Solution: Building a Flexible IoT Platform
CFL found a small set of vendors who offered the type of flexible and horizontally scalable IoT platform it needed and invited each of them to demonstrate their solutions.
“It quickly became clear that Hitachi Vantara was the right partner for us,” says Arnaud. “Hitachi’s experience in the rail industry meant that they spoke our language, and we were impressed by the team’s project management and customer relationship skills, not just their technical expertise.”
For example, during the evaluation phase, Arnaud deliberately asked each vendor to solve a problem that he knew was technically unfeasible, to see how they would respond.
“The Hitachi Vantara team were the only vendor that faced up to the challenge. They explained that they couldn’t do exactly what we were asking for, but they could achieve the same result by another method. And we’ve seen that same attitude throughout our relationship. ‘Impossible’ is not part of their vocabulary, they always find a way.”
The Operational Technology team made a business case for adopting Hitachi Lumada Data Integration, delivered by Pentaho as the core of its new IoT platform, and won approval from the CFL board for a one-year trial. The next challenge was to demonstrate to line-of-business teams that the platform could deliver real value.
The team began by porting CFL’s various existing IoT solutions over to Lumada Data Integration, which proved the technical capabilities of the new platform. It also led to savings of several thousand Euros per month by cancelling subscriptions to external IoT vendors—delivering immediate value to line-of-business stakeholders and convincing them to advocate for the new platform.
Next, the team began developing new IoT use cases that would demonstrate the full potential of the platform. For example, CFL is now using Lumada to capture video data from CCTV cameras in its station parking lots and apply artificial intelligence to predict the number of parking spaces that will be available.
“We can already predict parking availability with 95% accuracy 24 hours in advance, and over 97% accuracy one hour in advance,” says Arnaud. “And that’s just the start. As we improve the quality of the data coming in from the cameras, we should be able to get even better predictions.”
In the future, CFL aims to make parking insights available to passengers via an online portal, so that they can check whether a parking space will be available before they start their journey and can make smarter decisions about which station to use.
Another use case will make it easier for CFL to maintain regulatory compliance. When trains cross the border from Luxembourg into Germany, the German railway regulator stipulates that the wastewater tank in the train’s restrooms must be less than 50% full. CFL is now using onboard sensors to monitor the water levels in real time, so it can discharge any excess before entering Germany.