Hacker News Board Hacker News, browsed like a classic message board
Live data from Hacker News

New Study: Waymo is reducing serious crashes and making streets safer

waymo.com

Post reply on HN
Print view View original

21–30 of 413 posts

Re: New Study: Waymo is reducing serious crashes and making streets safer

#21

Post by somanyphotons »

buckle8017 wrote: post #14
carlgreene wrote: post #10

This will surely get some skepticism as it's a Waymo study, but it's nice to see a real‐world dataset this large at 56M miles. An 85 % drop in serious‐injury crashes and 96 % fewer intersection collisions is a strong signal that Level 4 ADS can meaningfully improve safety in ride-hail settings. Still curious about how much of that comes from operational design versus the core autonomy, but it’s a big leap beyond “nov…

It's an incentive problem. Uber/Lyft drivers are strongly incentivized to drive as quickly and aggressively as possible. The individual drivers are trading risk for cash. A company like Google isn't going to make that trade because it's actually the wrong trade across millions of hours.

I may be out-of-date here, but I had thought the accelerometers in the phone detected if drivers were too jerky in the movements of the car and that the drivers would be informed of poor service

Re: New Study: Waymo is reducing serious crashes and making streets safer

#22

Post by prasadjoglekar »

timewizard wrote: post #11
returningfory2 wrote: post #6

Earlier quoted context omitted.

Tens of thousands of Americas are killed every year in car crashes. The existence of a case or some cases where a self-driving car caused injury has zero value. What matters is the rate of cases per mile driven.

The US drives about 3.2 trillion miles per year. Waymo has 56.7 million miles over several years. Their percentage data is essentially useless. 50% of fatalities involve alcohol or drugs and are often single vehicle accidents. 25% involve youth or inexperience. 15% involve motorcycles. 15% involve pedestrians. What I really need to see is a complete breakdown of every accident a Waymo has had. Then I can start to com…

This is the statistic I was hoping to see. It wasn't in the paper as best as I looked.

Re: New Study: Waymo is reducing serious crashes and making streets safer

#23

Post by shermantanktop »

timewizard wrote: post #11
returningfory2 wrote: post #6

Earlier quoted context omitted.

Tens of thousands of Americas are killed every year in car crashes. The existence of a case or some cases where a self-driving car caused injury has zero value. What matters is the rate of cases per mile driven.

The US drives about 3.2 trillion miles per year. Waymo has 56.7 million miles over several years. Their percentage data is essentially useless. 50% of fatalities involve alcohol or drugs and are often single vehicle accidents. 25% involve youth or inexperience. 15% involve motorcycles. 15% involve pedestrians. What I really need to see is a complete breakdown of every accident a Waymo has had. Then I can start to com…

How would you compare single-incident reports to get to a meaningful conclusion?My guess is that Waymo makes mistakes that a human wouldn't, and vice versa. At that point the overall safety record, which normalizes those differences, seems the most relevant.

Re: New Study: Waymo is reducing serious crashes and making streets safer

#24

Post by pests »

somanyphotons wrote: post #21
buckle8017 wrote: post #14

Earlier quoted context omitted.

It's an incentive problem. Uber/Lyft drivers are strongly incentivized to drive as quickly and aggressively as possible. The individual drivers are trading risk for cash. A company like Google isn't going to make that trade because it's actually the wrong trade across millions of hours.

I may be out-of-date here, but I had thought the accelerometers in the phone detected if drivers were too jerky in the movements of the car and that the drivers would be informed of poor service

DoorDash (so food not people) will give you a report for hard braking / acceleration but it doesn't actually affect anything afaict.

Re: New Study: Waymo is reducing serious crashes and making streets safer

#25

Post by Terr_ »

buckle8017 wrote: post #14
carlgreene wrote: post #10

This will surely get some skepticism as it's a Waymo study, but it's nice to see a real‐world dataset this large at 56M miles. An 85 % drop in serious‐injury crashes and 96 % fewer intersection collisions is a strong signal that Level 4 ADS can meaningfully improve safety in ride-hail settings. Still curious about how much of that comes from operational design versus the core autonomy, but it’s a big leap beyond “nov…

It's an incentive problem. Uber/Lyft drivers are strongly incentivized to drive as quickly and aggressively as possible. The individual drivers are trading risk for cash. A company like Google isn't going to make that trade because it's actually the wrong trade across millions of hours.

