For years, Waymo has been claiming that its driverless vehicles have the potential to be safer than humans. Now the company says it has the data to support it.
Waymo analyzed 7.13 million fully driverless miles in three cities – Phoenix, Los Angeles and San Francisco – and compared the data to human driving benchmarks to determine whether its cars would cause less injuries and police- were involved in reported accidents. And it was the first time that the company only studied miles from completely driverless operations, rather than a mix of autonomous and human-supervised driving.
Waymo analyzed 7.1 million completely driverless miles in three cities
conclusion? Waymo’s driverless cars were 6.7 times less likely than human drivers to be involved in a crash that resulted in injury, or an 85 percent reduction compared to the human benchmark, and were less likely to be in a police-reported crash. The chances were 2.3 times lower, or a 57 percent reduction. , This means that in the cities where Waymo operates, there were an estimated 17 fewer injuries and 20 fewer police-reported accidents than if a human driver had driven the same distance.
Waymo’s analysis comes at a difficult moment for autonomous vehicles. The company’s main rival Cruise has halted operations nationwide after an accident in San Francisco that resulted in one of the company’s driverless cars dragging a pedestrian 20 feet. Cruz allegedly withheld video footage of the incident from regulators and now faces a fine of up to $1.5 million from the state. Lawmakers and labor advocates are demanding a crackdown on autonomous vehicles, regardless of company ownership.
But Waymo said the timing of the release of its safety analysis — coming at a time when its main rival is facing its worst crisis in years — was merely coincidental. Still, Trent Victor, Waymo’s director of safety research and best practices, said the data can help the public understand that not every company developing autonomous vehicles is the same.
“What we want to do is give a clear picture to allow people to see the difference between Waymo and other AV companies,” Victor said in a briefing with reporters. “Another difference is that we are moving forward responsibly.”
The company is using public data to draw its conclusions, and is inviting other, third-party researchers to replicate its results. Waymo and other AV operators are required to report every accident to the National Highway Traffic Safety Administration as part of the agency’s efforts to spotlight self-driving cars in the US.
Waymo’s millions of miles were not entirely incident-free. The company said that overall, throughout the more than 7 million miles driven in the three cities, its vehicles were involved in only three crashes that resulted in injuries: two in Phoenix and one in San Francisco. All three injuries were minor, according to Christopher Cusano, a safety researcher at Waymo and co-author of the study.
Waymo’s analysis comes at a difficult moment for autonomous vehicles
Still, this is much lower than the accident rate for human drivers. Another way to look at this is to look at the accident rate per million miles driven. The human benchmark is 2.78 incidents per million miles. Waymo’s benchmark for its driverless vehicles was only 0.41.
One of the biggest challenges for Waymo was controlling for various factors when comparing its vehicles to human drivers. To present a fair comparison, Waymo needed to address statistical biases in the data, such as human drivers failing to report minor crashes, or differences in driving conditions.
For example, Waymo’s vehicles operate in geo-fenced areas in the three cities where they operate, which do not include highways. Human drivers do not avoid these types of roads. Human drivers also do not report some low-level accidents, such as minor fender benders. In contrast, Waymo is legally required to report every contact incident with another vehicle, no matter how small. As a result, Waymo needed to adjust its model to take those factors into account.
Waymo’s analysis “included underreporting adjustments for crashes reported by police or obtained from naturalistic driving study databases,” the company said, based on “a literature review of 12 previous studies comparing AV and human crash rates.” 1 book”.
Those millions of miles weren’t completely incident-free
“The goal was to say what a human crash population would look like if it were operating under the same conditions as an autonomous vehicle,” said John Scanlon, a safety researcher at Waymo and co-author of the study.
The analysis comes after a study that Waymo published in collaboration with Swiss Re found that the company’s driverless vehicles reduced the frequency of bodily injury claims by 100 percent, while Swiss Re’s human base The line is 1.11 claims per million miles.
The current state of the self-driving car industry is very unstable. AVs are rolling out in a handful of cities, but many say they are skeptical of the technology and don’t necessarily trust it any more than their own driving. The negative attention surrounding the Cruise incident, as well as Tesla’s recent recall of Autopilot, is the latest in a series of bad headlines that have led many to conclude that self-driving cars are a fad, or that Worse, there are even more dangerous ones. human driver.
Waymo has gone further than other AV companies in its use of data and statistical analysis to make the safety case for autonomous vehicles. Waymo insists that driverless cars are necessary to tackle the crisis of traffic deaths, of which there are about 40,000 a year in the US. The company often points out that driverless cars are never drunk, tired or distracted and are able to avoid the human errors that often lead to accidents and deaths.
The current state of the self-driving car industry is very unstable
Last year, the company produced two scientific papers comparing autonomous vehicle performance to human driving. The first analyzed reaction times when a crash occurs, while the second presented a novel methodology for evaluating how well autonomous driving systems avoid crashes. Waymo has also sought to measure the safety of its AVs by simulating dozens of real-world fatal crashes that have occurred in Arizona over nearly a decade. The Google spinoff discovered that replacing any vehicle in a two-car collision with its robot-guided vehicles would virtually eliminate all deaths.
It is an undeniable fact that there are far fewer AVs on the road than human-driven vehicles and thus, less data to draw conclusions from. Humans travel approximately 100 million miles between fatal accidents. Some experts claim that we will need hundreds of millions of miles from autonomous vehicles before we can begin to make more meaningful comparisons about safety.
But Waymo’s efforts appear to be a step in the right direction. “These reports represent a good faith effort by Waymo to evaluate the safety of its [autonomous driving system] “It compares to the safety of human driving,” said David Zuby, chief research officer at the Insurance Institute for Highway Safety. “The results are encouraging and represent a step forward in our evolving understanding of ADS security.”
Other researchers agree. “This provides the strongest evidence I’ve seen so far that ADSs have lower accident rates than humans,” said Carol Flanagan, a researcher at the University of Michigan Transportation Research Institute. “I think the more complex question is what this means in the broader journey of determining whether ADS is ‘safer than humans.’