Uber's Self-Driving Car Saw Pedestrian 6 Seconds Before Fatal Strike, Says Report (tucson.com) 253
An anonymous reader quotes a report from Arizona Daily Star: The autonomous Uber SUV that struck and killed an Arizona pedestrian in March spotted the woman about six seconds before hitting her, but did not stop because the system used to automatically apply brakes in potentially dangerous situations had been disabled, according to federal investigators. In a preliminary report on the crash, the National Transportation Safety Board said Thursday that emergency braking is not enabled while Uber's cars are under computer control, "to reduce the potential for erratic vehicle behavior." Instead, Uber relies on a human backup driver to intervene. The system, however, is not designed to alert the driver.
The report comes a day after Uber announced it will be ending it's self-driving vehicle testing in Arizona. The full NTSB report is available here.
$100 million? (Score:2)
How much will Uber be paying that family?
Re: (Score:3)
They paid them off within days.
But she was a crackhead who wasn't in touch with her family. They likely got off real cheap.
and hardtime for the CEO/VP's at uber! (Score:2)
and hardtime for the CEO/VP's at uber!
Re: $100 million? (Score:4, Insightful)
A hero implies courage. She didn't volunteer to test it. More like a martyr if you ask me.
Re: (Score:3)
We're _all_ test rats.
Re: $100 million? (Score:4, Insightful)
Re: $100 million? (Score:3)
So we are agreed that she was neither a hero nor a martyr, but rather a random nobody high on weed and meth, randomly walking into traffic?
Re: $100 million? (Score:4, Funny)
She has the legal right to walk anywhere, against any light, under any conditions, and even leap out in front of a car in traffic.
ItÃ(TM)s the law.
Kellogg's called and they want you to return your law degree. It was a mixup; you were supposed to receive a decoder ring.
Re: (Score:2)
Wait, what now? (Score:5, Insightful)
This can't be right. They are saying that Uber's self-driving car rig is neither designed to stop for nor alert the driver about pedestrians obstructing the path of the vehicle. It's just designed to... log them?!
What part about this is considered "self-driving" then, exactly?
Re: (Score:3, Insightful)
I really liked that comment from the last story: "Oops, we left it in murder mode."
Just sums up the whole situation so well.
Re:Wait, what now? (Score:5, Funny)
http://dilbert.com/strip/2018-01-30 [dilbert.com]
Re:Wait, what now? (Score:5, Interesting)
What did you expect from a company like Uber, exactly? Their entire business model was built around the concept of "laws don't apply to us, because, internet." This is the same company that beyond running an unregistered taxi service (cheaply disguised as "ride sharing", as if people just happened to be going in the same direction), was tracking critical journalists with "God Mode" and hiring private investigators against them to find things to blackmail them with, ran an active campaign to track law enforcement officers in order to evade them, ran a campaign against competitors like Lyft involving the massive use of fake ride requests, and literally dozens and dozens more types of general scumbaggery.
Anyone shocked that this same company shut off all pedestrian safety controls in order to get their "self driving cars" on the market sooner? Bueller?
Be vewy vewy quiet, were hunting... (Score:2)
humans. What's up Doc?
Re: (Score:3, Insightful)
This can't be right. They are saying that Uber's self-driving car rig is neither designed to stop for nor alert the driver about pedestrians obstructing the path of the vehicle. It's just designed to... log them?!
Apparently the way they had it was that the computer would drive and the driver would stop it from driving, if needed. That doesn't seem like an obviously ridiculous arrangement, even if having the computer ping the driver would have been better.
Except the driver they hired to do the stopping thin
Re:Wait, what now? (Score:4, Informative)
If you RTFA (I know, I know), you'll see that the backup driver said she was monitoring the self-driving interface, not fiddling with a phone.
Re: (Score:2)
Yeah right, so she should have noticed something with 6 seconds to do something about it.
It doesn't look like she was either monitoring a good enough system and she didn't look up for nearly 6 seconds.
I feel sorry for the driver but she is likely at fault. Except if Uber didn't tell her what the situ was with the car not looking out for objects.
Re: (Score:3)
Why do you assume that just because the car could notice something in the dark that human eyes could?
The whole point of self-driving cars is that they have better 'vision' than we do, that they (should) see things sooner and clearer.
At what point is it said that the unfortunate pedestrian was visible to the naked eye for six seconds rather than visible to radar, lidar, laser scan or whatever it is Uber cars use for area awareness?
