Slashdot Mirror
Uber Ordered To Take Its Self-Driving Cars Off Arizona Roads (nytimes.com)
After failing to meet an expectation that it would prioritize public safety as it tested its self-driving technology, Uber has been ordered to take its self-driving cars off Arizona roads (Warning: source may be paywalled; alternative source). "The incident that took place on March 18 is an unquestionable failure to comply with this expectation," Gov. Doug Ducey of Arizona wrote in a letter sent to Dara Khosrowshahi, Uber's chief executive. "Arizona must take action now." The New York Times reports: Uber had already suspended all testing of its cars in Arizona, San Francisco, Pittsburgh and Toronto. "We proactively suspended self-driving operations in all cities immediately following the tragic incident last week. We continue to help investigators in any way we can, and we'll keep a dialogue open with the governor's office to address any concerns they have," said Matt Kallman, an Uber spokesman. The rebuke from the governor is a reversal from what has been an open-arms policy by the state, heralding its lack of regulation as an asset to lure autonomous vehicle testing -- and tech jobs. Waymo, the self-driving car company spun out from Google, and General Motors-owned Cruise are also testing cars in the state. Mr. Ducey said he was troubled by a video released from the Tempe Police Department that seemed to show that neither the Uber safety driver nor the autonomous vehicle detected the presence of a pedestrian in the road in the moments before the crash.
If there was someone at the controls, and he failed to correctly 'supervise' the self-driving - which presumably he was there for, and trained to do, why is is this not simply a case of dangerous driving/driving without due care and attention? The purpose of this exercise is presumably to test and improve the software/sensors - we all know they are not good enough 'yet', so surely it's down to human failure, as any ordinary accident involving a motor vehicle.
I wonder if there may be a class action suite to the city for letting unlicensed technology on the streets that caused human death?
If you put any simple, and common sense second thoughts aside, just because that's what you (and the entire planet) get:
Global warming "dispute".
Trump winning, albeit the Dems got the common vote.
Bipartisan battle.
NRA dictating weapon laws.
CRISPR
Non standard unit systems.
Disruptive technologies does follow its name. Globally. A farmer in Botswana has no say if his market is flooded with cheep US crops. If his family's lives have been destroyed? Simply unfit for free enterprise and trade. Same happens to Ted Rustbelt? Hell no!
Globalisation: 30 years in it's accelerating death spiral. If it's (finally) falls apart, blame China and EMEA.
Yes, that biker/pedestrian was a huge problem nobody noticed before.
Kill all bikers.
Testing is done for finding problems. They found one. Don't stop testing now!
Testing is also done (or should be done) in controlled environments until you get way past the alfa and beta stages. Putting the autonomous car on the road can be justified when the car doesn't need human supervision and it can deal with normal traffic conditions in day and night with the same performance as that of a human driver.
I seriously have no idea why autonomous cars in pre-alfa stage are on the roads.
Unfortunately most people are treating autonomous cars as software. And we know how software engineers think. Throw the alfa software to the public and fix mistakes afterwards. Oh and we're not responsabile for anything the software might do that brings down your house, empties your bank account etc....
If I remember correctly it was concluded that it was not clearly the autonomous driving software or sensors fault for the incident?
It was dark, the guy didn't have any reflexes on his bike or clothes, wore black clothing and the car only detected the person 2.5 seconds before the crash and only reacted about 1 second before the crash (which is possibly even faster than a human would have reacted).
IMO until we have a case where it's clear that we have a case where the fault was clearly on the software/sensors we should not hinder autonomous driving.
Is it
Gren-witch
or is it
Gren-itch
If you have a guess, by all means, express yourself.
We must forge ahead with this brave technology!
After all, the guy totally Darwin-ed himself by walking into the road in front of a car!
Robot cars are the future! Homeless people suck balls!
The problem they found though was pretty bloody serious. A driver that is completely distracted and a vehicle not capable of picking up what should be an easy spot for an automated system. bugs are fine, these seem like pretty serious design flaws that need rethinking.
Alpha.... obviously you don't do software or math (or greek)....sheesh.
that is.. self driving cars killing people...
when they invited uber to go there after california, et. al. told uber to fuck off (read: had regulatory requirements uber didn't want to meet).
the state of arizona is just as liable for this 'accident' as uber and its 'operator' are.
