A real-world demonstration in the City of York
European cities face increasing challenges in managing traffic congestion, ensuring reliable travel times, and responding rapidly to unexpected disruptions — all while working with fragmented, incomplete, or costly data sources. Within the EMERALDS project, the City of York was used as a real-world demonstration to show how extreme-scale mobility analytics can support both day-to-day traffic operations and strategic transport planning.
This EAD demonstrates how EMERALDS tools can transform raw mobility data — such as GPS traces from vehicles and public transport data — into actionable insights. The result is a dual approach that combines near real-time traffic monitoring with in-depth offline analysis, enabling cities to better understand current conditions, anticipate congestion, and design more resilient transport networks.
The video below illustrates the full demonstrator, from data ingestion to traffic monitoring, analysis, and forecasting.
How EMERALDS addresses the challenge
Urban transport authorities often face a combination of constraints:
- Limited access to real-time, high-quality traffic data
- Partial coverage of probe vehicle data
- Strong privacy requirements that restrict detailed trip analysis
- The need to combine multiple data sources (private vehicles, public transport, maps)
- Tools that are either too complex to deploy or too rigid to adapt to local needs
EMERALDS addresses these challenges by providing a modular set of analytics services, called Emeralds, designed to operate under extreme-scale conditions — large volumes of data, high update rates, and heterogeneous sources.
In the York demonstration, EMERALDS shows how cities can:
- Monitor traffic conditions in near real time, even with limited probe vehicle penetration
- Detect congestion hotspots by analysing traffic patterns across space and time
- Forecast congestion before it happens, enabling proactive traffic management
- Leverage public transport data (GTFS) as an alternative or complement to car-based data
- Analyse historical traffic behaviour to support long-term planning and policy decisions
Rather than replacing existing tools, EMERALDS is designed to extend and enhance them, making advanced analytics accessible to cities through interoperable and scalable components.
How it works
The EMERALDS solution demonstrated in York combines real-time analytics with offline exploratory analysis, using a shared data foundation.
1. Data ingestion from multiple sources
The demonstrator integrates several types of mobility data:
- Raw GPS data from vehicles, received as a continuous real-time stream
- Public transport data (GTFS and GTFS-Realtime) from buses operating in the city
- OpenStreetMap-based road network data, used as the reference map
All data are anonymised and processed in compliance with privacy requirements.
2. Real-time traffic monitoring and forecasting
At the core of the real-time component is the Real-time Extreme-scale Floating Car Data (FCD) Analysis Emerald. This service estimates traffic speeds on road segments directly from raw GPS points, without requiring pre-aggregated speed data.
In York, this capability was integrated into PTV Flows, creating a live demonstrator that allows users to:
- Visualise current traffic conditions on the road network
- Monitor key corridors using configurable performance indicators
- Trigger alerts when travel times exceed defined thresholds
- Replay past traffic situations for retrospective analysis
- Forecast travel times up to one hour ahead
This enables traffic operators to move from reactive monitoring to anticipatory traffic management.
3. Offline analysis for deeper insights
To complement real-time monitoring, several EMERALDS components were applied offline to support exploration and planning.
- Trajectory analysis and travel time analytics transform raw GPS and bus data into meaningful vehicle trajectories, enabling detailed analysis of speeds, stops, and travel behaviour.
- Extreme-scale map matching aligns GPS points to the road network and reconstructs high-quality trajectories, even when data are sparse.
- Hotspot analysis identifies statistically significant congestion areas by considering not only individual road segments but also their spatial and temporal context.
- Highway congestion prediction models learn from historical and real-time patterns to estimate the probability of congestion on specific links.
- Public transport–based traffic estimation uses buses as moving probes to estimate general traffic speeds across the network, reducing reliance on costly car data.
Together, these tools enable cities to understand where, when, and why congestion occurs, and to assess how traffic conditions evolve across the day.
4. From data to decision support
The York use case demonstrates that EMERALDS can deliver value even when:
- Probe vehicle coverage is low
- Data sources are incomplete or heterogeneous
- Cities operate under strict data protection constraints
By combining multiple data streams and analytics services, EMERALDS produces insights that are directly usable by:
- Traffic control centres
- Transport planners
- Policy makers evaluating long-term interventions
Impact and replicability
Although the demonstrator focused on York, the approach is fully transferable to other cities and regions. The EMERALDS architecture allows components to be deployed independently and connected to existing platforms, making it adaptable to different data availability and operational needs.
This use case illustrates how EMERALDS contributes to:
- More resilient and efficient transport networks
- Better-informed mobility decisions
- The transition toward data-driven, sustainable urban mobility