The Hidden Data Map That Could Transform How Railways Talk to Passengers

Every time a traveler checks a platform display, opens a journey planner, or receives a delay alert on their phone, a vast and largely invisible machinery of databases springs into action behind the scenes. The quality of what passengers see on the surface depends entirely on what flows underneath: timetables, occupancy figures, infrastructure records, and real-time feeds that must be accurate, accessible, and correctly classified. Yet while researchers and operators have long focused on refining the information they already possess, far less attention has been paid to a more fundamental question — how do railways actually obtain, organize, and map the data their passenger information systems require? A new study published in Scientific Reports tackles that question head-on, offering a Europe-wide conceptual framework that could reshape how the rail sector thinks about its most urgent data needs.

The research, led by Pranjal Mandhaniya, Nils O. E. Olsson, and Patrick Emanuel Urassa of the Norwegian University of Science and Technology in Trondheim, together with Malcolm Lundgren of the Swedish Transport Administration and Helge Bontveit of the Norwegian Railway Directorate, emerges from the FP6 FutuRe project, partially funded by the European Commission through Europe’s Rail Joint Undertaking under the Horizon Europe program. The team’s motivation is deceptively simple: improving railway passenger information systems is a potential method to increase ridership and customer satisfaction, but high-quality information can only reach travelers if the underlying data are sufficient. In an era when rail operators across Europe are competing with cars, airlines, and each other for climate-conscious passengers, the difference between a seamless journey and a frustrating one often comes down to the data plumbing that nobody sees.

What makes the study unusual is its focus on the road to obtaining the data rather than the data themselves. Most researchers, the authors note, naturally concentrate on the datasets they happen to be using, leaving a conceptual gap in how the rail sector characterizes and prioritizes the databases required for passenger information systems. To close that gap, the team encapsulated experiences with mapping databases across a Europe-wide focus, drawing on the perspectives of national railway directorates, infrastructure managers, and transport authorities. The result is not a new algorithm or a flashy machine-learning model, but something arguably more consequential: an applied conceptual refinement that gives operators a shared language for discussing what data exist, what data are missing, and which gaps matter most for the passenger experience.

At the heart of the framework lies a three-axis classification scheme that characterizes railway data along an accessibility-suitability spectrum, a demand-supply spectrum, and a dynamic-static spectrum. The accessibility-suitability axis captures how obtainable a dataset is and how well it fits the purpose at hand — a timetable feed that is technically public but buried in incompatible legacy formats scores very differently from one exposed through clean, standardized interfaces. The demand-supply axis contrasts what data consumers actually need with what data producers currently deliver, exposing mismatches that can silently degrade information quality. The dynamic-static axis, meanwhile, separates data that change by the second, such as train positions and occupancy levels, from data that remain stable for months or years, such as station geometry and infrastructure layouts. Together, these axes allow any dataset to be located precisely within a conceptual space, making priorities visible at a glance.

The classification expands on categorizations from previous literature while covering the three data families that underpin transportation systems: timetable data, occupancy data, and infrastructure data. Timetable data form the backbone of every journey planner, defining when trains are scheduled to depart, arrive, and connect. Occupancy data describe how full trains actually are — information that has become increasingly valuable as post-pandemic travelers seek quieter carriages and operators seek to manage crowding dynamically. Infrastructure data encompass the physical fabric of the network, from track sections and electrification boundaries to platform lengths and accessibility features. By mapping these families against the three classification axes, the team built a data requirement matrix that visualizes which data are most urgent for a case of passenger information systems and which data require simulations to obtain — a distinction that determines whether an operator can simply query an existing database or must model scenarios computationally before reliable information can be generated.

The technical significance of this matrix approach should not be underestimated. In complex engineering systems, the hardest problems are often not computational but organizational: different stakeholders hold different fragments of the truth, and without a shared classification, nobody can agree on what is missing. The refined model functions as a coordination device, allowing an infrastructure manager in one country, a train operator in another, and a software developer building passenger-facing applications to locate every dataset within the same conceptual space. When a gap is identified — say, dynamic occupancy data that are technically accessible but poorly suited to real-time dissemination — the matrix makes the gap explicit and prioritized, rather than leaving it buried in institutional memory. This is the kind of applied conceptual refinement that rarely makes headlines but quietly determines whether multi-billion-euro digitalization programs deliver on their promises.

The timing of the work is notable. Europe’s rail sector is in the midst of an ambitious digital transformation, with the European Union investing heavily through programs such as Horizon Europe to modernize signaling, traffic management, and traveler-facing services. Passenger information systems sit at the intersection of these efforts, and their performance depends on data flows that cross organizational and national boundaries. A journey planner serving a cross-border trip must reconcile timetable data from multiple national infrastructure managers, occupancy feeds from several train operators, and infrastructure records from yet more sources — each with its own formats, update cycles, and accessibility constraints. The study’s Europe-wide focus reflects exactly this reality, and its classification scheme offers a way to characterize data from any source consistently, regardless of where it originates or who controls it.

For the traveling public, the implications could be tangible. Better-mapped data mean journey planners that know not just when a train is scheduled but how likely it is to be crowded, which carriages have space, and how a delay on one line propagates through the network. They mean accessibility information that is accurate and current, allowing passengers with reduced mobility to plan with confidence. They mean disruption messages that arrive before the frustration does, because the underlying feeds are classified, prioritized, and maintained according to a shared standard. The authors’ data requirement matrix, by visualizing the most urgent data needs, gives operators a practical tool for deciding where to invest scarce digitalization resources first — a question that every railway directorate in Europe is currently asking.

There are also lessons here for the broader data science community. The study demonstrates that classification and mapping — tasks often dismissed as unglamorous cataloging — can be elevated to a rigorous conceptual discipline with direct operational consequences. The three-axis scheme echoes classification frameworks used in other domains, from geographic information systems to supply chain analytics, suggesting a transferable pattern: characterize data by how obtainable it is, how well it fits demand, and how rapidly it changes, and the priorities for improvement become self-evident. For researchers entering the railway domain, the expanded categorization provides a starting taxonomy that covers timetable, occupancy, and infrastructure data in a single coherent scheme, sparing future teams from rebuilding the conceptual foundations from scratch.

The work, published open access under a Creative Commons Attribution 4.0 license, was carried out in accordance with relevant guidelines and regulations, with the authors confirming that the underlying survey used no personal data and that records were anonymous. The authors declare no competing interests. As Europe’s railways race to win back passengers with better digital services, the study’s central message resonates far beyond the sector: the journey to great passenger information begins long before the platform display — in the careful, disciplined mapping of the data that make information possible. In the emerging economy of mobility, the railways that map their data best may be the ones that travelers trust most.

Subject of Research: Engineering

Article Title: Data classification and mapping for supporting passenger information systems in railways

Article References: Mandhaniya, P., Olsson, N. O. E., Urassa, P. E., Lundgren, M., & Bontveit, H. (2026). Data classification and mapping for supporting passenger information systems in railways. Scientific Reports. https://doi.org/10.1038/s41598-026-75445-9

Image Credits: AI Generated

DOI: 10.1038/s41598-026-75445-9

Keywords: railways, passenger information systems, data classification, timetables, train occupancy, railway infrastructure, Data, classification, mapping, supporting, passenger, information