I think when you find a bug like this, you need to halt testing fix the issue (or attempt a fix) before resuming testing.
It is not the states job to put their citizens at risk to benefit a technology company and it's shareholders, in fact they may be liable for a civil suit (if they are not already) if someone was injured or died. As of course Uber now is, I imagine the family of the person who died is getting condolence calls from lawyers who are trying to make them feel better by taking Uber to the cleaners.
It might be valuable to put some sort of basic license test for self driving cars before they take to the road. Things like detecting and avoiding pedestrians could be a helpful start. I am sure the military prefer to test weapons before they kill the wrong people in a conflict, testing does not start with putting the weapon into the heat of battle.
You can't handle the truth! - Because I don't post left all my comments get modded down, bye bye Karma.
When did we become such cowards over a little death? We let our mining companies frack our water and turn our kids into mercury addled morons. We tested posilac on the general public with little concern about how it leeches into the milk we drink. We KNOW roundup causes cancer and I don't see anyone putting down thier forks. When we sprayed agent orange on school children in a pool they laughed their heads off, heck we even let loose nerve gas on sailors from '66 to '68 and nobody had a problem with it. Hell we used live human beings to do the tuskegee experiments! So what if a few people get crushed by a 500lb steel and rubber death machine once in awhile, that is progress dammit! Did you really think we would get self driving cars without a few deaths? Go back to your fantasy land you hippie communists! I bet your expecting flying cars without decapitations as well, bloody dreamers. Same self serving buggers will probably be crying in their milk the first time an AI pulls a loved one in half, but thats how we get a decent god dammned roomba that cleans in the corners!
Surely Governor Ducey is not going to be a hypocrite, particularly when lives are at stake: "Arizona must take action now!"
Surely there must be some criminal negligence here also, on part of the safety driver and also Uber.
Have they confirmed just what the safety driver was doing in that video? I imagine shouldnâ(TM)t be too hard to find out if the phone was sending or receiving messages at that time stamped moment.
Is it legal to operate a mobile device whilst driving in Arizona? Is safety driver considered driving if theyâ(TM)re not doing anything?
You sure that wasnt a sly pun?
the problem they found was that the cars wouldn't force the driver to take over if their systems are fucked.
lidar didn't work. apparently the driver wasn't aware of that it wasn't working for whatever reason.
and look, it's pretty easy to make a test track that should have exposed this problem long before driving on public roads.
world was created 5 seconds before this post as it is.
Because the driver in this instance is a bogus legal device to pretend that self driving cars running self driving tests are somehow not doing the driving.
He can only judge the cars choices by the outcome of those choices and that would be too late to intervene. He is not making the driving decisions, he is not driving the car.
Instead of testing self driving cars until they are safe to unlease on public roads, they let car makers get away with this bogus legal device.
More troubling in this case is the doctored video they released. A normal car recorder video in the dark, would have the gain turned up, and would be bright but grainy. That video was dark, suggesting it had been darkened. I ask again, did the police obtain the video from the car recorder themselves, or did they ask Uber's technical assistence, because the video shouldn't be dark. Especially darker around the edges which suggests intentional vignetting.
I guess you're not aware that the governor of Arizona is a Republican? I don't know much about him, but I'd presume that means he leans conservative, not liberal. Self-driving cars are being tested in California as well, and now perhaps you see why you can't always trust a corporation to self-regulate in the absence of government oversight. This is, sadly, how government regulations tend to come about.
The pedestrian was technically in the wrong, but we've heard a lot of rumors recently that Uber's self-driving software was being pushed forward recklessly. And given the wonderful people at Uber and their stellar track record, this isn't exactly hard to believe. A competently programmed car with a properly functioning lider probably should have seen that woman on the bike and reacted to it by braking far earlier than it did.
Yes, deaths will inevitably occur, but let's at least try to make sure there are as few as possibly going forward. This is a good reminder that machines can be just as fallible as the humans who create them.
Irony: Agile development has too much intertia to be abandoned now.
Where are the liberals here?
Alpha may be positive to you. In software terms pre-alpha is like v0.1, something that may produce some output, but will fail horribly at many common cases and lack crucial features making it even sellable. v0.1 / pre-alpha is not even something Google Beta.