Re: (Score:2)
Apparently the way they had it was that the computer would drive and the driver would stop it from driving, if needed.
So the HAL 9000 can rightfully claim that AI is NOT to blame, and this accident was the result of "human error".
Re: Wait, what now? (Score:2, Insightful)
HAL9000 was human error, too.
So many people who watched that movie walked away from it with the wrong moral in their heads. It wasn't a tale of technology gone wrong, it was a tale about what happens when someone has no choice but to comply with all orders fully and completely, and you don't think things through.
So yeah, this is a HAL9000 scenario. The AI was not allowed to slow down, stop, change lanes, or notify the driver. There were no available actions it could take to avoid the collision, because Uber
Re: (Score:2)
Apparently the way they had it was that the computer would drive and the driver would stop it from driving, if needed. That doesn't seem like an obviously ridiculous arrangement
You mean... like Autopilot?
Re:Wait, what now? (Score:5, Interesting)
Apparently the way they had it was that the computer would drive and the driver would stop it from driving, if needed. That doesn't seem like an obviously ridiculous arrangement, even if having the computer ping the driver would have been better.
I suspect that the alerting the driver wasn't gonna be any better. This is all speculation, but if Uber turned off the AEB (automatic emergency braking) system in self driving mode because it would have actuated the brakes too often making the driving erratic, simply notifing the driver about the same potential collisions would eventually result in alert fatigue in the driver.
Then the driver would start ignoring the alerts for basically the same reason that they turned off the AEB system originally.
If you've ever taught someone to drive who was overly cautious and braked all the time you know how that goes. The flaw is that simply seeing the pedestrian isn't enough to make the braking decision and the driver has to learn to anticipate the actions in the environment to drive successfully.
The driving instruction not only has to anticipate the environment, but also the driver to know when to intervene (grab steering wheel or apply the brake). Most self driving systems don't give these driving observers (not instructors) enough information and training to anticipate what the system is going to do to make them effective at intervening.
This is the problem of allowing people to walk before they can crawl... Sometimes you can't bypass stages of learning. If the system's AEB system activates too much, you kind of have to let it do that until it can learn not to do that.
Re: (Score:2)
Paying attention all the time while driving is hard if you do the driving yourself. When you are just watching for problems, it becomes pretty much impossible. Whoever thought his set-up was a good idea is the one that actually is responsible for the kill.
Re: (Score:3)
My new Prius nags me if it thinks my hands aren't on the wheel - I discovered while creeping along in a straight line in nearly stopped traffic with my hands resting lightly on the wheel. Apparently it senses torque inputs (even very slight ones) from the driver to determine whether someone is holding it, and in the very slow traffic the fact that I hadn't needed to apply a steering input for awhile was flagged as "not holding the wheel" by the software.
Why didn't the Uber car have some similar sort of mech
Re: (Score:3)
Why didn't the Uber car have some similar sort of mechanism to detect a non-responsive "driver"?
Uber has always been a "let's just try something now and worry about the what-ifs later" kind of company. That they jumped into the self-driving car business without adequately thinking things through doesn't surprise me at all. I imagine we'll be witnessing a few more fiascos, if/when they make good on their plan to start flying people around in oversized drones.
Not According to the FA (Score:5, Interesting)
Re: (Score:3)
She was, without doubt, looking at something much more interesting than a list of vehicle status updates.
Re: (Score:2)
QA always gets short shrift, the first place where companies try to save money. So hire someone at very low pay and have them do the on-the-road testing with minimal training (which is expensive). A quality company would have had at least two testers in the car, and it would have enabled audible and visible alerts. Now if you had two people both texting instead of testing, then you need to hire better QA.
Re: (Score:3)
Re: (Score:2)
Where's 110010001000 [slashdot.org] when you need him?
Re: (Score:2)
Still waiting on the Uber apologists to show up and tell us how much safer these systems will be and how these accidents will only happen once ever. Like developers never reintroduce errors.
Seems that one death has made Uber pull out of developing SDCs, at least they've handed in the license they had and are laying off the staff so looks fairly permanent. That's probably the best that could happen to SDCs because Uber's business is built on playing it fast and loose with the law. Which is one thing when it's about medallions and whatnot, but when they get people killed it's a big deal. Hopefully the serious players will keep going slow and steady, Waymo has been working for it for 9 years and
Re: Wait, what now? (Score:2)
Re: (Score:3)
Re: (Score:2)
! [rejected] master -> master (non-fast-forward)
error: failed to push some refs to 'basic-freaking-safety-features-that-youd-be-criminally-negiligent-to-disable'
To prevent you from losing history, non-fast-forward updates were rejected
Merge the remote changes before pushing again. See the 'Note about
fast-forwards' section of 'git push --help' for details.
