The data coming out from an independent study of Waymo autonomous vehicles is, frankly, amazing. Swiss Re, one of the largest global insurance firms based out of Zürich, reports that 25.3 million fully autonomous miles drive by Waymo vehicles resulted in a 92% reduction in car crash injuries.
In plain English, Waymo self-driving tech is 12.5x safer than human drivers.
Let’s dig into what that means!
It’s waymo safe!
If you take less risks, of course you will have less accidents. Waymo drives under very strict and controlled circumstances. Humans drive in way more varied and risky conditions.
And humans are also unquestionably idiots with their agenda.
Bots just do their job.
Waymos are on city streets. Hardly “controlled conditions.”
City streets are very much controlled conditions in what they’re referring to. Kept clean, typically with decent signage and repainted lanes.
You should go for a drive in San Francisco someday. If they can nail fully autonomous driving there, that shit can probably damn near drive in India.
I’m not going to fucking California 😭
Have you been to SF? Decent signage and repainted lanes would not be a great description.
I work in vehicle testing, specifically on ADAS and autonomous driving features. One thing that gets overlooked with Waymo is the fact that that fucker has a 6-figure price tag on the sensing equipment on it. You’ll never see that on a production level vehicle
The real point of self driving cars is not to have to buy one. They should be ubiquitous fleets of taxis that never sleep, not personal possessions that sit in the driveway 90% of the time.
From where I’m sitting that’s very far from reality for a whole lot of reasons. I’d be happy to be wrong (because cars are unacceptably unsafe as-is) but there’s a long way to go
Okay, feel free to tell me all the reasons.
I’m not going to write you an essay
There are good video essays on the subject by Benn Jordan, Not Just Bikes, and Adam Something that (mostly) are a good overview of the problem and their videos explain things better than I’m ever going to write
Also the eternal taxi idea just seems like a shit excuse for not building other transit options. Like why not build trains and bike lanes instead of that
Even if it was a quarter million bucks, it would quickly pay itself off. Keeping a taxi car available and operating 24/7 is a big deal for operational costs. $25/hour on a $250k device is a recovered in just over a year. It’s all profit after that.
“Never” is a long time. Technology is always getting cheaper, I see no reason why that sensing equipment won’t end up on a production vehicle at some point.
In the auto industry pennies matter. Inevitably some higher-up will convince himself that they can accomplish the same thing with just a camera and some software. Margins tend to be more important than quality. I’m not saying that’s right, its just what tends to happen
Isn’t that already happening with Tesla?
Happened awhile ago. It was and is still a bad idea IMO
That sensor array is fuckin sweet. If I’m going to trust a car to drive me, I want it to have laser beam eyes that see through pitch blackness and blizzard conditions.
You’re the engineer, I’m just a pickup truck driving comedian, so I’m assuming that I’ve just accurately described a commercial-grade LiDAR array.
The LiDAR arrays are dropping very quickly in price - they’re now low six figures. I anticipate they’ll eventually make production, probably with fewer sensors, but sensors of equal quality. Probably sooner than most folks realize.
Next week at your local big box for the price of a DJI drone
Of course the user named leadfoot is thrilled with auto-driving.
I’m personally skeptical, probably won’t enjoy for myself (at least not til I’m REAL old), but hey, I love this for folks who can’t or don’t like to drive!
Fewer car accidents is good for everybody.
I don’t like driving and am not “good” at it, but I’m suspicious by my nature, so probably won’t use it even if I end up with a car capable of it.
In the end, insurers will be the harbingers of autonomous vehicles.
In 2050, the insurance will be twice as high if you insist on having a steering wheel, and it will have a major impact on buying decisions.
Never trust an automated vehicle you have to buy insurance for. If it’s truly autonomous, then the actual person in the driver seat is irrelevant. There is no need to price risk individually. Any true self driving car should have a lifetime insurance policy included in the purchase price. The manufacturer is the one determining if crashes will occur. The liability should be entirely on them. Any company selling you a “self driving” card that still requires you to buy insurance is selling snake oil.
Mm. No. Where / how much you drive it has everything to do with the amount of risk and is completely personal.
Well, self-driving cars should refuse to drive in conditions that are outside the safety parameters they are designed for, regardless of exposure (time, amount, and condition driven), so the result should be the same.
Even standard warranty doesn’t define conditions that strictly, just time & mileage - and even that bcs of the wear.
No dude. If you live in a high crime city your insurance is going to cost more than if you live in an affluent gated community. Because the risk of theft or vandalism or damage is just greater. I’m not saying that’s right or wrong, it’s just a fact. But maybe facts don’t belong in this fantasy session about how we all wished insurance worked.
Sorry, was this a reply to my post?
I got a bit lost.Also, in the first and second world countries we don’t have crime-area based pricing tables, only like natural disasters, but I’ve never heard about it applied to car insurance (tho it’s def possible, but they wound need to be less detailed).
Watch these Tesla’s eat shit in “Full” Self-Driving mode because there’s fog, it’s sort of darkly hilarious that human needed to intervene to avoid hitting a fucking freight train:
I agree with you, but it will never happen without legislation forcing it. The insurance companies don’t care who the money comes from (for the most part), so take them out of the equation. The person purchasing the car will (rightfully) feel that they shouldn’t have liability because they’re not driving the car, but the manufacturer/dealer will also (rightfully) feel that they can’t control the environment that the owner subjects the car to, so the liability should be on the purchaser.