When stakes are low, you are free to experiment, with as small userbase as possible first and minimizing impact of mistakes in many other ways. Unfortunately, people cargo cult and apply it to everything. If you do that, then you're no longer running a business. Yes, a pun.
California caved not long after Arizona let Uber more there and SF gave them a fucking license to operate as well.
I personally believe these autonomous cars can be made to work today, just not at the price point these guys are pushing. They need a two or three blade rack of high end servers with a ton of memory, a couple gpus or special purpose machine learning boards each, and sufficient cameras to cover all four sides of the car, plus LIDAR, plus hard brake if anything gets within 300 feet, or 2 car lengths per 10mph in front of them.
This might seem excessive, but like training a young person to drive, you need to give them wider margins than you tend to practice in the real world, and once they can prove they can avoid accidents with those restrictions, you can start letting them drive like the 'real world' does. But until that point they will be dangerous and reckless and screw up in the sort of life and death situations that will make things messy, as happened with this uber car due to both the failures of its autonomous systems to detect an obstacle/pedestrian, as well as the operator's inattentiveness to the road.
Since the driver was unable to detect this incident too, they better remove all drivers as well!
Change is certain; progress is not obligatory.
What is that f*cking banner ad on slashdot that does not go away when I scroll down? Already I hardly visit slashdot, now this infuriating banner ad that does not go away might totally make me never visit slashdot again. Slashdot, if you are reading this, get rid of it already or atleast make it non floating over the content.
That puts you at the top of your list, no?
SAE Level 3 automation should be illegal. Period. The backup driver simply cannot be expected to go from "no interaction with the vehicle for 1500 miles on average" to "rescuing the vehicle from an emergency".
Automation should be locked at Level 2 (ProPilot, Autopilot, Supercruise, etc, etc) until vehicles are at least ready for Level 4, if not Level 5. Level 2 = hands on the wheel, ideally with eye tracking, ideally making the user drive for themselves for at least a couple minutes every hour to stay alert.
And it should not be up to companies when their vehicles are ready to put them on the roads, as they're in a race to be seen as first movers. Governments should have their own review and testing processes, which involve both code audits (in the case of neural nets, audits of the net core and how the nets are trained) and real-world testing of simulated hazard scenarios.
"99 dead duelists of Dios on the wall. 99 dead duelists of Dios! Take one's ring, pass it around..."
How dare they violate ubers 1st amendment rights (or is it 2nd?) to test however they want.
Regulation kills!!
Level 3 is fine if there is plenty of warning to take over, say a minimum of 30 seconds. Time to put away your phone, pack away the sandwich and slip your shoes back on.
Unfortunately what we have is crap like this from Audi: https://youtu.be/WsiUwq_M8lE
Note the way it disengages suddenly and the guy has to instantly take over. That's dangerous and unacceptable.
const int one = 65536; (Silvermoon, Texture.cs)
SJW, n: "Someone I don't like, and by the way I'm a fuckwit" - AC
Level 3 is fine if there is plenty of warning to take over, say a minimum of 30 seconds.
In other words level 3 is never fine.
But that more or less gets you to Level 4, which IIRC is just Level 3 + pull over safely if something goes wrong.
Socialism: a lie told by totalitarians and believed by fools.
Sweet
Test tracks are useless here. That's the problem with machine learning - if you test too much in the same environment, the algorithms will "memorize" the test data and still be useless in the real world.
What these companies need to do is put all the needed sensors on millions of cars and collect the data, without actually being self-driving. Problem is, what consumer would pay for that?
Socialism: a lie told by totalitarians and believed by fools.
Having a human that can get bored babysitting automation is fundamentally the wrong way to do it anyway. Until the automation is reliable to not need human supervision, it should be the other way around: the automation monitoring the human to provide an extra safety net, so that the human has to be actively driving and alert rather than getting bored and not monitoring the automation properly.
Oolite: Elite-like game. For Mac, Linux and Windows
More importantly, the systems should be required to be certified througout for have a functionnal safety SIL level 3 or 4.
aaaaaaa
What's your proposed model for testing an autonomous car driving amidst normal traffic conditions that does not include actually having it drive among normal road traffic?
