Bringing more data to Uber Movement with street speeds

By Emily Strand and Jordan Gilbertson

We’re excited to announce the release of a new Uber Movement dataset — speeds. In the coming months, we’ll be making street segment level speeds data, a valuable and traditionally costly dataset for cities, free and publicly available. We will share this data using the open standard created by SharedStreets — a nonprofit initiative supported by NACTO and the Open Transport Partnership. Additionally, we’re deepening our commitment to long-term, open standards with a $250,000 grant to SharedStreets to help fund their valuable work.

We’re excited to see how this data will be used to spur new ideas that improve urban mobility.

Average speeds on New York City’s road network visualized in kepler.gl

A commitment to cities

After expanding our existing zone-to-zone travel time data to more than 20 cities around the globe, we’ve seen insightful work produced from understanding the impact of road closures in London to measuring traffic conditions after flooding in Nairobi. As we’ve spoken with our city and industry partners over the past year, it’s become clear that we could take this a step further by providing more fine grained street-level speed data that’s needed for everything from congestion management to road safety analysis.

Visualizing the impact of the 2016 Tower Bridge closure on travel times across London

In making this traditionally difficult to compile and costly dataset freely available, we’re hoping to inspire more and better conversations about the future of our streets. We’re also working to avoid common pitfalls like cost-prohibitive contracts and restrictive data sharing agreements.

Investing in open standards

Our speed data leverages a new standard for describing the street: the SharedStreets referencing system. SharedStreets provides a global, non-proprietary system for describing streets, designed to incorporate any source of street-linked data. This allows public and private entities to communicate with clarity and precision about streets while ensuring full compatibility with organizations’ internal map data and current daily workflows. We’re proud to be working with SharedStreets in creating global standards that will help us reimagine how we produce and share data about public roads.

More about our speeds data

Our speeds data is aggregated and anonymized from billions of Uber trips across the globe.

As always, our first priority is the privacy of our riders and drivers. All data is anonymous and aggregated. In fact, we’re not providing individual trip data at all. Instead, we’ve aggregated data to the street segment level, and are providing speed values averaged from all on-trip traversals across that segment.

We’re also employing the same privacy filtering approach that’s used for our existing travel times data set — we only show data for a given street segment when a minimum number of readings are available to ensure individual rider or driver behavior is anonymous.

This new dataset will allow for a fine grained look at speeds within a city. This data can help support planning for a variety of transportation changes we think are critical to the long term health of cities: bike lanes, new transit routes, carpool lanes, or new dedicated transit infrastructure, just to name a few examples.

We’ll be bringing this new data set to at least 6 cities in the next year. Stay tuned for future announcements about individual pilot cities.

Uber thrives when cities thrive. Today, we bolster our commitment to providing powerful, useful and open tools for cities as we work towards building safer and more sustainable streets, together.

Have a question or research you’re interested in pursuing with this data? Contact us at movement-research@uber.com — we would love to hear from you!