Open Data Transition Report: An Action Plan for the Next Administration

• • The Department of Transportation National Highway Traffic Safety Administration (NHTSA) should support industry efforts to share automated vehicle data that could lead to safer vehicle development, such as testing data or challenging scenarios that have led to crashes or near misses.
• • In releasing templates and additional guidance for the Safety Assessments described in the Federal Automated Vehicles Policy over the next several months, the NHTSA should ensure these assessments are born digital. Rather than collecting PDFs or documents, NHTSA should collect key data fields in a format that enables automated analysis.
• • As the technology moves toward large-scale deployment, the NHTSA should partner with industry and the research community to work toward developing an open set of criteria for an “automated driver’s test” that will eventually enable automated vehicles to self-qualify for operating on public roads.

In early 2016, Secretary of Transportation Anthony Foxx declared that “we are on the cusp of a new era in automotive technology” as fully automated vehicles transform from a distant vision to a near-term reality.  At least 33 companies are researching and/or developing automated vehicle technologies,  with some claiming they will deploy automated fleets broadly in the next three to five years.  A small number of automated vehicles are already on the road (with a human at the wheel to seize control if necessary) in California, Texas, Washington, Arizona,  Pennsylvania,  and several large urban areas abroad. Automated vehicles can usher in a new era of transportation efficiency and safety while providing mobility to millions of additional Americans. Open data can help ensure that both the technology and the public are ready for the road ahead.

The Department of Transportation is already preparing for this shift. The National Highway Traffic Safety Administration (NHTSA) released the Federal Automated Vehicles Policy,  which provides safety guidance for the development of automated vehicles, in September 2016.  The policy offers vehicle performance guidance that gives industry a flexible approach to pursue the technology, summarizes the need for a consistent national framework for operating automated vehicles, outlines the current regulatory tools available to the NHTSA, and examines potential future tools and authorities. NHTSA is continuing to solicit feedback on the policy and plans to release additional guidance or updates in the future.

As NHTSA continues its research to examine the unique opportunities and risks provided by automated vehicles, such as cybersecurity and performance metric development, the administration should lay the foundation for a robust open data ecosystem to support the automated vehicle industry in the future. They can support this vision by:

• Fostering data sharing across the autonomous vehicle industry and government. The NHTSA Policy outlines a plan to establish a mechanism to “facilitate autonomous data sharing,” which is a critical next step to support safer vehicle development. The shared data should include testing environments and scenarios as well as data on accidents or near-misses, which would help the field of developers to address known weaknesses in the technology. Several industry leaders are already voluntarily sharing information with the public, such as monthly reports, that include much of this information but not the datasets themselves.
• Ensuring automated vehicle system safety assessments are born digital. The NHTSA Policy already requests that manufacturers provide 15-point safety assessments as well as collect and share data on incidents, crashes, and near-misses. NHTSA plans to release templates for these assessments in the near future and eventually make them mandatory.  NHTSA should ensure these templates collect data in a structured and reusable format that is readily available for analysis and potential future sharing in machine-readable formats. Rather than collecting information in PDF or separate documents, NHTSA should embrace best practices in electronic data collection and replace forms and reports with data fields.

As the technology continues to develop, NHTSA should work with industry to develop a standard automated driver’s test based on open data that will enable future vehicles to self-qualify for market entry. Developing this standard in an open format will support a level playing field for industry, enable the research community to participate, and ensure the public has a voice in this epic shift in modern transportation. This role of public participation aligns with the Department of Transportation’s view that “larger questions [concerning automated vehicles] will require longer and more thorough dialogue with government, industry, academia and, most importantly, the public.”

NHTSA should then work with companies to open their autonomous driver’s test data, including videos and quantitative data. It would function similarly to open crash data from the Insurance Institute of Highway Safety.  Opening this data would simultaneously make manufacturers accountable for ensuring vehicle safety, and would give the public confidence in the vehicles. Several surveys show a majority of Americans have safety concerns with automated vehicles.  By developing an automated driver test and publishing the results, NHTSA and the automotive industry could help make the public more receptive to automated vehicles’ safety benefits.