Dutch Railways is using big data to keep trains moving — here’s how

While most Dutch trains arrive on time — 92.6 percent, to be exact — some delay is inevitable. Falling leaves can make the railway slippery. Snow sometimes causes the railroad switches to freeze up.

Machinal issues, on the other hand, such as broken doors or overheated brakes, can increasingly be fixed before they cause severe delays. Thanks to thousands of sensors and big data analytics, train maintenance has become much more efficient.

It’s a development that took a flight since 4G became available in the Netherlands, allowing data to be processed at a much higher speed. This includes check-in data, status measurement points on tracks, and information sent by sensors in the trains. All this data continually flows through the Data & Analytics (D&A) division of Dutch Railways (NS) through more than 140 sources.

“What we do is crucial. Without data analytics, we’d be lost,” Martijn Scheele, head of the D&A department, says.

It’s this data that keeps the Dutch rail network moving, and helps NS deliver a better, safer, and easier journey for train passengers. Scheele’s team services the whole Dutch rail network, its customers, and the government.

“We are one of the few departments that serve the entire NS organization, which makes the work here special,” Scheele says. More than 160 NS employees, who all work on innovative applications, use big data every day. And because we expand our team with 20 new colleagues every year, big data is becoming a big deal.”

Over the years, D&A has taken a central role at the NS. Business Intelligence converts a wealth of data into reports and dashboards to aid continuous improvement and the management information for the organization.

Although the average traveler may not notice it, behind the scenes their travel experience is being continually improved by an orchestrated use of Advanced Analytics.

Using existing data, new applications are constantly being designed by the creative minds in D&A. For example, the “seat seeker” app was developed on this basis. Prorail measuring stations were used to calculate the passenger numbers based on the weight of trains. This data is then used and presented in such a way to help travelers find a train that’s likely to have seats available.

NS’s IT department is a diverse bunch, ranging from traditional developers to statisticians.

“There’s a certain degree of standardization in the products we make, which means that creativity is appreciated, but you don’t have to invent everything yourself,” Scheele says.