Slashdot Mirror
Self-Driving Cars In California: 4 Out of 48 Have Accidents, None Their Fault
An anonymous reader writes: The Associated Press reports that 48 self-driving cars have been navigating the roads of California since the state began issuing permits last year. Of those, only four have been in accidents, and none of the accidents were the fault of the autonomous driving technology. Seven different companies have tested autonomous cars on California's roads, but Google, which is responsible for almost half of them, was involved in three of the four accidents — the other one happened to a car from Delphi Automotive. All four of the accidents happened at speeds of under 10 mph, and human drivers were in control during two of them. The Delphi accident happened when another car broadsided it while its human driver was waiting to make a left turn.
The AP pieced together its report from the DMV and people who saw the accident reports. But critics note that there aren't direct channels to find this information. Since one of the chief selling points of autonomous cars is their relative safety over cars piloted by humans, the lack of official transparency is troubling. "Google, which has 23 Lexus SUVs, would not discuss its three accidents in detail." Instead, the company affirmed its cars' accidents were "a handful of minor fender-benders, light damage, no injuries, so far caused by human error and inattention."
The AP pieced together its report from the DMV and people who saw the accident reports. But critics note that there aren't direct channels to find this information. Since one of the chief selling points of autonomous cars is their relative safety over cars piloted by humans, the lack of official transparency is troubling. "Google, which has 23 Lexus SUVs, would not discuss its three accidents in detail." Instead, the company affirmed its cars' accidents were "a handful of minor fender-benders, light damage, no injuries, so far caused by human error and inattention."
I expect the number haven't been publicized, because they are still to limited to have any significance, and also because the cars have been running under fairly tightly controlled conditions.
When there are a few hundred cars, running in all kinds of weather and traffic conditions, with millions of miles - if the numbers are still good, you can bet that they will be plastered all over the internet
Enjoy life! This is not a dress rehearsal.
4 Out of 48 Have Accidents, None Their Fault
I think that compares well to the average Californian.
4 out of 48 have accidents within one year. That's a lot.
Computers are still too stupid. The giveaway with autonomous cars is the need for centimeter precision in their navigation maps. These are trains on computationally enforced tracks, not free driving autonomous vehicles.
I'm beginning to think that this autonomous vehicle stuff will end up in tears.
Or is it normal that one out of twelve cars that is involved in an accident each year? And by calling it "only" the submitter suggests that the regular accident rate is much higher than that.
Perhaps you would like to read the story somewhere other than the NYT because paywalling is stupid and offensive, even if you know how to bypass it. Thanks, Seattle Times, for just showing me the flipping article.
Here's a news flash: You can get the same AP newswire article anywhere. Yet people still link the NYT. That's poor internet etiquette given that they paywall.
"You're right," Fisheye says. "I should have set it on 'whip' or 'chop.'"
what does it do at four-way stop if it was the second car there but the human in the first car is waving it on?
has it been tested on drivethru fastfood?
if someone is shooting at car, how does the car react?
does it get up to speed on onramps or does it merge into highway traffic going 45mph?
'48 self-driving cars have been navigating the roads... Of those, only four have been in accidents'
I know that the bigger point is that zero (known) incidents can be traced to the software making a 'mistake' (though even if the other driver is 'at fault', hard to say if a human would have done better at avoidance). The thing that strikes me though is the editorial bias here. *Only* 4 out of 48.. that's nearly 10%. That's far far above the percentage for the general population. It's perfectly likely that is simply a fluke of the small sample size, but implying that 4 out of 48 is a very promising rate of incident is pretty silly.
XML is like violence. If it doesn't solve the problem, use more.
This is all well and good having self driving cars, however, (ok, so this is through reading the unfortunate things that happen to people through stupidity, and malice on reddit, liveleak, etc)....
Hopefully we'll never get to the stage where you just 'pop the infant' into the car and tell it to go to grandmas (assuming grandma is there, and nothing happens along the way). Sad, far fetched, but you can bet that this will happen somewhere and some unfortunate may be hurt as a result in something entirely preventable).
This was a waking dream this morning. Sad thing to wake up to. Hopefully never happen.
While I get that the autonomous system wasn't at fault, the real question is, "Were the accidents something a human driver could have avoided?" I'm a professional delivery driver and one of the things that's constantly drilled into us is to essentially watch out for the stupid people. Sometimes you have to yield right-of-way because it's clear the other driver isn't going to. Do autonomous cars know that?
