Making Sense of Mobility Data

Mobility data is the connective tissue of modern transport. Every bus schedule published in GTFS format, every e-scooter pinging its location via GBFS, every real-time arrival prediction powered by SIRI — all are expressions of a data ecosystem that now underpins how billions of trips are planned, operated, regulated, and optimized worldwide. For anyone building, governing, or investing in transport systems, understanding this ecosystem is no longer optional. It is foundational.

This primer maps the full landscape: the standards that structure mobility data, the stakeholders who produce and consume it, the regulations that mandate its sharing, and the strategic tensions that shape its future. The goal is not just comprehension but decision-readiness — whether you are a transport authority evaluating data infrastructure, an operator navigating compliance, or a platform developer choosing which standards to support.

Defining mobility data and the value chain that makes it useful

"Mobility data" is broader than most people assume. It encompasses all digitally captured information about how people, vehicles, and goods move through space and time. This includes the obvious — transit schedules, real-time vehicle positions, ridership counts — but extends to parking availability, curb usage, fare structures, emissions per vehicle-kilometer, accessibility of stations, demand patterns from ride-hailing, and aggregated movement flows derived from telecommunications signals.

A useful way to categorize this diversity is along three axes. Temporal: static data (routes, stops, timetable structures that change infrequently) versus dynamic data (real-time positions, delays, and availability that update every few seconds). Access: open data published freely for anyone to use, versus proprietary data held by operators as competitive intelligence, versus regulated data that authorities can mandate sharing of under specific conditions. Purpose: traveler-facing data designed for journey planning apps, regulator-facing data used for compliance monitoring and urban planning, and internal operational data that never leaves an organization's systems.

The data value chain runs from collection (transit agencies' CAD/AVL systems, IoT devices on shared vehicles, GPS traces from navigation apps) through standardization (formatting raw data into specifications like GTFS or NeTEx) to aggregation (combining feeds from dozens of operators into a single platform like Entur or Trafiklab) to distribution (APIs, bulk downloads, real-time streams) and finally application (Google Maps routing a commuter, a city planner analyzing service gaps, a MaaS app selling multimodal tickets). Bottlenecks exist at every stage, but the most persistent is standardization — getting diverse operators to produce consistent, high-quality data in formats that machines can reliably consume.

A common misconception deserves early correction: "just publish GTFS and you're done" profoundly understates the challenge. A GTFS feed is a starting point, not a finish line. Without ongoing quality management, real-time supplements, fare integration, and governance frameworks, published data quickly becomes stale, incomplete, or misleading.

Five standards that define the technical landscape

The mobility data standards ecosystem can feel bewildering, but five specifications account for most of what matters in practice. Understanding what each does — and where they compete or complement — is essential.

GTFS (General Transit Feed Specification) is the global workhorse. Born in 2005 from a collaboration between Google and Portland's TriMet transit agency, GTFS is elegantly simple: a ZIP file containing CSV text files describing stops, routes, trips, schedules, and optionally shapes, fares, and pathways. Over 10,000 transit agencies in 100+ countries now publish GTFS feeds. Its companion, GTFS Realtime, uses Protocol Buffers to deliver trip updates, vehicle positions, and service alerts, typically refreshed every 15–30 seconds. Recent extensions include GTFS-Flex for demand-responsive transport (officially adopted March 2024) and GTFS-Pathways for modeling station accessibility. GTFS's strength is its simplicity and massive ecosystem of tools and consumers. Its limitation is scope — it was designed for rider-facing trip planning, not operational management or complex multimodal fares.

NeTEx (Network Timetable Exchange) is GTFS's European counterpart, but far more ambitious. Published as a CEN standard (CEN/TS 16614), NeTEx is an XML-based format built on the Transmodel conceptual reference model. Where GTFS has roughly 20 simple files, NeTEx contains thousands of XSD elements organized across six parts covering network topology, timetables, fares, European passenger information profiles, alternative modes (sharing, rental, taxis), and accessibility. NeTEx can represent virtually anything in public transport — but this comprehensiveness comes at a cost. Implementation requires significant expertise, and national profiles (Norwegian, Italian, French) add further complexity. The EU mandated NeTEx as the required format for National Access Points under MMTIS regulation, making it the de jure European standard even as GTFS remains the de facto global one.