Re:Wait, what now? (Score:4, Insightful)
A very poorly written article which resulted in a poorly written summary.
Buried in the article: "Uber also disabled the Volvo's factory-equipped automatic emergency braking system when the vehicle is in autonomous mode, the report said."
The Volvo XC90 comes with a feature they call "City Safety". https://www.media.volvocars.co... [volvocars.com]
This is an auto-braking system with sensors. Uber's autonomous system has its own braking and sensors. It's understandable from a system perspective that they don't have two separate, independent, systems deciding when to apply the brakes operational at the same time.
The poorly written article makes it sound like Uber's system either didn't have a feature for braking for obstacles or that it was disabled. This is not accurate. It does appear that Uber's system failed to either detect the pedestrian or to brake when detected.
It's probably also true that when testing they don't use the factory cruise control to maintain speed on the highway. There are likely other standard functions not used when the autonomous equipment is under test.
Re: (Score:2)
Replying to my own comment for those who don't follow the link to the details of the standard Volvo system.
Cyclists crossing the path of the car or suddenly swerving out in front it. Depending on the situation, City Safety is able to avoid a collision if the relative speed difference is up to 45 km/h (28 mph). At higher speeds, the automatic braking can mitigate the consequences of the collision.
Pedestrians walking out in front of the car. City Safety is able to avoid a collision at speeds u
Re:Wait, what now? (Score:4, Insightful)
Then again the pedestrian not wearing dark clothes or having reflectors on the bike while walking in front of a car at night would probably have helped more.
How would that have helped? The car detected her and didn't brake, and the "driver" wasn't looking and so couldn't brake.
Re: Wait, what now? (Score:4, Interesting)
How would that have helped? The car detected her and didn't brake,
As he pointed out, they're talking about two different systems. It's hard to parse out, but I think what they're trying to say is that the built in factory system detected her but couldn't break because it was disabled, but Uber's own system didn't detect her.
If that's the case then yeah, maybe reflectors and such would have helped the Uber system detect her and stop. On the other hand if the factory system detected her and theirs didn't then it also means that their system is pretty shit.
Re: (Score:2)
I trained drivers for a decade. Six seconds is enough time to stop a semi at 55mph, but not enough to stop a car at 45mph???? Six seconds in a car isn't even an emer
No, you are wrong. (Score:5, Informative)
And I suspect intentionally so.
It has been clearly stated, BY UBER that they both disabled the Volvo emergency braking AND their own emergency stop/avoidance code.
To imply it was just the Volvo system is trying to avoid the fact that Uber had intentionally removed their own code that would make emergency stops.
They had intentionally made a system that had NO automated method of making an emergency stop, and no method of warning the secondary driver that one was needed (which would be stupid anyway,. they are supposed to be testing a system for future use without such a driver).
They should be hammered for this - it would be the equivalent of removing the emergency brakes from a buildings lifts, or the airbags from a car, without telling anyone.
Re: (Score:2)
They had intentionally made a system that had NO automated method of making an emergency stop,
Somewhere in an alternate universe your hypocritical clone is complaining that an automated vehicle stopped suddenly and resulted in an inattentive motorcyclist rear ending the vehicle and dying.
They had an automated method of making an emergency stop. It was the brake pedal and a human brain. Currently that's the most robust method with presumably the best false positive\negative ratio.
Re:Wait, what now? (Score:5, Informative)
The poorly written article makes it sound like Uber's system either didn't have a feature for braking for obstacles or that it was disabled. This is not accurate.
No, you're the one spreading misinformation. To quote the NTSB report directly:
According to Uber emergency braking maneuvers are not enabled while the vehicle is under computer control to reduce the potential for erratic vehicle behavior. The vehicle operator is relied on to intervene and take action. The system is not designed to alert the operator.
Basically they had a system that calculated it would crash, but did nothing and warned no one. In other words, the same as having no system at all.
Re: (Score:2)
" It's understandable from a system perspective that they don't have two separate, independent, systems deciding when to apply the brakes operational at the same time."
You missed the bit where they disable their own system too, this is why the car didn't stop. Not 2 systems but zero braking systems. If they were going to disable their own system then they should of enabled the original Volvo system if possible.
