Danske Statsbaner’s increases productivity up to 30% with help from Microsoft AI solutions

Talking about his company's mission to deliver better services for its passengers and how technology can help it achieve that goal, Emil Møller Marker, Vice President for AI at Danske Statsbaner, says, "If we want to get more people out of their cars and into our trains, the question that always has to be top of mind is: How can we use data and new technologies to provide efficient, more convenient, and more relevant rail journeys?”

Only 8% of journeys in Denmark are by train, while the majority are made by car. Motor travel is a huge part of the personal transport market, though often not as sustainable. “So, how can we convince them to leave their cars behind and trust in rail?” asks Marker.

Danske Statsbaner (DSB) is the largest Danish train operating company, and the biggest in Scandinavia, with around 6,500 employees. Because trains are working across Denmark all day, every day, DSB needs robust solutions that deliver all year round. “We think part of the answer involves AI and machine learning,” explains Søren Friis, IT Director at Danske Statsbaner. “That was why we’ve decided to launch a program that aims to transform the processes that we have today using emerging technologies, making it not only IT-led but more business-led. To do so, we need to have every part of the business working with us on this journey.”

Friies concludes, “Microsoft has been important; it assisted us by providing the technology and expertise to explore the challenge—and also to help us power the cultural change we need to make the AI-powered solutions we need a reality.”

AI for all fuels business transformation

In recent years, the company has embraced a variety of AI-powered and machine learning initiatives, at the heart of which is a belief that the technology could help transform the way DSB does business. One area where the company saw an immediate opportunity to improve the rail system was in the cleanup of graffiti. Ensuring the stations and rolling stock are kept clean was costing the company a substantial sum each year. The rail operator founded a digital lab to investigate how new technologies could best be developed and applied practically. But the advent of generative AI (GenAI) has been a game-changer for the company’s approach to adoption.

“We saw huge potential in using GenAI when it first came out,” says Friis. “Up until then, we would start individual initiatives where a department would be the internal customer of a very specific solution—whereas GenAI andMicrosoft 365 Copilot are potentially for everyone.”

The company realized that besides GenAI being a more democratic (and democratizing) technology, blanket deployment was essential to maximizing its utility for the organization. Friis explains, “It’s hard to see how the payback is on the individual little task that you’re setting up something for. But cumulatively they contribute to the overall business case.”

The team also had an intuition that AI and machine learning would have an important role in taking best advantage of the vast pools of data DSB continually collected, but so far made limited use of. “We had a lot of existing data at our disposal, distributed in various silos in a rather unstructured way,” recalls Marker. “But we could not easily make a strong business case for centralizing, organizing, and structuring this potentially valuable resource. We just thought it must be a positive thing that would help create a cultural transformation in the company which would pay off in the longer term. But essentially, we were making a leap of faith.”

Pooling data and making use of assets

The rail operator decided to wholeheartedly embrace AI technology and was keen to investigate available technologies and their potential impact on the business. “Microsoft started talking to us about how Microsoft Copilot could be used as part of a Microsoft 365 solution and how that could support DSB, which is why we chose Microsoft 365 Copilot,” explains Marker.

“DSB needed to collect more of our data on the same Microsoft Azure cloud,” adds Friis. “So we started to centralize more on one single data lake. Similarly, with OpenAI coming in, it was Microsoft who advised us on the technology. Importantly, we could also take advantage of Microsoft partner networks to enlist the right resources for our objectives."

“Finally, the team at Microsoft has supported the education and training of the employees, not only within specific organizational areas, but also in the way we are trying to adopt the new technologies on the more general business side,” Friis concludes.

Spotting graffiti and enhancing security with AI

DSB has been experimenting with a range of different initiatives that lean on both tech and data. One is how best to use AI to help deal with the company’s graffiti challenges. "We’re using cameras to identify graffiti on our trains and carriages and identify which ones to prioritize for cleaning and removal, saving us the time and costs that we usually spend on manual inspections,” explains Friis.

Another idea is to use a similar camera-linked system to improve railway safety by recognizing when people unlawfully enter the railway line, posing a danger to themselves and others. “Our next step will be to build a proof of concept for one station to test the viability of such a system,” says Marker. “Then we need to consider scalability.”

DSB is not alone in its commitment to AI as an integral part of its business strategy in Denmark. According to business journal Børsen, the rail operator is part of a wave of early adopters that are enjoying 15% to 30% improvements in productivity.

“We’re starting to get a better understanding of what is actually needed to deliver durable solutions which can work in an operational environment,” explains Marker. Enthuses Friis, “Right now, we have a lot of data and a lot of tech, but we need more of both to achieve our aims in full.”

Powering rail travel with data

The possibilities to use data to improve rail service in Denmark are set to grow considerably, according to Friis. “The rolling stock that we’re procuring now comes with the latest technologies and IoT enablement. So, the number of data points is mushrooming. Our data strategy is also moving up the agenda, and we’re starting to get a better understanding of what is needed to deliver these solutions and how to make them durable enough to work in an operational environment.”

Friis also understands the need for a commitment to ongoing cultural transformation. “We need to ensure that we never lose sight of how to apply that data in the business,” he says. “Lack of data is no longer the main hindrance for achieving our goals. What we need to do is keep on applying radical thinking to our business processes, and we know that Microsoft can help us get there—and beyond.”