"a handful of minor fender-benders, light damage, no injuries, so far caused by human error and inattention"
In case any of those were done by human co-drivers in automated vehicles, this does not exonerate the automation from some share of responsibility. For example, if the presence or habitual use of the automation makes it more likely for the co-driver to become inattentive, it's partly to blame.
But would a human driver have been able to avoid the accident? On more than one occasion I've escaped a fender bender that would not have been my fault.
The autonomous may not have been at fault, but one wonders whether some of the accidents would have been avoidable by a fleshy driver.
That the vast majority of human drivers around them were able to avoid accidents despite the presence of dangerous automated cars.
Self-driving cars need to be banned.
Why - it wasn't their fault?
American components, Russian components, they're all made in Taiwan.
the preceding comment is my own and in no way reflects the opinion of the Joint Chiefs of Staff
One can be "in the right" and still not have done the right thing. For example, if the light is green I'm in the right not to slow down for the intersection. But that doesn't mean I shouldn't take precautions to check if someone is coming the other way. If I had I might have avoided the accident that was not assigned to my "fault".
On the other hand it's also possible that google cars will be better drivers than the average person. One might hope they use different CPUs for the texting and the driving.
Some drink at the fountain of knowledge. Others just gargle.
How does one make eye contact with an autonomous vehicle at an intersection, or when merging lanes? Human drivers will have to learn a separate protocol.
>> I don't want to drive a car, I might break a nail!
Yep, that's me. It won't surprise you how I travel around the world either: I've never actually flown a Boeing or an Airbus.
So who is a fault when breaks fail. It has happened. Who is a fault when tires blow out? Or fuel tanks catch fire, or airbags improperly deploy?
Liability is nothing new and ToS cannot waive rights that are not waive able.....
If information wants to be free, why does my internet connection cost so much?
Big issue with AI controlled cars in a human-dominated traffic is that AI doesn't react the same way people do. Sure, all-AI traffic would likely be more efficient and less prone to accidents, but we are nowhere near this. Instead we have AI that to humans is hard to predict.
For example, huge puddle on the road, most humans would unwisely drive through it. What would AI do? No idea, and I wouldn't want to be driving behind it when that happens. What about a hobo at the end of the offramp begging for change? Would AI freak out about pedestrian on the road? No idea, and I wouldn't want to be driving behind it to see what happens.
There is a difference between following traffic laws and not being at fault and not failing to avoid an accident an alert human driver would have.
If that were to be the case, although I'm not meaning to imply that it won't be, then the purchaser of a vehicle would be assuming such liability.... if people don't to bear that responsibility, then they won't buy such cars in the first place. The demand will stay low, and there will be no need to create any new laws prohibiting them.
File under 'M' for 'Manic ranting'
Honestly, you can say it wasn't their fault, but nearly 10% of them in 6 months have been involved in accidents. Even if it wasn't the fault of the technology itself, why is the accident rate so high?
NON-Self-driving cars need to be banned. There, I fixed it for you. :D
It's better to be the foot on the boot than the face on the pavement. ~~ tkx Kadin2048
Actually theres a historical precedent for legislation concerning semi autonomous non human traffic. Horses.
It's not entirely as ridiculous an analogy as it initially sounds.
This is why "no fault" insurance was invented. Even with people, the liability in traffic accidents is often complex. Require no-fault insurance, you insurance is responsible for you, mine is responsible for me, and you are basically done - at least as far as car insurance is concerned.
What may be left is civil liability. If a manufacturer produces a genuinely faulty product, they can be sued in a completely separate action. The laws are already in place for that. The problem I see is the exact opposite of what you are worried about: People will sue, regardless of the quality of the product. Because bad things aren't supposed to happen, and if they do, someone must be to blame.
So someone without insurance t-bones my autonomous car, and I want compensated. The other driver is broke, so I sue my car manufacturer. Stupid, but entirely possible under US law. Loser pays would be the simplest solution: if you file a stupid lawsuit, you'll be paying the other side's legal costs. In the case of class action suits, the attorneys for the class should be liable for the loser's legal costs if they lose.
Before any company brings autonomous cars to market, the US tort system has got to be fixed.
Enjoy life! This is not a dress rehearsal.
You are trolling, you trolly troll.
Maybe they are sticking to the rules - and the meatbags in the other cars assumed that they would behave with the standard nonchalance towards some of the rules that meatbags frequently display...
