sakkmesterke - stock.adobe.com
NS has 150 different data sources that are currently being brought together in Microsoft’s Azure cloud. That project started in 2017 and was called ACE: Adoption Cloud Ecosystem. “The most important sources have now been upgraded and are available from the cloud,” he said.
The biggest challenge now lies in changing the mindset of employees, he explained. “Maintenance engineers have been working from a maintenance book for years,” said Scheele. “When we, IT-people, suddenly told them: ‘Based on the data we get from various sensors, this certain train needs maintenance now,’ it was sometimes difficult to convince people to work in a new way.”
After all, on the basis of data from trainsets, NS is now able to optimise many operational processes and customer services. For example, the Dutch rail network is equipped with so-called Gotcha sensors, which measure the load on the train’s wheels.
These were traditionally used for freight trains, but now the data is also used to predict when a trainset needs maintenance due to wear and tear on the wheels. “Every trainset has a tag which allows us to recognise which trainset it is,” said one of the maintenance development engineers at NS Techniek, in a video. “When such a trainset passes over a Gotcha measurement station, the rails bend, so we can determine the weight on each wheel of the train.”
This system has been in use for years to measure wear points on the wheels. “But it now turns out [through data technology] we can also use this system to [automatically] see whether the weight on the train is neatly distributed over all those wheels. If that distribution is not optimal, there is a greater risk of derailment, which is dangerous.”
It was possible to measure this before, but it was a very time-consuming and complex job. Now, it can be measured automatically at any time throughout the country.
The data from the Gotcha measurement stations is received by ProRail, the rail infrastructure management company in the Netherlands. NS requests the data set and with the help of the Data & Analytics department it combines the data with other sources to make results available in an orderly manner.
The calculations of the Gotcha data are shared with the mechanics in the workshop via a dashboard, so they can see whether the trains coming in for maintenance need additional adjustment at the wheels to optimise the weight distribution again.
“By signalling a sub-optimal weight distribution in time, we can increase passenger safety and also prevent further damage to the rolling stock,” said Scheele.
Read how rail companies are using the latest technology
- The Trainline is encouraging customers to discover cheaper rail fares for the same journey with its AI-enabled SplitSave option.
- Northern Rail will be resorting to technology to solve some of its many issues as the troubled rail operator is brought under government control.
- Swedish rail company tells Computer Weekly about its successful mission to improve customer experience.
Scheele pointed to another service that was developed with the help of his department: the NS Seat Finder. “Everyone who travels by train wants to sit down, so for us it’s a continuous puzzle to make the right amount of carriages available on a route.”
KPMG calculated that optimal seat occupancy yields many millions in economic value for NS. Moreover, the correct information ensures that NS does not keep more trains in reserve than strictly necessary.
“The savings we were able to achieve there ran into tens of millions of Euros,” said Scheele.
But even more important is the customer experience, one of the pillars of NS. By combining data from the trains with passenger movement data on the platform and rush-hour statistics, the carrier can distribute passengers evenly across train sets.
“Within the NS travel app, we’ve built a feature where the passenger can see which carriages are full and where there is still room before the train even enters the station,” he said. “We can show this in real time using the Gotcha data. This way, a passenger can stand at the right point on the platform to get a seat.”
When it’s extremely busy, NS can also deploy extra trainsets on the basis of this data. “We do everything to make the journey as optimal as possible for the customer,” said Scheele.
Save on waste water
Another way NS is able to make concrete savings using data is its approach to train toilet waste water tanks. One of the latest train models, the Flirt, has a highly advanced toilet, as NS believes it’s important that passengers can always use a toilet during their journey, and why the holding tank was emptied daily.
However, using sensors in the waste water tanks, it could be determined that this was much less often needed. “We even found that we could do the maintenance of the toilets ourselves, instead of outsourcing it to a maintenance company, because we didn’t have to do it so often. That has saved NS a lot of money,” said Scheele.
NS is determined to make increasing use of data in the coming years in order to achieve further process improvements and cost reductions. In the coming period, for example, the emphasis should be on the 400 stations in the Netherlands.
“If, for example, we can combine GIS data with real-time images of a drone flying around the stations, and with historical maintenance data, I think we can predict much better when we will need to carry out maintenance on stations,” he said. “If we can do that predictive maintenance better, we can save many millions.”
Scheele’s greatest wish, as the driving force behind the Advanced Analytics area he introduced at NS, is that information will soon be available in such a way that good decisions are always made by the company or the passenger.
“This means that as an organisation we will be able to provide a very smooth service, with travellers having no idea what’s happening in the background,” he said. “As a traveller, I want that. Whenever things do go wrong or threaten to go wrong, I immediately receive the most up-to-date information to complete my journey in the smoothest way possible. Even if it’s not with the NS. Getting that information to the passenger at the right time is crucial.”
Data has to get into the DNA
Scheele’s biggest challenge is to create a solid data culture throughout the company. He works with 175 people in the Data & Analytics department, which increases by about 20 people a year.
“We’re one of the few departments in the NS that serves the entire company, but we need to get more out of the technology and build trust in all parts of the company to get started with data,” he said.
“That ‘data integration first’ policy must be embedded in the DNA of the entire organisation. That means that when we’re building an app, we need to think about how we’re going to implement it so that other people will also act on it. That really is a big challenge for us.”
Another challenge for the company is to properly unlock data, and Scheele is setting up a separate programme for this next to ACE.
Ultimately, he said, all data should contribute to NS’s core objective: providing a comfortable journey in a clean train with sufficient seating and a timetable in which trains run on time between stations.