Traffic Data supporting HERE and TomTom maps using OpenLR
Traffic powered by 500 million MAUs updating every 5 minutes for 2.3 billion road segments around the world
By: Ben Levin
New Mapbox blog posts are now published on www.mapbox.com/blog. Leave your email to receive email updates when new posts are published or subscribe to the new RSS feed.
We are now offering stand-alone Traffic Data for use outside of our Navigation APIs — for companies and governments that require the most comprehensive and accurate traffic data to use in customized routing engines, direct analysis, or congestion analysis.
The key to providing accurate traffic data with broad geographic road coverage is to get data from as many people moving on the road as possible. Our Traffic Data provides real-time and predictive traffic patterns drawn from a globally distributed user-base of more than 500M MAUs. Only a handful of companies around the world have access to this volume of data, and no one has made it commercially available. Until now.
Our Traffic Data is now compatible with other proprietary data sets; Mapbox Traffic Data uses OpenLR encoding for maps like HERE or TomTom. From automakers, to large-scale logistics, fleets operations, or urban planning use cases — improve your intelligence with the highest quality commercially-available traffic data via the Public Beta of Mapbox Traffic Data.
“With approximately fifty thousand apps currently using Mapbox tools, the wave of adoption has been considerable. Additionally, the ability to process over 300 million miles of road data per day is unprecedented. Mapbox Data Service’s strategic move to sell live traffic data is unique to the industry and will be particularly well suited for urban planning, route planning at scale, and other use cases that require the analysis of raw traffic data or the simulation of billions of potential routes. Importantly, Mapbox’s offering, and associated pricing, is simple, transparent, and scalable.” — Brent Iadarola, VP, Frost & Sullivan
Highest quality traffic data
Mapbox Traffic Data leverages the power of global crowds at a scale available to only a handful of companies on the planet, paired with advanced data processing and machine learning algorithms to offer accurate live and typical traffic profiles. We produce updates at a 5-minute granularity for 2.3 billion roads segments around the world.
The power of the crowd: A massive, distributed global user base
They key to Mapbox Traffic Data is our wide adoption and data availability. Every month, more than 500 million monthly active users touch our maps, via 45,000 different Android and iOS applications built with Mapbox — generating more than 300 million miles of aggregated, anonymized and privacy-filtered driving data every day. The breadth of this data pipeline allows us to offer more detailed traffic coverage, on more streets than existing datasets.
Accuracy powered by AI
We continually benchmark our traffic predictions against ground truth calculations and use machine learning techniques to continually tune predictions and minimize error. When integrated with a routing engine, Mapbox Traffic Data can achieve predicted ETAs within 10% of actual arrival time (for instance, on a 20-minute trip the ETA would predict 18–22 minutes.)
Real-time coverage
Mapbox receives more than 100,000 location updates per second, and uses data to continuously detect unexpected traffic and slow-downs. Live data in Mapbox Traffic Data includes traffic profile updates every 5 minutes, based on information not more than 15 minutes old.
Using Traffic Data
We’ve built Mapbox Traffic Data to directly plug into common applications and unlock value for on-demand services, logistics companies, and analytics:
Matching to a HERE or TomTom map
Mapbox Traffic Data contains map-neutral OpenLR linear referencing. OpenLR is an open, vendor-neutral standard that uses metadata about the road such as start point, end point, length, and bearing to allow cross-map referencing. With OpenLR, Mapbox Traffic Data integrates directly into a HERE or TomTom basemap. We have also added support for native matching to OpenStreetMap node-pair IDs for companies using OSM node IDs with their custom data.
Integration with a custom routing engine
Some sophisticated routing and analytics users want to perform custom routing analysis and integrate proprietary datasets, such as driver-specific speeds or historical trip information, into their routing engines. Blending this proprietary information with Mapbox Traffic Data can improve accuracy, leading to more accurate ETAs and more deliveries or pickups per hour.
Mapbox Traffic Data is designed to integrate with popular routing engines like Graphhopper and OSRM. For a live traffic use case, take map-matched typical traffic data, overlay live data where available, and upload it as edge weights into the routing graph. An example of OSRM’s implementation for this functionality can be found here.
Once you have a custom routing engine established, you can easily scale to millions or billions of requests, continually leveraging improving Traffic Data, without incurring additional costs.
Directly analyzing Traffic Data
Mapbox Traffic Data is accurate, comprehensive (coverage of all major roads and many small roads), and detailed (5-minute traffic increments; block-level traffic estimates). With Mapbox Traffic Data, you can directly analyze typical traffic data to discover patterns, model changes, and determine typical congestion. Our documentation has an example of working directly with traffic metrics: Calculate Aggregated Traffic Metrics
Test immediately and start building
Check out our FAQ for more details and detailed documentation for traffic data to begin integration testing today. For production or evaluation access, Mapbox Traffic Data is licensed annually and available on an Enterprise plan. Contact our sales team for more information.