I know that if I were to stop at every yellow light where I could stop (with a bit of effort), I would soon have another car in my trunk. So I utilize the 'if it can be done safely' part of the rule fairly frequently... 9 times of 10, the car behind me also goes through on the yellow light...
From the story:
2 out of 48 have crashed by computer.
1 out of 48 have crashed with human driver.
1 out of 48 had someone crash into them at a junction
Denial doesn't fix bugs here.
It's likely they just miss a lot of the subtlety of driving, the "I think person X will do Y so I'm going to adjust my driving by Z" that goes on in peoples minds.
They've managed it with software on your devices, what makes your car different than any other device?
How safe autonomous vehicles will be when most of the vehicles on the road are autonomous. There will then be wars about which companies system is safest.
I think the war will be how fast and reliable they are. A system that's safer but takes longer to get people from A to B, or gives up and stops for any little thing in order to increase safety won't be too popular.
My life has only so many minutes. I don't want to spend more of them than I have to being slow cargo.
Because someone being blamable for accidents is much more important than having fewer accidents in the first place, right?
The major part of safely driving is anticipating another driver breaking the law. Doesn't look like computers can do that yet.
Way to not demonstrate any understanding of that axiom. If self-driving cars are highly correlated with an increased number of accidents, is that not something to be concerned about?
If you want a vision of the future, imagine a youtube comments section scrolling - forever.
You realise that humans were driving them in 50% of the 4 cases, right?
Honestly, you can say it wasn't their fault, but nearly 10% of them in 6 months have been involved in accidents. Even if it wasn't the fault of the technology itself, why is the accident rate so high?
Well two of the accidents were while the cars were under manual control, and two under automatic. That doesn't point to it being more likely to have an accident
There are plenty of times this happens with normal cars. There was one where the car accelerated on it's own when going off of a freeway, killing 1 person. The driver was sent to jail.. until several other people had this same issue. The car manufacture knew about this, didn't say anything, and issued a recall. Everything that can happen with an autocar has happened already, thing only otherthing would be programming errors.
People are stupid, and news sites like to print the most horrible, twisted and eye grabbing thing they possibly can. In that kind of environment, it only takes a single crash due to a software glitch to get the project outlawed statewide. I'd keep the information private too, and probably pay off the guilty party anyway just to keep them quiet.
Alter Aeon Multiclass MUD - http://www.alteraeon.com
From an interview with a Google Lexus 9000:
"How would you account for this discrepancy between you and the twin 9000?"
"Well, I don’t think there is any question about it. It can only be attributable to human error. This sort of thing has cropped up before, and it has always been due to human error."
"There has never been any instance at all of a computer error occurring in the 9000 series, has there?"
"None whatsoever, Frank. The 9000 series has a perfect operational record."
Accident rate in general: 4-5%
Accident rate so far with only 48 vehicles: 8-9%
Without details I'm not going to accept "someone else's" fault as fact.
Since the byline mentions mister Bond, presumably a derivation from the '64 007-flick Goldfinger ("No, Mr. Bond, I expect you to die!"), I need to point out that "fender" is a decidedly American English term, which one would call "mudguard" in proper English. Not that I care much one way or the other, but I have this vivid picture in my head of agent 007 feigning not understanding the reply.
Free, as in your money being freed from the confines of your account.
Or who knows what's happening, but there's a correlation, so we should be concerned with the situation and further investigate.
Saying "correlation does not equal causation" and washing your hands of the whole thing is intellectually lazy and does nothing to solve the problem.
No they haven't. At least in every country i have been. See "critical" software. Hell in Aus and NZ the EULA have often been overruled in courts for just video games! And cars are not a phone. If car makers could have forced you to sign such right away, you think they wouldn't have done that already? Oh and well its not like there already isn't critical software in cars already. ie ABS breaks.
If information wants to be free, why does my internet connection cost so much?
Whatever the lawmakers decide it will be. These aren't laws of physics. People just make it up and that's how it is.
it might not be going far enough though. I argue that committing an infraction that causes no accident, when an accident is highly likely if one does not commit the infraction, is preferable to following the rules more rigidly and having the accident.
I've had far too many times where I've had to cross the double-yellow line, or make an unsafe lane change, or had to drive into the bike lane, or had to pull out into an intersection against the light to avoid a collision to want to prohibit a computer from being able to do the same. Unfortunately it's difficult to teach even a seasoned driver to do this; I expect it might be very difficult for a computer to learn to predict when other drivers will engage in actions necessitating this kind of rulebreaking.
