Shooting for Utopia: The Future of Railway Data Management

Railways have always been about connections — connecting places, connecting people, connecting economies. Yet behind every train movement, every passenger journey, every bridge and every station lies something less glamorous but far more critical: data.

Data is now the true connective tissue of the railway. It tells us how passengers move, how assets behave, how projects succeed (or fail), and how designs become reality. But here’s the question: what should we be shooting for? What does “good” data management in rail actually look like when we take a hard look across customer experience, asset management, project delivery, and project design?

Let’s dive deep into each of these domains, explore where the industry stands today, and sketch out the “utopia” we should all aspire to.

1. Customer Experience: From Tap-In to Seat Finder

Railways are ultimately judged by customers on one thing: the experience of travel. Delays, overcrowding, confusing disruption messages — all of these are data problems as much as operational ones.

The data we need:

• Demand and flow: Contactless transactions, tap-in/out data, Wi-Fi presence (de-identified), and journey times.

• Crowding and seat availability: Weight sensors, door counters, seat sensors, and reservations.

• Service quality: Punctuality is measured at the journey level, missed connections, and cancellations.

• Feedback and incidents: Complaints, NPS, CSAT, mapped back to service and geography.

Why this matters:

• Imagine boarding a train knowing in advance which carriages are quietest.

• Imagine receiving a disruption alert that doesn’t just say “delayed” but offers your best alternative based on origin, destination, and preferences.

• Imagine performance reporting that isn’t about “PPM at 3 minutes” but about actual journey time metrics that reflect the real passenger experience.

Real-world signals:

• TfL’s Oyster and Contactless data has long been a foundation for customer insights — recently expanded with de-identified Wi-Fi analytics across the Underground to map flows and improve disruption response.

• Southeastern’s SeatFinder shows carriage-level crowding in near-real time, leading to higher app engagement and smarter boarding decisions.

• LNER’s seat sensors distinguish between reserved and genuinely occupied seats, a step-change for long-distance travel.

• Elizabeth Line’s Journey Time Metric is a world-class example of customer-centred performance reporting.

KPIs to aim for:

• % of trains with live crowding/seat data

• % of passengers receiving personalised disruption comms

• Journey Time Metric variance against the target

• Complaint rate per million journeys

2. Asset Management: Knowing, Predicting, Preventing

Behind the scenes, the health of tracks, trains, and stations underpins the reliability of service. Asset management is the invisible engine of rail performance.

The data we need:

• Asset register and hierarchy: Standardised taxonomy, linked to design data and linear-referenced locations.

• Condition and state: From inspections, telematics, sensors, and event logs.

• Work management: Work orders, failure codes, MTBF, MTTR, costs, and spares.

What good looks like:

• ISO 55001-aligned systems.

• Condition-based maintenance is the default, predictive when cost-effective.

• Risk-based prioritisation of renewals.

• Seamless project-to-asset handover with rich Asset Information Models (AIMs).

Real-world signals:

• Network Rail ORBIS (£335m) created a national asset information backbone, enabling mobile inspections and better analytics.

• Plain Line Pattern Recognition (PLPR) now replaces thousands of manual track inspections with AI-powered high-speed imaging.

• SNCF with IBM Watson IoT links trains, tracks, and stations for predictive maintenance at scale.

• Alstom HealthHub aggregates train conditions every 30 seconds for proactive fleet management.

• SMRT Trains (Singapore) found tangible organisational benefits from ISO 55001 adoption.

KPIs to aim for:

• Unplanned failures per million km

• Secondary delay minutes due to asset failure

• % of assets with live condition data

• Reduction in manual inspection hours

3. Project Management: The Digital Thread of Delivery

Rail projects are notorious for delays, overruns, and messy handovers. A significant part of the problem is that data — fragmented, duplicated, and often lost between delivery and operations.

The data we need:

• CDE content: Drawings, models, issues, RFIs, safety records, cost and schedule data.

• Decision logs and approvals: To preserve intent and reduce rework.

• Structured handover packages aligned to asset registers and AIMs.

What good looks like:

• One Common Data Environment (CDE) governed by ISO 19650.

• 4D/5D BIM integrated with cost and schedule risk.

• Automated EIR→AIM handovers, not manual document dumps.

Real-world signals:

• Crossrail’s Bentley CDE hosted millions of documents across thousands of users, with structured handover into TfL’s ops systems.

• HS2’s digital blueprint has formalised model-based workflows and will provide a valuable legacy for future projects.

• Elizabeth Line’s performance packs now tie project-delivery data into operational performance reporting.

KPIs to aim for:

• % deliverables approved first time

• Handover completeness score (data vs. documents)

• RFI cycle time

• Schedule/cost variance at P50 and P80 confidence

4. Project Design: Building for Reuse, Not Just Delivery

Design data is the raw material for every stage of a railway’s life cycle. But too often, it’s locked in proprietary formats and never reused beyond construction.

The data we need:

• Authoritative alignment and chainage models.

• Discipline models (track, OLE, signalling, power, civils) with shared IDs.

• Requirements traceability and validation logs.

Standards that unlock value:

• IFC 4.3: Adds alignments and linear infrastructure into openBIM.

• RailTopoModel and railML: Define network topology and enable timetable–infrastructure exchange.

• EULYNX: Standardises signalling interfaces across Europe.

Real-world signals:

• IFC 4.3 is now production-ready and being mandated for linear infrastructure by progressive clients.

• European programs like Digital Rail Germany and EULYNX are showing the power of shared signalling standards.

KPIs to aim for:

• % models IFC 4.3-compliant

• Requirements verification rate

• % of design data reused in operations

• Defects caught in the model vs. on-site

Pitfalls to Avoid

• Perfect data paralysis: chase “fit for purpose,” not perfection.

• The project–operations gap: appoint an Asset Information Manager on every project.

• Vendor lock-in: mandate open standards (IFC 4.3, railML) in contracts.

• Privacy missteps: follow TfL’s transparent, de-identified approach.

Roadmap to Utopia

First 90 days:

• Define core IDs and the location model.

• Stand up a lakehouse with lineage and catalogue.

• Ingest timetable, movements, and work order data.

• Pilot crowding info and condition-to-work-order integration.

3–12 months:

• Scale crowding and disruption of communications line-wide.

• Roll out PLPR-style analytics for track.

• Implement IFC 4.3 workflows on new projects.

12–24 months:

• Deploy predictive maintenance on the top 5 failure modes.

• Integrate worksite/possession data into the digital twin.

24+ months:

• Push towards ETCS/ATO-ready timetable data.

• Adopt EULYNX signalling interfaces to break vendor lock-in.

The Utopia

The vision is clear:

A standards-based, privacy-safe railway data mesh feeding a living digital twin. Every decision — from passenger disruption messages, to tamping schedules, to resignalling — is based on the same trusted, linearly referenced data. Project models hand over seamlessly into the asset register. Customer KPIs, delay minutes, and maintenance risks are visible in one pane of glass.

This is not fantasy. The building blocks already exist. TfL, Crossrail, Network Rail, SNCF, LNER, and others are already demonstrating parts of it. The challenge now is to stitch these examples into a system-wide fabric — and finally give the railway the digital nervous system it deserves.

👉 What do you think? Where do you see the biggest gaps — customer data, asset health, project handover, or design interoperability?