None (Score:5, Insightful)
Uber wants data and they don't care how they get it. You can't get data on risky events if you're too cautious. kinda like how they used to vivisect criminals.
Re:Wait, what now? (Score:5, Insightful)
"They are saying that Uber's self-driving car rig is neither designed to stop for nor alert the driver about pedestrians obstructing the path of the vehicle."
What they seem to have done is taken a vehicle with substantial safety equipment, left the sensors operating, but turned off the accident avoidance features. Then they added a distracting task (monitor the system) for the driver. Then, SURPRISE!!!, something went lethally wrong.
We'll have to wait until folks with time and full information perform an analysis. But it sure looks at first glance like this was/is a questionably well designed test program.
Re: (Score:2)
Fully automated Trucks of Peace. Put an Arab patsy in one, direct the drone truck to ram into some people, and terrorize your population into accepting whatever totalitarian decadence you can dream up!
Re: (Score:2)
This can't be right. They are saying that Uber's self-driving car rig is neither designed to stop for nor alert the driver about pedestrians obstructing the path of the vehicle. It's just designed to... log them?!
What part about this is considered "self-driving" then, exactly?
And people going to jail for this will be exactly zero.
Re: (Score:2)
It did drive by itself, obviously. It drove over somebody, but it was clearly self-driving doing that. I think you are confused about the terminology used here.
Re: (Score:2)
It's a self-driving car, not a self-stopping car.
Re: (Score:2)
I think they assumed (incorrectly) that the testing drivers will be alert. It was required to have an alert human beind the wheel during testing, but given that this was Uber they undoubtedly cut corners.
Re: (Score:2)
Re: (Score:3)
Re: (Score:2)
I can vouch for that. I was on the road with a pedestrian standing on the medium strip. She slipped and fell right in my path, If I had not been playing close attention I would not have already started shifting lanes, due to concern about risk and thus avoiding action at the risk being actually expressed was readily achieved. Smart driving is all about minimising risk, not just for yourself but for others. It seems like automated vehicles under current design standards will kill without blinking an eyelid.
Re: (Score:2)
So... (Score:2)
Criminal negligence, then.
Classy.
No Carmageddon then, this time (Score:2)
Might still happen though, if the manufacturers keep up the shoddy security we've come to expect from them.
Time for Regulatory Control (Score:5, Interesting)
Get this: The car "saw" the person 6 seconds before striking the person. The emergency system was disabled. The emergency system was not set up to alert the driver. So many things wrong here that would have been avoided had there been sensible regulations written by sensible engineers.
Re: (Score:2)
Re: (Score:2)
Hmm...you seem to think that regulations (Rules of the Road, that sort of thing) are written by engineers. Alas, the truth is that they're mostly written by politicians....
Re: (Score:2)
Hmm...you seem to think that regulations (Rules of the Road, that sort of thing) are written by engineers. Alas, the truth is that they're mostly written by politicians....
No, I'm fully aware of who writes them... that's why I specified "by sensible engineers" in my statement.
Re: (Score:2)
Quick and easy. Just expropriate the FAA regs, global replace 'aircraft' to 'vehicle'. Delete the silly stuff. That's an alpha.
Give the auto industry five years to get the ECUs up to snuff. Then they can get _started_ on self driving cars.
Re: (Score:2)
Re: (Score:3)
These vehicles are already regulated.
Re: (Score:2)
Re: (Score:2)
Uber self-driving car (Score:2)
Hmmn
on an FYI basis, my self-driving car (written in Python using PyTorch) does not work either...
Just saying !
Uber Solves Los Angeles Homeless problem (Score:3)
The New Uber Terminator Car will eliminate any unwanted pedestrians.
Ask us how!
Re: (Score:2)
It'll knock them clear to Seattle.
Split attention (Score:4)
There was a human in the Uber car. Theoretically she was there to provide a human element that backed up the self-driving. But she was not giving her full attention to safety.
https://www.theverge.com/2018/5/24/17388696/uber-self-driving-crash-ntsb-report [theverge.com]
I think Uber should change their procedure. Have two humans in the car: one to provide safety backup to the driving, and one to manage the iPad app. That will cost more than just having one human in the car, but would have saved a life in this case.
Note that the NTSB says the pedestrian was crossing unsafely, which contributed to the incident. Again quoting from the above-linked article in The Verge:
I don't see what the drugs in her system had to do with anything; I think this would have happened the same way even if she was completely sober. But she was doing an unsafe thing when she died.
