheesemaker approached them for help improving quality and output while also reducing food waste.

“Of course, our team knew zero about a cheese factory or creating cheese, but there are a few cheese technologists – that’s an actual job – and they are very knowledgeable about cheese,” Nugteren says. “They know everything about this whole process, but it takes them a lot of time to extract data to put it in Excel or small tool so they can analyze it.

“They said, ‘We have all this data now, but analyzing it to optimize the factory yield or to improve some processes takes us a lot of time. If you have any idea on how to improve that, that would be great.’ So we told them: challenge accepted.”

 

DAIRY MEETS DATA

The four major factors the Notilyze team looked at that can affect the quality and yield of cheese are salt, fat, moisture and pH. For a cheesemaker that produces 100 million kilograms of cheese annually, any imbalance of those ingredients can be costly – to the bottom line and the environment.

“You need to recognize trends soon as possible and make changes to make sure that the ingredients are optimized for each other,” Nugteren says. “They were really surprised by the amount of power that is in the data and how much we were able to gather, in just a few weeks from a lot of data, learning so much about producing cheese without having a cheese background.”

During the month-long Hackathon, the Notilyze team was able to design a dashboard exploring all components of the cheesemaking process, build models to predict yield based on these components, and develop an API that suggests adjustments to the process based on the type of cheese, current control settings and actual observed data.

cheese-sidebar-middle-right

Needing to do all of that in such a relatively short period of time, Nugteren was confident that SAS Viya was the right AI and analytics platform for the job.

First off, a factory produces a lot of data, so the solution had to be scalable. It also required performance levels that would let the team iterate through a variety of models quickly given the time constraints. And finally, it had to be fully integrated – for the experts at Notilyze to collaborate with one another and for the cheesemaker’s non-programmers to take advantage of low- and no-code options.

“The two technologists that were working with us were impressed with the possibilities and how smooth the whole Viya interface was,” Nugteren says.

 

FOCUSING ON THE FOOTPRINT

An expected increase in yield of 7% when optimized – that would be roughly 7 million kilograms of cheese annually – would also mean far less food waste in the manufacturing process.

Perhaps that’s one reason why the cheese technologists’ colleagues were excited, if initially a little incredulous, that this finicky biological process of making cheese could be visualized, analyzed and refined using data.

“We explained what we were aiming for, but also how much collaboration happened so that we could actually understand what we were doing,” Nugteren says, “and that our goal was not to replace them, but to help them become more efficient.”

With the approach embraced, the next step is making sure it continues to be effective once implemented. While cheese gets better with age, the same can’t always be said for analytical models.

Nugteren says he was comforted on that front by Viya’s model management and governance, which automates the process of ensuring models are still working as they should over time. That will allow the cheesemaker to maintain its data-driven gains over time and be more sustainable not just from an environmental perspective but from an analytical one.

“There’s a huge change coming for these businesses, and it’s very interesting,” Nugteren says. “There are many factories that need to be more aware of their ecological footprint, and it’s something that our colleagues really appreciate, the way you can combine something for business and something for good.”

RECOMMENDED FOR YOU

GEN Z

A SHARED VISION FOR PROTECTING SEA TURTLES

How crowd-training a computer vision AI model can help track sea turtles' health over time.

→ READ MORE

DATA VIZ

FILLING IN THE BLANKS

Can a writer prove that even data newbies can gain insights from exploring the data in their daily lives?

→ READ MORE

Curious about SAS and the analytics that empower organizations everywhere?

GET TO KNOW US