With a network of over 3,500 kilometres of railway lines serving nearly 250 million customers a year, NMBS/SNCB, Belgium’s national rail company is responsible for ensuring passengers can travel safely throughout the country. In February and March 2021, Belgium experienced some unexpectedly sunny weather, which resulted in higher than anticipated numbers of customers travelling to the coast by rail.
Against a backdrop of Covid regulations and BE regulations such as sufficient aeration and space, this unforeseen spike in passenger numbers culminated in the limitation of train capacity to coastal cities for safety reasons. NMBS realised that being forewarned is to be forearmed and needed a solution whereby passenger data from website and app activity could be processed in real time to predict customer behaviour based on travel intentions.
These analyses could then be disseminated to internal stakeholders (communication staff, security staff) and external stakeholders (coastal cities and tourism office). This would allow both communication and operational opportunities to make rail travel more efficient, as different stakeholders would be able to prepare better and mitigate potential problems such as overcrowding.
“Our Innovation Lab works on projects designed to test and offer new features and products to customers but also to internal and external stakeholders. Our goal is to deliver value and learnings, fast. This specific project was very important for us, as it allowed us to tackle a big challenge in these Covid times. We had to move quickly.”
Radix was tasked with helping NMBS’s Innovation Lab to develop a data pipeline and dashboards that would let NMBS/SNCB know in advance the travel intentions of their passengers. To reach that goal, the two partners used non-personal, GDPR compliant, data about the number of requests on the website or app showing origin-destination.
The main goal was to provide real-time visualisations of customer behaviour and predict the affluence of passengers from and to the Belgian coast. This would allow NMBS/SNCB personnel and external stakeholders to use the data to take appropriate action, such as organizing passenger flow in stations or sending announcements through the app and website.
The platform had to be flexible, stable and produce precise and usable information with a lightweight front-end. It also needed to be easily accessible online by all relevant stakeholders. End-user satisfaction with the dashboards was paramount.
Radix initially decided with NMBS/SNCB to start with the coastal city of Oostende. The goal was to develop a reliable first sprint very quickly to be used right away to predict passenger flows. This first sprint could then be extended to other coastal stations and improved upon.
Radix worked closely with the team at NMBS and had the first version up and running within just three weeks. The solution architecture was set up through Microsoft Azure.
Radix used passengers’ itinerary requests as a basis to build their predictions. Data generated by the website and app is automatically picked up by a data pipeline in Azure every hour. This pipeline processes the new requests and updates the dashboards, making sure that stakeholders always see the latest information.
There was constant communication and feedback from end-users, enabling Radix to adapt these first sprints quickly and share results. Mock-ups and data were shown to end-users at NMBS/SNCB, who in turn shared their experience and expertise to further improve the solution.
The solution, dubbed the “frequentation barometer”, now allows stakeholders to view passengers’ travel plans to a specific station up to 24 hours in advance.
“We were very impressed by Radix’s agility and expertise. Their ability to deliver after only 3 weeks was remarkable. They managed to keep everything simple and transparent while delivering great value for us.”
The tool has been a huge success within NMBS/SNCB and the company plans to build on it in the months ahead. The first tests were very successful. The results were consistent and allowed NMBS/SNCB to highlight a direct link between the trip searches on its website/application and the effective occupancy of the trains.
This allows the company to anticipate the flow of people on the trains and take the necessary measures. The tool will help the company regulate passenger flows during summer 2021, which is expected to be very busy, as Belgium is coming out of lockdown regulations.
“This project was very successful. It allowed us to provide internal and external stakeholders with reliable and actionable data that they can use to make important decisions. Radix helped us prove the value and the potential of this idea internally. We’re planning on industrializing the tool, and I hope we’ll be able to collaborate again with Radix in the future.”
“We really enjoyed working with Christophe and his team on this project. Our collaboration went really well, and we were able to work together efficiently in a very agile manner. We proved the potential of this idea, and we can’t wait to see it in action this summer.”