SIRI (Service Interface for Real-time Information) complements NeTEx for dynamic data. Also a CEN standard, SIRI provides eight specialized services — from stop monitoring and vehicle tracking to connection protection and disruption messaging — delivered via request/response or publish/subscribe patterns. SIRI dominates European real-time infrastructure, while GTFS Realtime dominates elsewhere. A new Part 7 published in 2025 establishes an EU-wide real-time interoperability profile.

GBFS (General Bikeshare Feed Specification) handles shared mobility — bikeshare, e-scooters, mopeds, shared cars. Created in 2014, now at version 3.0, GBFS provides JSON endpoints describing vehicle and station availability, pricing, and geofencing zones. It is deliberately public and unauthenticated by design, optimized for consumer-facing apps. Over 230 operators in 45+ countries publish GBFS feeds.

MDS (Mobility Data Specification) serves the opposite audience: city regulators. Originally created by Los Angeles DOT in 2018 to manage the sudden arrival of dockless scooters, MDS (now managed by the Open Mobility Foundation) provides APIs through which cities receive trip data, vehicle status, and fleet information from operators, while sharing digital policies like geofences and speed limits. Over 160 cities use MDS, but it remains controversial — Uber sued LADOT over privacy concerns, and civil liberties organizations have raised constitutional questions about real-time location tracking. MDS 2.0 expanded scope to passenger services, car-share, and delivery robots.

The divergence between European and North American standards reflects deeper philosophical differences. Europe's approach is top-down: formal CEN standards processes, regulatory mandates, and comprehensive specifications designed for cross-border, multi-operator interoperability. North America's approach is bottom-up: community-driven formats, voluntary adoption, and deliberate simplicity. Both have trade-offs. NeTEx is richer but harder to implement; GTFS is simpler but less expressive. In practice, many European agencies publish in both formats — NeTEx for regulatory compliance, GTFS for developer adoption.

Conversion between the two is possible but imperfect. NeTEx's richer semantics mean information is lost when converting to GTFS, while GTFS-to-NeTEx conversion produces structurally valid but semantically shallow outputs. Projects like DATA4PT maintain detailed mappings, but the dual-standard reality persists.

Who produces the data, who needs it, and who captures the value

The mobility data ecosystem involves a complex web of stakeholders with misaligned incentives.

Producers span the entire transport sector. Public transit operators generate schedules, real-time positions, and ridership data. Micromobility companies (Lime, Tier, Voi) produce vehicle availability, trip records, and fleet telemetry. Navigation platforms — particularly Google Maps — aggregate probe data from billions of devices to generate real-time traffic estimates. Connected vehicles now generate roughly 50 GB of data per day internally, with OEMs like Stellantis launching dedicated data marketplaces. Telecommunications companies sell anonymized movement data derived from cell tower signaling — Telia's "Crowd Insights" product in the Nordics is a prominent example, used by Sweden's Public Health Agency during COVID-19 to track population mobility changes.

Aggregators occupy a critical middle layer. Public aggregators include Entur (Norway), Trafiklab (Sweden), Rejseplanen (Denmark), and Digitransit (Finland). Private aggregators include Google Maps, Apple Maps, Transit App, and Citymapper. The EU mandates National Access Points (NAPs) as official aggregation portals — over 30 now operate across Europe, coordinated through the NAPCORE initiative.

Consumers range from journey planning apps and MaaS platforms to urban planners, researchers, and the traveling public.

The power dynamics are striking. Google Maps has become the world's dominant trip planning interface by consuming GTFS data from thousands of agencies for free, integrating it into an advertising-supported platform, and creating dependencies that no regulation currently addresses. Academic literature describes this as "quasi-sovereign power" in shaping mobility choices. Transit agencies provide their data voluntarily because Google visibility drives ridership — but this creates a structural data asymmetry where agencies contribute enormous value and capture relatively little. When you record your parking location in Google Maps, you get a reminder; Google gets a global parking dataset.

