New Data Hub Shows How Waymo Improves Road Safety

We are introducing a new resource that offers deeper insights into how the Waymo Driver is helping make our roads safer.  

The new data hub compares the Waymo Driver’s crash rates to humans’ in the cities where we operate. It leverages best practices in safety impact analysis, invites researchers to replicate the results, and will be regularly updated. This road safety hub provides an unprecedented level of transparency within the autonomous driving industry. 

Waymo’s Positive Impact on Road Safety

Transparency and accountability are essential to achieving our mission to be the world’s most trusted driver. As we continue to scale our operations incrementally and deliberately across Phoenix, San Francisco, Los Angeles, and Austin, the Waymo Driver’s positive impact on road safety becomes increasingly evident. To empower communities with a deeper understanding of our safety record, the new data hub shows how the Waymo Driver’s crash rates compare to humans’. 

The analysis builds on the robust methodology that Waymo first introduced in 2023 and has consistently used to benchmark its safety record. Peer-reviewed* by industry experts, the methodology advances the state of the art of crash rate analysis and enables valid comparisons of autonomous and human driving performance. It adheres to the industry best practices, including the recently introduced RAVE (Retrospective Automated Vehicle Evaluation) checklist for safety evaluation that was jointly developed among industry, academic, and insurance collaborators.

The most recent data shows that over 22 million rider-only miles driven through the end of June, the Waymo Driver has been involved in 84% fewer crashes with airbag deployment, 73% fewer injury-causing crashes, and 48% fewer police-reported crashes compared to human drivers**. Notably, this is the first time the airbag deployment metric has been introduced in our crash rate analysis, providing valuable context regarding the severity of crashes with other vehicles.

Additionally, we provide local metrics for San Francisco and Phoenix, cities where we have driven millions of rider-only miles over time. This allows for statistically significant comparisons with local human benchmarks and demonstrates how the Waymo Driver is improving road safety in each of these cities.

To ensure the transparency and validity of our findings, we’ve made it easy for others to reproduce the results. By leveraging publicly available data — specifically, Waymo’s crash reports submitted under NHTSA’s Standing General Order (SGO) — and disclosing vehicle miles traveled (VMT) data we enable road safety researchers to replicate, and therefore verify, our analysis. 

As we log more miles, we will regularly update the webpage as NHTSA SGO reports are being released.

How the Waymo Driver Reduces Crash Rates

There are a number of factors that contribute to the Waymo Driver’s strong safety record. Because the Waymo Driver is programmed to respect the rules of the road like stop signs and speed limits, and because it never drives drunk, drowsy or distracted, the Waymo Driver prevents many common types of crashes from occurring at all. And thanks to its advanced detection and collision avoidance capabilities, it is continuously looking to respond quickly to potential hazards — like another car about to run a red light — in real time. 

Our comprehensive safety framework is a crucial part of how and why the Waymo Driver improves road safety, providing data supported guidance for our deployments. 

Just as a single musical instrument can’t capture all the richness and complexity of a symphony, so too a single metric can’t capture all the complexities and nuances of safe driving. Waymo’s safety framework includes multiple methodologies across hardware, behavior, and operations that each provide a different, complementary part of the answer, coming together to create a rich harmony that’s far more informative than its individual components.

By introducing the new safety data hub, we aim to provide riders, communities, government officials, and researchers with valuable insights into the impact of Waymo’s driving on road safety. We will also continue to share meaningful information about our safety approach and the complementary methodologies and metrics we employ to evaluate the safety of the Waymo Driver, building on dozens of safety publications we’ve shared over the years.

Safety transparency, like what we are providing in this industry-leading data tool, is key to building public trust as we bring autonomous driving technology to more riders. The U.S. Department of Transportation can help in this effort too, and we urge NHTSA to also collect vehicle miles traveled data when it conducts its planned rulemaking to codify the Standing General Order on Crash Reporting. This will allow safety researchers to do industry-wide analysis along the lines of what our tool enables relating to the Waymo Driver. 

To learn more about Waymo’s approach to safety and real-world performance, visit https://waymo.com/safety.

_{*Scanlon, J. M., Kusano, K. D., Fraade-Blanar, L. A., McMurry, T. L., Chen, Y. H., & Victor, T. (2024). Benchmarks for Retrospective Automated Driving System Crash Rate Analysis Using Police-Reported Crash Data. Traffic Injury Prevention (In Press). doi:10.1080/15389588.2024.2380522.}

_{Kusano, K. D., Scanlon, J. M., Chen, Y. H., McMurry, T. L., Chen, R., Gode, T., & Victor, T. (2024). Comparison of Waymo Rider-only crash data to human benchmarks at 7.1 million miles. Traffic Injury Prevention (In Press). doi:10.1080/15389588.2024.2380786.}

_{** The analysis includes all Waymo crashes, regardless of the Waymo vehicle’s role in the crash.}