Hundreds of driverless cars are now an everyday part of traffic in San Francisco as they shuttle passengers across town, but how safe are they and how do they compare to human drivers? Waymo, owned by Google’s parent company Alphabet, just released a new study that attempts to shed light on its own track record as one of San Francisco’s largest driverless car companies. The findings, described as first-of-its-kind research, make the case that Waymo’s driverless fleet is able to steer clear of certain types of collisions at a higher rate than human drivers.

“It shows that we’re safe,” said Trent Victor, Waymo Director of Safety Research and Best Practices. “We’re seeing the fruits of our focus on safety.”

Waymo, in partnership with insurance giant Swiss Re, analyzed insurance liability claims in an effort to collect collision data involving both driverless vehicles and human-driven cars. The study utilized 600,000 claims stretching over 125 billion miles to create a “baseline” for human driver collisions that could be compared against Waymo’s record involving 3.8 million miles traveled without a human in the driver’s seat.

Attempting to draw fair comparisons between the safety record of driverless cars versus human drivers is a complex feat that has repeatedly proven difficult for the autonomous vehicle industry, according to experts.

“The kind of reporting that is used in insurance is much more standardized than any other kind of collision data that you have around,” said Luigi Di Lillo, Head of Products and Partnerships for Swiss Re. “The novelty of this study is to connect the insurance claims data to safety directly.”

Waymo’s driverless vehicles, which currently operate in San Francisco and Phoenix, experienced 76% fewer accidents involving property damage compared to human-driven cars, according to the study. For every million miles traveled, human drivers filed 3.26 property damage claims compared to 0.78 for Waymo’s driverless cars.

“This is an ama