Good point: Part of Waymo's safety stats comes from settings that are probably tuned right now in favor of safety stats even if it means a longer or less-profitable ride. It doesn't care if you're going to lose your job if you're five minutes late.

So a fairer comparison would be contrasting Waymo rides to trips conducted by the Ultra Safe Even If It's Slower Chauffeur Company.

Re: New Study: Waymo is reducing serious crashes and making streets safer

#26

Post by crazygringo »

timewizard wrote: post #11
returningfory2 wrote: post #6

Earlier quoted context omitted.

Tens of thousands of Americas are killed every year in car crashes. The existence of a case or some cases where a self-driving car caused injury has zero value. What matters is the rate of cases per mile driven.

The US drives about 3.2 trillion miles per year. Waymo has 56.7 million miles over several years. Their percentage data is essentially useless. 50% of fatalities involve alcohol or drugs and are often single vehicle accidents. 25% involve youth or inexperience. 15% involve motorcycles. 15% involve pedestrians. What I really need to see is a complete breakdown of every accident a Waymo has had. Then I can start to com…

> The US drives about 3.2 trillion miles per year. Waymo has 56.7 million miles over several years. Their percentage data is essentially useless.

No, that's not how statistics works.

The percentage data's accuracy depends mainly on the number of incidents recorded (and somewhat on the rate of incidents). But the percentage of the whole is completely irrelevant.

If you are basing something on 10 incidents but it's 50% of the total, it's still terrible accuracy.

Whereas if you are basing something on 100,000 incidents but it's only 0.1% of the total, it's still going to be quite accurate, assuming the incidents come from the same overall distribution.

Re: New Study: Waymo is reducing serious crashes and making streets safer

#27

Post by Scaevolus »

AIPedant wrote: post #7
floxy wrote: post #2

That's great news. I wonder how much insurance rates will go down when autonomous vehicles get popular. Seems like the liability portion could go way down.

The problem is that Waymos can't replace cars (or even Ubers!) since they depend on human oversight and problem-solving. Even at a 20:1 ratio it is not feasible for autonomous vehicles to employ 10m people to account for every American on the road.

The human oversight is mostly to tell off people breaking rules in the cars, not to fix the driving, and is quite easy to automate as they scale.

Re: New Study: Waymo is reducing serious crashes and making streets safer

#28

Post by femiagbabiaka »

Anecdotally, as someone who bikes a lot in SF, Waymo's are a lot safer than human drivers simply because they follow the letter of traffic laws. Stopping at stop signs, waiting for pedestrians to clear the box, following the posted speed limits, etc.

Re: New Study: Waymo is reducing serious crashes and making streets safer

#29

Post by Veserv »

returningfory2 wrote: post #6
cratermoon wrote: post #3

Waymo says Waymo is safe. https://www.latimes.com/california/story/2024-05-16/woman-ge...

Tens of thousands of Americas are killed every year in car crashes. The existence of a case or some cases where a self-driving car caused injury has zero value. What matters is the rate of cases per mile driven.

For which any such system must be assumed unsafe until proven safe and the burden of proof lies with the manufacturer to present sound, robust, transparent, third-party audited evidence.

You do not get to counter-argue: “What matters is the rate of cases per mile driven” without actually presenting that number with supporting evidence. Otherwise the only sound conclusion is the default presumption of non-safety.

In the case of Waymo, we have some tentative supporting evidence from this and other studies Waymo has run. However, that is still insufficient, even ignoring the lack of audits by non-conflicted parties, to strongly conclude Waymo is safer than a human. The evidence is promising, but it is only prudent to wait for further confirmation.

In contrast, Cruise was almost definitely not safer than a human driver.

In 2023, Cruise ADS cars drove 2,064,728 miles [1] and were involved in, by my count, 29 collisions with 5 causing injury [2], namely incidents on 2023-05-04, 2023-05-21, 2023-06-09, 2023-08-18, 2023-10-02.

That is ~72,000 miles per collision and ~400,000 miles per injury in contrast to the national human averages of ~500,000 per reported collision (which is non-comparable) and ~1,270,000 miles per injury (which is comparable). So, absent a more detailed analysis, Cruise ADS cars were ~3x MORE likely to be involved in a injury causing collision per mile.

Details and evidence matter in these discussions. Blanket rhetoric and optimism is not prudent when discussing new safety-critical systems.

[1] https://thelastdriverlicenseholder.com/2024/02/03/2023-disen...

[2] https://www.dmv.ca.gov/portal/vehicle-industry-services/auto...

Post reply on HN