Mobility Data: How Your Travel Patterns Could Inform Policy
Data generated by travel modes can inform planners and regulators in improving the transportation system, but private mobility companies often restrict their access for concerns about privacy and competition.
Information about how you choose to move around your community—your mobilitydata—is a coveted commodity. Did you drive, rideshare, or take the bus? How much did you pay and how long did it take you? Answers to questions like these can help city planners and regulators design transportation systems that best meet your needs.
Unfortunately, planners and regulators have access to only a small subset of the mobility data out there. Most mobility data is collected by private companies, such as app developers, automakers, and ride-hailing and ride-sharing services. These companies are rarely enthusiastic about sharing their data. Planners are also rarely able to gather high-resolution data on personal vehicles, which still represent the vast majority of all travel and traffic. As a result, many officials feel as if they’re increasingly “flying blind” when it comes to making smart transportation decisions.
A recent example of the resistance of mobility companies to data sharing is in the fight to learn how much Uber and Lyft are contributing to congestion. Rather than share their data with public officials or planners to analyze their impact on traffic, the ride-hailing companies jointly hired a consulting firm to analyze their impacts in six cities. While the findings are useful for understanding larger impacts, full data sets are still not available for public sector agencies to review and analyze in more detail for unique jurisdictions. Further, the data used are high-level mileage data, showing contributions to overall vehicle miles, excluding New York City, the largest market for these services.
Policymakers could require companies to share mobility data with the public sector, but the prospect of such a mandate raises real issues. Proprietary concerns loom large. If one company’s data makes its way into the public domain, other companies could seize competitive advantage. Shared data could also compromise customer privacy. Even if data are anonymized, travel patterns can be enough to link trip records to specific individuals. Aggregating mobility data so that individual trip routes aren’t shared can alleviate these security concerns, but too much aggregation makes data less useful for planning. Finally, meeting data-sharing requirements can be burdensome and costly for companies. This could discourage new market entrants and could increase prices if costs are passed onto consumers.
Mobility Data Solutions
A new paper [pdf] from the University of California, Davis presents a framework for how to improve mobility data sharing to respond to the priorities of public officials and private companies alike. The paper argues that successful mobility data sharing involves four steps: 1) establish data standards, 2) develop clear data-sharing requirements, 3) securely store shared data, and 4) apply mobility data to make smart transportation decisions.
The first step is to establish a common language practice for shared data, or data standards. Standards make it possible to compare data from different sources and make data collection and sharing easier. Data standards are already well established in certain parts of the transportation world. For instance, more than a thousand public-transit operators use the General Transit Feed Specification (GTFS) to format and share routes and timetables. This is what enables information for transit all over the world to be accessed by the public through a single integrated platform like Google Maps.
The GTFS recently inspired the city of Los Angeles to launch the Mobility Data Specification (MDS) [pdf] in 2018, a new data standard for scooter-sharing services. MDS is now administered by the Open Mobility Foundation. The MDS allows real-time communication between mobility providers and public agencies. Mobility providers participate in an API that shares real time data on vehicle location, while public agencies input data important to mobility-service operations (such as information on places where scooters and similar mobility devices can park legally). This innovative strategy has quickly become popular—the MDS has only been around for about a year, but is already being used by more than 50 cities. There are concerns about how to ensure secure collection and storage of disaggregated trip data. But there is great potential for further expansion of the MDS to facilitate a standardized exchange of mobility data for cities.
The next step is for cities and public agencies to apply the standards and establish clear data-sharing requirements for mobility companies. To date, most governments do not have mobility data sharing requirements, and there are no ubiquitous norms for data sharing policies. Among the handful of places with data sharing requirements, unsurprisingly they all look different, varying considerably between state agencies versus small cities, because they have such different purviews. State-led efforts to date tend to collect data at relatively infrequent time intervals and aggregated levels (e.g., data reported annually by zip code). City-led efforts may be better suited to collect the fine-grained data needed for certain types of local planning. However, setting consistent requirements at higher levels will make compliance easier for companies operating in multiple jurisdictions.
Data collection creates a need for data storage. Given the scale of mobility data and the costs of secure storage, government is likely to have a role in establishing centralized mobility-data repositories. Centralized repositories will provide users access to data from multiple regions, which is often key to identifying trends. Partnering with a trusted third party such as a university or national laboratory for repository management can help ensure that public transparency laws do not result in accidental exposure of personally identifiable information.
A great example of a centralized data repository is the U.S. Department of Transportation’s Secure Data Commons. This repository offers different levels of access privileges to different types of data users (e.g., researchers, city planners). Users can also perform some statistical analyses within the system, meaning that raw data isn’t downloaded and stored non-securely on local computers or servers.
Finally, mobility data is only useful if data-driven insights are accessible to multiple audiences, including elected officials and the public. A large number of proprietary data analysis platforms exist to address this issue. However, relying on private analytical platforms to store or visualize public data may set a troubling precedent, due to concerns about who maintains long-term access to the data. SharedStreets, an open-source nonprofit platform has also emerged, potentially offering a cheaper and attractive alternative for public agencies and cities.
As mobility options proliferate, policy frameworks to govern mobility data sharing are overdue. Threading the needle between oversharing and undersharing is key to ensuring that the data we collectively generate works to our benefit.