Re:Split attention (Score:4, Insightful)
If she was sober she might have looked both ways before crossing a street or noted that an oncoming vehicle wasn't slowing to avoid her. If I was crossing the street at night sober and a car wasn't stopping for me I would stop and let it pass and then proceed once it was safe.
Re: (Score:2)
The problem that we're seeing with auto pilot cars with humans as backups is that there is lag time for the human to realize that they need to become rapidly involved in dangerous situations. That lag time is much greater than a full-time driver that isn't expecting the car computer to drive the car.
We automatically react to dangerous situations and try to avert them as full time drivers.
A "almost driver" in a self-driving car is likely not as attentive as a full time driver. The person in the Uber car wa
Re: (Score:2)
I don't see what the drugs in her system had to do with anything; I think this would have happened the same way even if she was completely sober. But she was doing an unsafe thing when she died.
She was crossing at a place everyone crosses. A bike path crosses the street there, though it's not printed on the pavement, and the right hand lane, which she was half in when struck, is a bike lane. The drug testing is obvious, it's the same reason it's done for worker comp. If she was high in the last week then all fault is hers, despite all facts and evidence to the contrary. Everyone knows this is true because drugs are the devil and the cause of all problems.
Re: (Score:3)
To quote the NTSB report [ntsb.gov] itself:
In this area, northbound Mill Avenue is separated from southbound Mill Avenue by a center median containing trees, shrubs, and brick landscaping in the shape of an X. Four signs at the edges of the brick median, facing toward the roadway, warn pedestrians to use the crosswalk. The nearest crosswalk is at the intersection of Mill Avenue and Curry Road, about 360 feet north of where the crash occurred.
The accident could've been easily avoided, if either she or the driver been paying any attention at all. But really, what do you expect when a phone-addicted ex-con encounters a homeless meth-addict while "driving" a vehicle with experimental software from a "ride-sharing" company?
Confusing article is confusing. (Score:2)
because the system used to automatically apply brakes in potentially dangerous situations had been disabled
Sensors on the fully autonomous Volvo XC-90 SUV spotted Herzberg while the car was traveling 43 miles per hour and determined that braking was needed 1.3 seconds before impact, according to the report.
A diagram in the NTSB report shows that the Uber system determined that the SUV needed to brake when it was at least 20 meters (65.6 feet) from Herzberg; it was traveling 39 mph (63 kilometers per hour) at impact. Kornhauser said that was enough distance for the SUV to stop, or slow considerably to mitigate damage from the crash.
Uber also disabled the Volvo's factory-equipped automatic emergency braking system when the vehicle is in autonomous mode, the report said.
So what was disabled? The factory auto brake, or Uber's auto brake? Surely they weren't allowing a car that wasn't able to brake itself out on the roads. Does the Uber system have a separate "emergency braking" subsystem?
Headline says the system "saw" her 6 seconds before impact, but it determined that it needed to brake just over 1s before impact? (traveling at about 59fps at impact)
This article sucks.
Re: (Score:2)
Seems pretty straight forward to me.
Uber disabled the factory emergency braking system. Understandable, if they have something better...
Uber's system detected the pedestrian 6 seconds before impact. Fine.
It them determined, 1.3 seconds before impact the the pedestrian was going to get hit if it did nothing. Makes sense, the woman was walking across the road, not standing still in the middle of the lane.
The system did nothing because Uber had disabled that functionality, apparently due to too many false posi
Re: (Score:2)
Uber's system detected the pedestrian 6 seconds before impact. Fine.
It them determined, 1.3 seconds before impact the the pedestrian was going to get hit if it did nothing.
What happened to the "swerve to avoid a collision" option? Maybe they were planning that for the v2.0 release?
Car Development (Score:3, Insightful)
Re: (Score:2)
They did the right thing. False positives of slamming on the brakes isn't safe for other drivers. Extreme\dangerous reactions should only be initiated by a safety driver until the false-positive rate is confirmed to be lower than the safety driver's false negative rate.
Re: (Score:2)
Re: (Score:2)
It wasn't an engineering failure that caused the system to neither alert the driver or react to the situation.
Someone at Uber decided to disable the system because they couldn't get it right.
Nothing failed, it was turned off, so never had the chance to fail.
Uber skipped Failure Mode Effects Anslysis (Score:3)
A bit of system engineering discipline, FMEA [wikipedia.org] , might have prevented this. It helps when innovators are trained to manage complexity, capability, and risk.
