In October, Waymo, a self-driving car company owned by Google’s parent company Alphabet, released its latest safety report from its autonomous ride-hailing service. The data is one of the most extensive public views into self-driving vehicle safety to date, claiming a 91% reduction in serious injury crashes compared to human drivers, alongside fewer airbag deployments, fewer crashes with pedestrian injuries, and zero fatalities. Although the data is self-reported, if autonomous vehicles are truly as safe as Waymo suggests, it would be a major leap forward for safety in a country where road fatalities are a leading cause of death.
Yet despite the potential public health benefits, polls consistently show public trust in autonomous vehicles (AVs) to be remarkably low. A recent survey found just 13% of U.S. drivers said they would trust riding in a self-driving vehicle, while 61% would be afraid to do so. Such fears have given rise to a wave of backlash against the technology, with fierce opposition to the authorization of driverless ride-hailing services in cities and states across the country, along with proposed federal legislation from Senator Josh Hawley that would effectively ban driverless cars nationwide.
Part of what fuels these fears is a lack of easily accessible, comparable, and independent data about AV safety. While the public can access some limited information about AV testing, it is fragmented across federal, state, and local lines, and cannot be directly compared because of differing reporting requirements and platforms. In the absence of high-quality data, eye-catching headlines and anecdotes about negative individual experiences with AVs dominate discourse.
Undoubtedly, AVs raise real questions around liability, jobs, cybersecurity, and ethics. But they also offer immense promise to reduce crashes, improve independence for people underserved by public transit, decongest urban roads, and lower transportation costs. High-quality data should be the foundation of discussion about these complex topics, not anecdotal speculation.
The solution is a unified federal AV reporting standard. We propose a two-layer approach, designed to build public trust and give regulators the details they need to properly oversee safety. First, a straightforward, public-facing dashboard allowing users to view crash rates and compare the safety of AVs directly to human-driven vehicles under various conditions. Second, a granular, comprehensive database accessible to researchers and regulators, giving them the detail needed to shape regulation. Built to preserve flexibility and privacy while limiting reporting burden, this approach focuses on public accessibility while protecting proprietary information and continuing to foster innovation.