The tension between data as competitive advantage and data as public infrastructure runs through every stakeholder relationship. Ride-hailing companies hold some of the richest urban mobility datasets but fiercely resist sharing them. Micromobility operators share data because cities require it for operating permits — making data a tool of regulatory leverage. Transit agencies publish openly because openness drives the ecosystem. The emerging consensus, articulated by the ITF/OECD, is to "think of mobility data as infrastructure" — linking sharing requirements to licensing, establishing minimum mandates, and building public analytical capacity.

Europe's regulatory architecture and the Nordic laboratory

The regulatory landscape for mobility data in Europe is shaped by several interlocking instruments. The most important for transport practitioners is MMTIS — Commission Delegated Regulation (EU) 2017/1926, significantly amended by Regulation 2024/490 in March 2024. MMTIS requires transport authorities, operators, and service providers to make specified data types accessible through National Access Points in standardized formats (NeTEx, SIRI). The 2024 revision expanded scope from scheduled public transport to include taxis, demand-responsive transport, car-sharing, micromobility, and parking, with a 1 December 2025 deadline for dynamic data types.

The practical challenges of MMTIS compliance are substantial. Research conducted for Entur's knowledge base reveals that most mobility operators outside traditional public transport have never heard of MMTIS. The regulation is written in "train language" — concepts like service disruptions and replacement services don't translate to on-demand or sharing services. One taxi representative reportedly "read it multiple times without understanding what's expected." Operators question why they should share data that they consider commercially sensitive, particularly historical demand patterns and pricing structures. And the mandated EU standards (NeTEx/SIRI) were developed for scheduled transit, creating friction when applied to bottom-up industry standards like GBFS and MDS.

The Open Data Directive (EU 2019/1024) complements MMTIS by requiring public sector data to be open by default, with the High-Value Datasets regulation (applicable since June 2024) specifically designating mobility transport network data as requiring free, machine-readable, API-accessible publication under open licenses.

GDPR adds a critical constraint. Mobility data is often personal data — research by de Montjoye et al. demonstrated that 95% of individuals in a 1.5-million-person mobility dataset could be re-identified from just four spatiotemporal data points. Vehicle location data, trip records, and even vehicle IDs linked to operator databases qualify as personal data requiring full GDPR compliance. Denmark's data protection authority fined taxi company Taxa 4x35 in 2018 for retaining 9 million trip records with insufficient anonymization. The result is a persistent tension: MMTIS mandates data sharing while GDPR restricts personal data processing.

The four Nordic countries represent a fascinating laboratory of different approaches. Norway built Entur — a state-owned company with approximately 300 employees that operates the national journey planner, data registry, and ticketing system. Entur went "NeTEx-native," aggregating data from roughly 60 operators covering 75,000 daily departures. All data is open and free. Built on OpenTripPlanner (open source) and Google Cloud, Entur has become effectively the shared digital infrastructure for Norwegian public transport. Sweden took a federated approach through Samtrafiken, an industry consortium of 38 operators that runs Trafiklab as an open data hub. Sweden used GTFS as its primary format (adding NeTEx more recently) and preserved regional autonomy — each region maintains its own apps atop shared data infrastructure. Denmark's Rejseplanen operates with a remarkably lean team of about 7 people, using the proprietary HAFAS journey planning engine rather than open-source tools. Finland went furthest on legislation, passing the Transport Code (liikennepalvelulaki) in 2017 — the world's most comprehensive mobility data sharing law, requiring both public and private operators to open data and ticketing APIs. Finland's approach explicitly aimed to enable MaaS, and the country's Digitransit platform (open source, GTFS-based) has been adopted internationally from Estonia to Romania.

Building infrastructure that works — and recognizing what fails

The Nordic cases, combined with international examples, reveal clear patterns about what makes mobility data infrastructure succeed or fail.

Transport for London's open data program provides the most compelling ROI evidence. At an annual operating cost of approximately £1 million, TfL's 80+ data feeds support 13,000 registered developers, 600 apps, and generate an estimated £130 million per year in economic benefits according to a 2017 Deloitte study — a 130:1 return. The program spawned companies like Citymapper and proved that open data creates ecosystems no single organization could build alone. Critically, the initial decision to open data was a "leap of faith" — the benefits could not be proven in advance.