Or common sense, but such things can be on short supply on large projects run by program managers who have an 'MBA' mentality: https://www.inc.com/nathan-fur... [inc.com]
Re: (Score:2)
Indeed. And apparently a mindset of "it is only software", just that it is not your browser that crashes here. The MBAs are the Plague and Cholera rolled all in one.
Arizona saved (Score:2)
So it's ok to kill people in every state except Arizona? Ridiculous! If you aren't operating in Arizona then you shouldn't be operating anywhere.
Breathtakingly irresponsible (Score:2)
This is utter madness. I would say 'pants on head' crazy but there's something a tiny bit endearing about that phrase, suggesting a gentle kind of crazy. No, this is 'dissociative, dig out your own eyeballs and eat them, psychotic break' insanity.
How about this: lets make a machine gun that's just always shooting bullets. It chooses which direction to point, and it's usually right. Sometimes, though, it might
Re: (Score:3)
It is probably a case where a lot of pressure to perform was applied to the engineers by some clueless MBA bean-counter and, in addition, nobody apparently had the experience and training to realize how woefully inadequate a "safety driver" is in such a set-up actually is.
On the positive side... (Score:3)
At least when the machines are actively trying to kill us, we can be fairly sure that once in a while they'll screw up and leave their victim alive.
GDPR (Score:3)
"
Weâ(TM)re sorry. This site is temporarily unavailable.
We recognise you are attempting to access this website from a country belonging to the European Economic Area (EEA) including the EU which enforces the General Data Protection Regulation (GDPR) and therefore cannot grant you access at this time.
For any issues, contact us. gdpr@townnews.com
(403 error.)
"
Re: (Score:3)
most of whom are probably pedestrians on account of the cost of owning a car.
EVERYONE is a pedestrian. And we have a society with specified rule of law carving out certain precautions DRIVERS of dangerous vehicles must take to help pedestrians stay alive. One of those being the operator of a vehicle is legally responsible to account if they kill a pedestrian that they had 6 seconds warning about
Re: (Score:2)
Except they didn't really have six seconds of warning... while it's true that the car saw the person 6 seconds before the collision, the car did not alert the driver in any way of this, so in fact the driver did not have any warning at all.
That said, the driver was not being attentive, and it is possible that the driver might have even not responded to a warning quickly enough if one had been given, but it's still entirely true that the driver did not have any warning about the collision.
Re: (Score:2)
If I was driving my non-autonomous car and didn't pay attention to the road, would I be responsible for killing someone, despite not seeing them walk infront of my car?
Yes, I would.
If my car had the capability to keep to a lane, maintain speed and stop at traffic lights, would I still be responsible?
Yes, I would.
That's about the limit of the capabilities of the Uber car. I assuming it can stop at traffic lights...
Re: (Score:2)
Re: (Score:2)
Streets that are paid for with things like gasoline tax, and sales tax on automobile purchases...
Re: (Score:2)
The law of man says the person controlling that big hunk of high speed metal needs to make sure they do so safely, and defines a set of rules to make it explicitly clear what is regarded as safe.
It says if you're negligent in your responsibility to driver safe and you kill someone, you go to jail.
Look both ways. Also don't run over pedestrians.
The driver was not a computer, the computer had no way to stop the car in an emergency. The person behind the wheel had that responsibility.
Re: (Score:2)
Re: (Score:2)
The street was lit perfectly well. The video they released was of poor quality.
Any person paying attention while driving would have seen them.
Cellphone cameras would have seen them, proven by people driving the exact same road at night and posting it to youtube.
Uber released the footage as media damage control. I wouldn't be surprised if they deliberately reduce the dynamic range to make it look like it wasn't entirely their fault, which it is.
Re: (Score:2)
The driver is at fault, unless a law has been past in Arizona that says a driver in an autonomous car is not require to pay attention to the road while testing.
If Uber told the driver to monitor the system instead of the road, they're telling the driver to break the law.
Just because your boss told you to break the law, doesn't get you off the hook.
Re: (Score:2)
Re: (Score:2)
Seems like something that needs to go through the court to figure out who's liable, Uber or the driver.
Most driving laws around the world would place the blame with the driver for not paying attention while driving.
It's probably going to depend on what Arizona set out when they allowed the testing on public roads in the first place.
Re: (Score:2)
If the backup driver was paying attention, she would have been likely able to avert the situation.
Instead, it appeared she was looking down in her lap doing "something".
She's there to prevent these situations from happening. She was negligent.,
Re: (Score:2)