Do not look into laser with remaining eye.
Click-wrap agreements have been upheld by North American courts.
Who will be the first person to die from a driverless car?
Think your lawyer can stand up to Google and DARPA?
https://www.youtube.com/c/BrendaEM
Accident rate in general: 4-5% Accident rate so far with only 48 vehicles: 8-9%
Ah, no, the meaningful rate is accidents per miles driven, not per car. These cars are driven constantly (for testing, data collection, etc), surely way more that twice the average.
Honestly, you can say it wasn't their fault, but nearly 10% of them in 6 months have been involved in accidents. Even if it wasn't the fault of the technology itself, why is the accident rate so high?
Because they are constantly on the road, for testing and data collection. I would imagine that they drove more in these 6 months than many cars do in 6 years.
Get back to me when a few thousand of these cars are on the road.
If the computer had been in control, it might have been able to avoid the accident that the human was unable to avoid. But in the rush to find fault with the computers, nobody is considering that angle.
Is it just my observation, or are there way too many stupid people in the world?
Except, as delt0r pointed out, there's a difference between liability for "Office crashed and destroyed my resume" and "My car's air bags deployed when we weren't involved in a crash." The former is an inconvenience so a click-wrap agreement absolving the company from any damages due to their software might be annoying but isn't life threatening. The latter involves actual lives. Car manufacturers have already been held accountable for faulty automotive systems. Self-driving will be another feature of the car like ABS and air bags. If the self-driving mechanism decides that the two lane road actually has three lanes, the car manufacturer will face a recall at best and lawsuits at worst. I don't see the courts treating a car feature like software instead of like other car features. To quote delt0r: If car makers could have had "click-wrap agreement" equivalents, they definitely would have and you'd never see any recalls. ("Thousands of our cars' ABS doesn't work when it is raining? Must be a faulty system. Oh well, they all agreed not to sue us. Those folks better pay to get that fixed.")
My sci-fi novel, Ghost Thief, is now available from Amazon.com.
That number is based on Canadian yearly accident rates with 32 million registered vehicles and, obviously, a variety of driving patterns.
Yes, but without details it is hard to know what the story really is. One big fear is that self driving will result in apathy of the drivers.
Had the humans *just* taken over? Were they merely shuttling the thing to the test site? Were the other drivers simply distracted by Sauron's eye being bolted to a Lexus? I see a huge danger in the hand-off of a self driving car throws its hands up at a construction site, or due to snow, or if it sees a paper bag that just might be a rock in the road.
A scenario I expect is that we will see a net increase in accidents by human drivers of autonomous cars that outweighs the gains while HAL is in charge. It will likely be a very long time before autonomous cars can handle 100% of the driving, leaving humans to tackle the hardest ones after possibly weeks of never being in control, and while they are completely tuned out of the situation. Would that be OK?
Should we instead compare the costs and safety benefits of adding HAL to every car vs. spending similar amounts to require all drivers to take a couple day driving refresher every 10 years to renew their license? Heck, it is pretty shocking that we let folks get a license for a car without ever taking a proper driving class first, but I digress.
I'm too lazy to do the math, but how likely is a sample of 48 from the population likely to turn up that kind of accident rate?
Jesus was all right but his disciples were thick and ordinary. -John Lennon
just posting to erase my mis-mod. I hit "redundant" instead of the intended "insightful"
Given your assertion, which I do agree with to an extent, what about non-injury causing accidents like those described in the article? Are those merely an annoyance? That lost resume could cost you a job that could have made you $100k/year and that small accident could ding a Bugatti Veyron.
The difference between an ABS system failing and software glitch is very different. The former is likely mechanical a failure to manufacture a part properly or install it to specification. A software glitch is very difficult to prove and there's no defined parameters as to what constitutes a negligent failure. Liability for previously unknown defects is far less than known defects, it's hard for a judge or jury to wrap their head(s) around what software developers should or should not have known about a software failure.
I believe the summary said that that humans were in control of the cars during at least 2 of the 4 incidents. You can't blame the self-driving car if the self-driving feature is disabled and the human takes over. That would be like blaming Google Maps for bad directions if you turn it off, take a left turn when Google had said to turn right, and wind up lost.
With a 2 out of 48 accident rate, that's 4%. Of course, that's a very small sample size. It would be interesting to see how the accident rate changes with many more autonomous cars on the road.