Right now, if you don’t maintain your tires, and you lose traction and cause a wreck, you’re at fault. If you don’t maintain your brakes and they fail and you slam into the back of another car, you’re at fault. Repeat ad nauseam for every part of the car.
Unless everything becomes leased (oh god, I can hear the comments about ‘you will own nothing, and you will be happy’ coming) and the manufacturer/dealer can force inspection of the car every x00 miles at the purchaser’s expense, they will happily (and successfully, because they’ll definitely sway the majority of american idiots with their ‘dire warnings’ about giving up ownership of your vehicle) that they shouldn’t be liable because they can’t ensure owners don’t set up a dangerous situation.
I also don’t see them ‘grounding’ a vehicle because a sensor says something is wrong. That is just screaming as the bad PR looms for the companies that would spearhead that thrust.
Your comment needs more paragraph-long sentences with multiple parentheticals. I can almost read it. /s
Actually, if you take out the parentheticals, your third paragraph doesn’t even make sense.
they will happily that they shouldn’t be liable…
Happily argue?
If you can’t keep track of your own sentences enough to not miss an important word, how can you expect other people to do so? Maybe ease up on the parentheses.
Sory.
Edit: Sorry
As it should. Nothing hurts your freedom more than an untimely death. And cars kill a LOT of people.
Sure, insurers balance portfolio & premiums, however most insurance products (and most insurance coverages) are designed (developed) to be profitable on their own.
So, insurance companies benefit vastly from higher premiums, not lower ones (short of something being uninsurable ofc).
(Eg healthcare in USA - insurers want higher prices too & are incentivized to make that happen.)Also depends what specifically WayOfTheMoo even insures. Maybe it’s just car liability insurance and not every scratch & ding comprehensive coverage. Or maybe SwissRe reinsures only damages over a certain amount or only (over) when total insurers claims sum up to a certain amount (reinsurance contracts can vary vastly … tho on the other hand they also like to collect all raw data from insurers to nerd over extra numbers).
This comment section is surprising me with both thoughtful and dark observations about the article. Well, that’s rad! I was expecting a more “good news” crowd on the Uplifting News board, but if y’all want Dark Futurism, I can hang.
You’re 100% correct. Non ADAS vehicles will be a luxury good. There will probably be social pressures, similar to seatbelt adoption, pressuring folks to not drive themselves. 2050 feels like a reasonable time horizon for that to start.
While I don’t disagree that autonomous cars are safer, I think they are comparing the Waymo city miles against mixed mileage and when you are driving at reduced speed the chance to have an accident lowers significantly.
But yes, people get distracted all the time, can be angry, tired, etc. that will negatively affect their driving.
The article covers an academic-style research paper. You might find that section of the full research paper interesting! You spotted something important, but I think you think that city driving is safer, when the opposite is true:
Here’s the part you might find interesting, the “12x safer than human” claim likely greatly understates the safety advantage, just due to the methodology of the study:
“The garaging zip code of the insured vehicle was used as a proxy for the city (Phoenix, San Francisco, Los Angeles, Austin) in which the vehicle drives. Waymo also almost exclusively operated on surface streets (non access-controlled freeways) with a unique distribution of driving that is representative of a ride-hailing fleet. In contrast, the benchmark represents the privately insured driver population that resides in these geographic regions. The associated benchmark mileage has more freeway driving than the Waymo ADS. There are several considerations when examining these results with respect to this limitation. First, freeway driving has a lower crash rate (Scanlon et al., 2024a). Including freeway driving makes this benchmark crash rate artificially lower, so, by including freeways in this study’s benchmark, the benchmark crash rate underestimates the true driving risk of where the Waymo ADS operates. Second, driving outside of these denser urban areas that the Waymo ADS operates would likely represent a reduction in overall relative crash risk. For example, commuters from the city would likely experience a reduced crash risk as they travel to less densely populated areas (Chen et al., 2024). Previous studies have shown that most injury collisions occur within a small radius from residency, and that American drivers rarely travel far from their place of residence, with approximately 80% of one-way household trips being less than 10 miles (DOE, 2022). Third, the benchmark drivers garaged in the Waymo deployment area are not operating with the same distribution of mileage within the geographical limits as the Waymo ADS. Chen et al. (2024) explored the effect of Waymo’s driving distribution on benchmark crash risk and found that - should the benchmark driving distribution match Waymo’s in San Francisco, Phoenix, and Los Angeles - the benchmark police-reported crash rates would have been between 14% and 38% higher. Due to all three of these limitations being expected to artificially suppress the benchmark crash rate (underestimation), the benchmarking results in this study are considered to be conservative. Surely, there is an opportunity in future work to leverage new data, such as insurance telematics, to more precisely define and leverage the benchmark driving exposure data to better account for this potential confounder.”
Less accidents per km on highways because it’s all people going in the same direction at a more constant speed.
Yep - of course they’re safer when they take at least 2-3x longer to get anywhere.
