Keeping London's roads moving with reliable data for decision-makers.

“EVERY DECISION THAT WE MAKE ON THE ROAD NETWORK IS HIGHLY SCRUTINISED AND WE NEED TO BE REALLY CONFIDENT IN WHAT THE DATA IS TELLING US, PARTICULARLY WHEN PREDICTING INCIDENTS AND PERFORMANCE. 1SPATIAL HAS DELIVERED A SOLUTION THAT HELPS GROW OUR CONFIDENCE, AND THAT'S WHAT MAKES A SIGNIFICANT IMPACT."

JAYMIE CROUCHER, GIS LEAD, TRANSPORT FOR LONDON

Aligning information from different road networks to predict incidents and improve performance.

Benefits

• Enhanced efficiency and coordination: Streamlines processes, improves data management and quality, and bridges internal gaps, enabling more effective coordination and decision-making across TfL teams.
• Cost and time savings: Saves approximately £15,000 annually in staff time and reduces potential program delays by eliminating the need for extensive manual data reconciliation and migration.
• Improved incident response and predictive modelling: Integrates near real-time traffic information into predictive models, supporting more effective and efficient incident response and highway capacity management.
• Consistent and accurate geographical alignment: Utilises OS MasterMap TOIDs to ensure accurate and consistent geographical alignment across TfL, enhancing reporting and operational efficiency.

Background and TfL's role

As one of the worlds megacities, with over 8 million residents, London faces a constant challenge: keeping its vast transportation network flowing. Transport for London (TfL) manages this intricate system, including a critical component – 590 kilometres of main roads carrying roughly 30% of the city's traffic. However, with ever-growing demand and competition from other transport modes, these roads are under immense pressure.

Every decision TfL makes regarding the road network is subject to intense scrutiny. To address this pressure, TfL has implemented the Surface Intelligence Transport Systems (SITS) program. This program aims to improve traffic flow and increase capacity through several key initiatives. The core concept revolves around ensuring the right data is available at the right time to deliver the right decision, so enabling the TfL’s Network Management Control Centre to identify, respond to, and manage traffic incidents faster and more efficiently.

TfL’s GIS Lead in Network Management and Resilience, Jaymie Croucher, heads the team of three experts responsible for facilitating geospatial data across operations and ensuring it supports decision-making and insights to support the day to day running of services. They are embedded within a 30-strong data analytics and modelling team.

Challenge 

Predicting where and when incidents may take place, as well as their impact on traffic, is a key element of SITS. However, with the modelling data and the incident management system used to deliver this component based on different geographical frameworks, a spatial lookup was needed to align the two networks which are:

• Historical and real-time data: Used by an internal modelling team to forecast potential incidents using the Operational Network Evaluator (ONE) network.
• High-precision road network data: The Common Operational Road Network (CORN) utilises the authoritative and detailed (Ordnance Survey) OS MasterMap Highways Network to provide TfL with a high-quality source of information.

Jaymie Croucher explains:

“The predictive component of SITS uses historic and real time data to infer what is likely to happen in a certain situation. We have an internal modelling team that provides data to do that using the Operational Network Evaluator (ONE) Model network – tactical highway assignment model built in the PTV VISUM software environment.

The Incident Management System uses a geospatial framework called the Common Operational Road Network or CORN. This aggregates OS MasterMap Highways Network – an authoritative and highly detailed node and link dataset – for around 600,000 road links in London to provide TfL with an accessible, high-quality source of information to meet GIS, analysis, modelling and reporting requirements.

With the CORN and ONE Model built using different geospatial networks and not linked identifiers, if you wanted to relate one to the other, you'd have to do it on the fly. This has implications on end performance to the system and can also impact the accuracy of the translation. For the predictive component of SITS to work effectively, we had to address the problem of the misalignment of datasets as there are key geospatial differences between them.”

Solution

Data integration, or conflation, enables the re-use of existing data investments by taking the best from each dataset to create a single source of spatial truth. 1Spatial has significant experience managing projects to create a central and authoritative database for clients across the UK, Europe and USA.

This insight, together with its expertise and reputation for enabling practical applications of OS MasterMap Highways Network, made 1Spatial the natural partner for TfL.

By translating TfL’s requirements into a workable three-step strategy, 1Spatial delivered an automated solution to realign ONE model links with the OS MasterMap Highways Network geometry. As a result, each ONE model link is assigned an OS MasterMap topographic identifier (TOID), which enables linking and sharing of data against the road and street networks.

The methodology and process use the tools available to define a set of rules that fine-tune the matching process to a high degree of confidence. The spatial lookup incorporates linear referencing to determine how far along a TOID the ONE model link start and end point should be. Where a ruleset cannot be applied, it automatically matches ONE model links to nearby features, such as complex junctions.

Furthermore, the workspace can be reused for OS and ONE model updates, as well as to enable future improvements, and with the work taking place offline, there was no downtime of the previous system during development.

The solution demonstrates how 1Spatial solved the issue of conflation (different yet complementary road data from multiple sources) by bringing together conceptual and physical representations of the real world. In doing so, it contributes to TfL’s vision for a unified high-quality geospatial network.

Jaymie Croucher concludes:

“We thought that inevitably one team or the other was going to have to change their way of working to deliver the predictive element of SITS. By bridging the gap between their processes, this solution enabled both our GIS and modelling experts to continue working in their own way – that's a real benefit."

"WORKING WITH 1SPATIAL, AND THE CAPABILITIES AND INSIGHT THEY HAVE BROUGHT TO THE PROJECT, HAS BEEN A REALLY POSITIVE EXPERIENCE.”