My sci-fi novel, Ghost Thief, is now available from Amazon.com.
Honestly, you can say it wasn't their fault, but nearly 10% of them in 6 months have been involved in accidents. Even if it wasn't the fault of the technology itself, why is the accident rate so high?
One does need to factor in the number of hours/distance driven. I would not be too suprised if their per mile or per hour accident rate was much lower than the average. Of course, small numbers of vehicles will tend to give larger variance - this 6 month period might just be a statistical outlyer.
once you control for percentage of time on the road, sure. What are the figures for accidents per vehicle-mile?
There are approximately 254.4M registered vehicles in the US and of those about 6M are in an accident each year. That equates to 2.4% of the registered vehicles are in some sort of accident. From the AP report, 4 out of 48 autonomous vehicles were in an accident which equates to 8.3%. Based on the information presented, autonomous vehicles are 3.5 times more likely to be in an accident than non-autonomous ones.
There are three kinds of lies: lies, damn lies and statistics. -- unknown, but known not to be Mark Twain
That's a good start, but there's a lot more to it than that. An experienced driver can tell when a car in the next lane wants over (even though it isn't bothering to signal), and can even guage its desperation based on how bad the "body" language is getting, and proximity to things like intersections, exits, etc. An experienced driver knows when the driver behind them in traffic is an accident waiting to happen (is that dude reading a copy of Ivanhoe or something?), and get into another lane. An experienced driver knows that the left lane across from a Starbucks is to be avoided in the morning, particularly if there are multiple late-model coupes and sedans in that lane.
You could probably put all that smarts into an AI, but I doubt its there yet.
NON-Self-driving cars need to be banned.
Let the self-driving cars get a lower accident rate than human drivers first. (And none of this crap about humans being at fault -- good drivers can and should dodge accidents that aren't their fault.)
Don't waste your vote! Vote for whoever you want, unless you live in a swing state it won't matter anyways
A few possible explanations:
1. As mentioned elsewhere, perhaps the time on the road is far greater than the average vehicle, thus the accident rate per mile actually being lower
2. The vehicles somehow stand out and represent a distraction to drivers who are curious and straining to get a look at the driverless vehicle
3. The autonomous vehicles behave safely but do not necessarily follow the typical patterns that other drivers expect thus indirectly causing the accidents (though not directly at fault).
Number 3 might be cause for concern and further research. Not sure what you do about 1 and 2.
Modelling with a binomial distribution? 20% chance of getting 4 accidents from a sample of 48 drivers when the true accident rate is 4.5%. 37% chance of 4 or more.
With a true accident rate of 4.5%, seeing an 8-9% accident rate in a sample of 48 is common and not cause for alarm. Now, if it was 480 trials (with a <0.1% chance of seeing even an 8% accident rate), I'd be worried, but it's not.
The average passenger car weight in the U.S. is close to 3000 pounds. http://faculty.washington.edu/dwhm/files/MacKenzie%20Zoepf%20Heywood%20as%20submitted.pdf
Contribute to civilization: ari.aynrand.org/donate
CA should say, "You can use our roads, BUT only if you share all accident info."
Table-ized A.I.
Why? Are you an ambulance chaser?
Considering that even a Civic weighs close to 3,000 (and considering how many 4,000lb German sedans, 5,000lbs pickups and 6,000lb SUV'S I see on the roads), the median weight must surely be a lot greater...
I seem to recall the speed limit out in west Texas being 85mph...
Sure it is, but with a sample size of 48, most assumptions of distribution are going to give you some very big error bars.
If there were 2 of them, and 1 got in an accident in 6 months, would you really think it said anything about the car?
Sigh. If you took a sample of 48 US human-driven cars, you'd have a 20% chance of getting 4 accidents out of that 48.
I'm not trying to say that's huge, but that sample size of 48 is far too small to draw conclusions from statistics that contain millions of samples. The error bars are massive.
Beats me. I'm not against self-driving cars at all and am genuinely curious to see the data.
As a scientist who occasionally works with correlated datasets, I'm just sick of people trotting out that meme, like Pavlov's fucking dog, every time they see the word "correlation". "Correlated" means very damn much related, but people around here treat it like it means the opposite. If you're concerned about an outcome (like vehicular collisions), a correlation is extremely useful in finding the root cause.
If you want a vision of the future, imagine a youtube comments section scrolling - forever.