The minimum viable data infrastructure for a city or region requires five elements: valid GTFS feed production from operators, a single aggregation point for accessing all operator data, open licensing (CC-BY or CC0), a journey planning engine (OpenTripPlanner is free and proven), and a developer portal with documentation. Finland's Digitransit provides the closest thing to a plug-and-play model — Docker-containerized, GTFS-based, deployable with local data and OpenStreetMap.

Failure modes are equally instructive. MaaS Global's Whim — the world's first commercial MaaS subscription app, launched in Helsinki in 2017 — raised €149 million from investors including BP Ventures, Toyota, and Mitsubishi before filing for bankruptcy in March 2024. Its 2022 financials showed a €9.3 million loss on €3.8 million turnover. The lesson: B2C subscription MaaS is economically unviable without massive scale, and the data platform layer must be separated from the commercial service layer. Cross-border integration remains largely unsolved despite Nordic cooperation — no unified Nordic journey planner exists. And version fragmentation in standards like MDS (where multiple live versions coexist) undermines the very standardization benefits the specifications aim to deliver.

Success factors crystallize around several principles: clear political mandate and ownership; open standards combined with open source to avoid vendor lock-in; starting with open data before building commercial services; separating neutral infrastructure from competitive applications; incremental development with fast initial MVPs (Entur delivered its first product in six months); active developer community engagement; and continuous investment in data quality governance. Perhaps most importantly, successful initiatives treat data as permanent infrastructure requiring permanent maintenance — not as projects with start and end dates.

The road ahead: federation, AI, and unresolved questions

Several trajectories are reshaping mobility data. Real-time coverage is expanding but remains unevenly distributed — many agencies still lack real-time feeds, and the December 2025 MMTIS deadline for dynamic data will test whether regulation can accelerate adoption. New data types are emerging: emissions per trip, station-level accessibility audits, curb occupancy, and demand-responsive transport specifications. The automotive data monetization market, valued at roughly $23 billion in 2024, is projected to grow to over $100 billion by 2034 as connected vehicles become universal data platforms.

The European Mobility Data Space (EMDS), backed by approximately €25.5 million in EU funding, represents the most ambitious attempt to create federated data infrastructure. Explicitly "not a centralized database," EMDS aims to interlink existing transport data ecosystems across Europe through shared governance and technical standards. The deployEMDS project (2023–2026) is piloting use cases in nine cities including Stockholm, Tampere, Barcelona, and Milan. Whether federated architectures can achieve the interoperability that centralized platforms deliver remains an open question.

AI is simultaneously driving demand for richer data and creating new analytical capabilities. Google's Mobility AI program uses aggregated Maps data to build digital twins of cities. Predictive maintenance systems process hundreds of millions of telematics data points daily. AI-powered demand prediction is enabling same-day paratransit booking where 24-hour advance notice was previously required. But AI's hunger for high-resolution, multi-modal, historically deep datasets intensifies every existing tension around data sharing, privacy, and commercial control.

Several governance questions remain genuinely unresolved. Who owns vehicle-generated data — the driver, the OEM, or the insurer? How should society prevent platform monopolization when Google Maps shapes billions of mobility decisions? When AI systems make routing recommendations that influence public transport ridership, who audits the algorithms? And who pays for the ongoing maintenance of open standards, National Access Points, and data spaces that the entire ecosystem depends on? These are not technical questions. They are political ones, and their resolution will determine whether mobility data serves as genuine public infrastructure or becomes another domain where private platforms capture most of the value.

Conclusion

Mobility data has matured from a niche technical concern into a strategic asset that shapes how transport systems are planned, operated, and experienced. The standards landscape — anchored by GTFS globally and NeTEx in Europe — is rich but fragmented, with interoperability remaining an ongoing challenge rather than a solved problem. Nordic countries offer genuinely different implementation models (Norway's centralized Entur, Sweden's federated Trafiklab, Finland's legislation-first approach) that provide actionable lessons for any region building data infrastructure. The most important insight from a decade of experience is structural: mobility data succeeds as open infrastructure and fails as closed competitive advantage. TfL's 130:1 ROI on open data and MaaS Global's €149 million path to bankruptcy tell the same story from opposite directions. The coming years will test whether Europe's ambitious regulatory and federated data space architecture can deliver on its promise — or whether Google, OEMs, and private platforms will continue to set the terms.