Karin Lindström
Redaktör

Democratizing IT at Munters

Case Study
31 Jan 20248 mins
Artificial IntelligenceCIOData Center Management

Kenneth Verlage, group CIO and CDO of global climate control supplier Munters, is on a mission to shift productivity from shadow IT to a technically sustainable platform in order to create a franchise operation for the business.

Kenneth Verlage stylized
Credit: Anders Fredriksson

Stockholm-based Munters, founded in 1955, works across a wide range of industries to supply innovative and energy-efficient air treatment and climate solutions. Today, the company, with nearly 4,000 employees in 30 countries and 2022 net sales of SEK10.4 billion [$1 billion], is supplemented with more specific niche systems for data centers and food tech software to support animal breeding and greenhouses.

“We’re growing not least because of digitalization,” says group CIO and CDO Kenneth Verlage. “Right now, we’re on a two-part digitization journey: one toward the customers and one toward our own production.”

The major investment in digitization began four years ago and since then, its budget has more than doubled according to Verlage, who was recruited three years ago as part of that investment.

When he joined Munters, the IT environment was extremely well managed and without a lot of technical debt — a godsend for a new CIO. But conversely, they hadn’t built much of their own, and digital innovation was low. As a result, much of the current work is about getting digital DNA into the company.

“IT at Munters can do different things now than it could three years ago,” he says. “We support business and factories, and they should know we’re there to help them. At the same time, it’s been important that everyone in IT is also part of that development, even those who have been involved for a long time.”

The want to democratize

An important part of that inclusiveness is to democratize IT so development can take place close to the business, in production, and business.

“The great opportunities exist when IT has become part of the normal business and not a separate department,” says Verlage. “That’s why I’m trying to make a franchise out of IT, to democratize it and let regular business development do the work. I want to reap the productivity of what’s been called shadow IT on a technically sustainable platform.”

This means the IT department sets up the systems and regulations, and then allows developers outside the business to build what they want within those frameworks.

“We must allow our factories to develop their own solutions,” he adds. “What we contribute centrally is we have a competence base and provide the technology and integrations to build the applications. But they must not integrate around our APIs or break our search solutions.”

The freedom it’s brought has meant that Munter’s factories now employ their own developers, something that hadn’t happened before. And things are moving so fast even Verlage finds it challenging to keep up with the pace.

For example, during a larger internal meeting at Munters last summer, Verlage expressed it would still take a while to lay the groundwork for generative AI.

“Some developers protested, and I liked that, it’s healthy,” he says. But after the meeting, it turns out gen AI was already implemented in a customer portal. Someone had entered all the manuals and then received answers in plain text where they could also be directed to the correct instructional video, and even where in that video the answer was.

“So fantastic things happen when developers get the opportunity to run away with the customer,” he says.

Pick up the pace

At the same time as they’re now building to increase internal innovation and bring out solutions that the business really wants, Verlage doesn’t think Munters is far enough ahead yet.

“We have to start scaling and we’ve spent a lot of time thinking out exactly how to do it,” he says. “We no longer let in difficult-to-implement platforms without getting people involved, because then they’re not used. We also need more data. But it’s going well now.”

And there are several things that hasten development in the right direction. For instance, Munters uses low code and no code in its factories. When it comes to the customer meeting, it’s about slightly more advanced development because it has to look better. Another contribution is the concept of microservices, piecing together applications that can be shared between business areas and developers.

“Then the organization is also more mature today than before and thinks more digitally,” says Verlage. “And with the help of this, I try to set the stage, let it out in the business, and have control only where you need to have control.”

Creating digital twins

To get the data in, Munters creates a digital twin for each machine it sells, and it also makes sure it’s clear it owns the data that’s generated.

From the moment an order for a machine comes in, the digital twin is born where all specifications are included. Then it follows through the production of its physical twin and knows exactly what fan and motor are inserted, and the serial number. It also has the information about where it’ll be delivered and when it’ll arrive.

“We get use of the digital twin because it keeps track of everything happening on time, and can warn if there are delays,” he says. “It’s very important for many customers.”

The digital twin also follows when the physical twin is installed — knows who’s doing it, notices deviations, keeps an eye on software, and then sends data from the physical part into the digital one.

“We only need to apply intelligence to the twin and not have it in the machine,” he adds. “It simplifies installation and maintenance, and we can handle everything comfortably at home from our servers.”

Safe from hacking

There’s also a security aspect to it all. For machines with sensitive information, it’s possible to ensure there’s only information going out to the digital twin, but it’s not possible to send information back to the machine to avoid it being hacked.

“If it were possible to shut down the cooling in a data center because we’re hacked, that would be a bad day at work for me,” he says. “So in some places, we might have a physical barrier where we need to go outside to apply intelligence to a machine inside the firewalls.”

Data coming in from all the digital twins can bring quick wins such as being able to save energy by setting up the machines as precisely as possible — an area where there’s still a lot to do, according to Verlage, who explains that today, it’s good to set machines with added difficulty to be on the safe side. But when you have accurate data, it’s also possible to set them exactly, so no extra energy is wasted.

But perhaps most importantly, if all the data that comes in can eventually be analyzed with AI, Verlage expects this can lead to many new insights and business opportunities.

Three guiding stars

Verlage has three guiding stars for how he should work with AI. The first is with agile to develop, implement, and go into operation quickly. The second is a methodology that comes from Google — pretotyping.

“Here, the ‘pre’ stands for pretend; you don’t have to have a prototype,” he says. “It’s enough to pretend you have one. Talk to the customer about something you don’t have but can do. If you’re going to test a new cell phone, make it in the shape of a wooden block and let people feel it in their hand and see if it initiates a spark.”

And the third star is what’s called “radical simplicity,” to make everything as simple as possible. To launch small, very simple applications and then improve them together with the customer.

Verlage says Munters recently built a customized customer portal where orders could be seen and tracked. And in one instance, on day two, the customer discovered an order was on its way to the wrong building and it was stopped and fixed.

“It was a wow experience for the customer, and shows that even simple things can quickly come in handy,” he says.

Leave the old

This approach to get things done and out quickly is very important to Verlage, and he thinks it’s often been strangely difficult for the IT business to cope with. Instead, large systems are often implemented for a long time, yet neglected.

“We can’t allow such waste and introduce expensive systems that aren’t used,” he says. “If we do, we’re doing something wrong. We have to leave the old world and digitize in a different way into the future. I’m almost obsessed with getting things into production.”