Apologies if this has been said already. Some number of accidents are inevitable in city traffic no matter who or what is behind the wheel. The only question I have to ask is this: If a reasonably competent human driver had been behind the wheel, was there an opportunity in any of these accidents for the human to take evasive action to avoid the accident, something that it may be currently impractical to program into the autonomous driving system?
Oliver's law of assumed responsibility: If you're seen fixing it, you will be blamed for breaking it.
Don't believe what they tell you. It's always someone else's fault.
Ah, I misunderstood your point of view. I agree with your ire at the over/misuse of that phrase.
And someone operating it outside their ToS would be using it illegally, thus reducing the liability for the corporate overlords.
Learn to love Alaska
The 6000lb pickups aren't "passenger cars". When your definition is wrong, then your observations will never match reality. So, do you change your definitions, or assert reality is wrong?
Learn to love Alaska
If we're strictly talking "cars," that technically rules out crossovers and even small SUV's such as a RAV4, hence the use of the [far-less arbitrary] phrase "passenger vehicles",,, ;)
Thanks. I had a feeling this was the case.
Jesus was all right but his disciples were thick and ordinary. -John Lennon
As stated, it's based on publicly available data from Statistics Canada. ~2000 fatalities for 32 million vehicles over the course of a year.
Can a good driver dodge accidents from the passenger's seat? Because 2 of those accidents were in self-driving cars with the self-driving turned off.
Without details on the accidents, we aren't in a good position to judge whether a good driver could have dodged this accident. The one detail we have is that the car was moving 10mph at the time. Not sure what to make of it -- was it read-ended, was it a parking lot accident, was it a bad lane change? Because being read ended is pretty cut-and-dried somebody else's fault (unless you *just* completed a lane change or something like that), but a parking lot accident could be very concerning.
Not that I disagree that the technology should be proven!
I assign the blame for the accident to the self-driving car, since they seem to have an unusually large accident rate and are keeping the details secret despite having video evidence of every accident. For all I know, control of the car could have switched from the computer to the fall guy one millisecond before the accident.
Don't waste your vote! Vote for whoever you want, unless you live in a swing state it won't matter anyways
Without more details, you have no basis to assume the claim of "someone else's" fault is false.
Well if it is based on fatalities, then:
Accident rate in general: 4-5%
Accident rate so far with only 48 vehicles: 0%
Without more details, you have no basis to assume the claim of "someone else's" fault is false.
Yes I do, for the same reason I assume the person on the phone who tells me I won a trip to the Bahamas is false. I believe things when I can see the details for myself and not simply going on faith.
Well if it is based on fatalities, then:
Accident rate in general: 4-5%
Accident rate so far with only 48 vehicles: 0%
It's based on accidents overall for which fatalities are a miniscule % (around 0.006)
Accident rate is uninteresting as a percentage of cars, only accidents/miles driven is meaningful. Considering that these are test vehicles and are likely driven far more than your average commuter car, I find the 2x increase in crashes/vehicle statistically meaningless. This of course becomes more complex when you consider other factors, such as that drivers that drive less currently are more likely to get in a crash per 100,000 miles.
Without details I'm not going to accept that you can avoid confirmation bias.
Except that 1) you're looking at prototypes 2) only 50% of the accidents were while the computer was in control 3) accidents per person/vehicle is not what you care about, it's accidents per mile.
No. You assume that a phone call stating that you won a free trip is false because of two things: you know that the odds of you winning such a thing is minuscule, and you've heard of phone scams and the odds of it being the later is higher.
I know it's false because I would never enter a contest for such a trip so I would have 0% chance of winning. Just like I know that the details of an incident are never reflected in the generalized statistics of such an incident.
NON-Self-driving cars need to be banned.
Ever driven a Mercedes for any length of time? Ever notice how expensive sensors are always failing?
So we have a German car manufacturer, known to generally make high quality vehicles, and the sensors are constantly failing.
How would an autonomous car know what is around it? Sensors.
Solve that first before you go banning NON-Self-driving cars.
"Someone needs to talk to the tree of liberty about its ghoulish drinking problem." by ohnocitizen
The only solution is to make it illegal to disengage the self-driving in non-accident situations and to have hefty fines for people gaming self-driving cars (perhaps by having obligatory dashcams on each of them).
Self-driving car already has excellent data from its radar and cameras. Just store accident data and report to police, immediately. People will learn FAST to avoid self-driving cars (and to hate 'em too).