Secondly measured in terms of accidents and fatalities, autonomous cars have already caused less accidents per miles driven than your average human driver, so if that's your metric, the argument can be made that said bar has already been passed, although that obviously does not mean that the safety cannot and should not be further improved until the fatalities drop to zero.
We should also know all know that alpha or not, there's no bug-free software. You can do all the simulations and all the testing you want, bugs and accidents will still happen. However, once they do and are fixed, the vehicles will not do the same mistakes again, which is not true for most human drivers. Human drivers also do not learn from the mistakes of other human drivers that they've never met. Autonomous cars do.
I fail to see what the problem is here. We all knew this was going to happen because it always happens with new transportation technologies that are in the process of being perfected. Plane safety has gone up dramatically as a result of tens and hundreds of accidents in both software and hardware, and the same's true with 'regular' cars. Your standard seems to me to be an illogical 'unless it's completely failsafe it should not be used at all', and if we followed that principle we'd still be using horses.
"It is the business of the future to be dangerous" -Alfred North Whitehead
Flash a dim light on the windscreen at random intervals, and ensure the human responds. No mobile phone usage then. This sort of thing has been done for trains for ages (I'm not talking about dead man, but attention monitors.)
Uber did not care. And that video they released was dubious, someone else took a dash cam of the area and it was reasonably well lit, even if there was no Lidar.
Uber were totally negligent. And I suspect they did not pay that driver very much. Monitoring a test car should not be a minimum wage job.
proposed model for testing an autonomous car driving amidst normal traffic conditions that does not include actually having it drive among normal road traffic?
Hire people operate vehicles on closed tracks to simulate traffic and specific scenarios.
Or, better yet, orchesate dozens of other self driving cars to move in fixed choreographed patterns that the autonomous vehicle must recognize and react appropriately to.
Thank you for your well research and thoughtful comment.
What's your proposed model for testing an autonomous car driving amidst normal traffic conditions that does not include actually having it drive among normal road traffic?
Perhaps they should focus first on not killing pedestrians on an otherwise empty road, and worry about "normal traffic conditions" later.
If autonomous cars could drive as well as humans, they wouldn't need testing. I would also like to know how you test against real world conditions without real world conditions.
30s? LOLOLOLOL
If there is traffic condition or emergency that gives you 30s to decide or react, then it's not an emergency.
autonomous cars have already caused less accidents per miles driven than your average human driver
Statistics are like bikinis. What they show is not anything special. What they conceal is vital.
Self driving cars only operate in a very subset of extremely limited conditions. When you compare like for like, limit the stats on human drivers to the same limited conditions that self driving cars, do you sell that have killed (3 people killed so far) and been the cause of far more accidents than human drivers.
we followed that principle we'd still be using horses.
We certainly would not! Horses are quite dangerous. Some statistics imply about 20X the risk of motorcycle riding per passenger hour.
How about just linking to the alternate? Then no need for your favorite phrase of caution.
Perhaps they should focus first on not killing pedestrians on an otherwise empty road, and worry about "normal traffic conditions" later.
And will you apply the same standard to human drivers, who kill 270000 pedestrians per year on average?
265,000 of them got themselves killed due to failing to have common sense and possessing too much arrogance about owning the road.
With level 4 you could potentially have someone unqualified to drive take a pre-planned trip, as long as the whole trip was on roads that the car can handle. That could be most roads, in fact, with a few exceptions for things like driving on to ferries, unpaved roads, certain types of car park etc.
With level 3 you would always need a qualified driver behind the wheel. Say Audi did their system properly, it would never disengage suddenly in traffic but you would have to be prepared to take over when traffic cleared or a change of lane was required (for navigation), or when roadworks were coming up etc. A limit of 37 MPH is probably a bit high for that and it would need to alert you the moment the traffic started to flow again, to give sufficient time.
const int one = 65536; (Silvermoon, Texture.cs)
SJW, n: "Someone I don't like, and by the way I'm a fuckwit" - AC
Shadow operations. Don’t control anything, just observe. If software action deviates from human action, review and analyize. It appears that this is the stage in product development that Uber needs to be in.
Once the shadow driving has advanced to a point where there are no negative deviations in a control environment, define that envelope and test rigorously within the envelope. Continue to shadow outside the envelope and slowly expand the envelope. Within test envelope, thoroughly validate performance with external telemetry— were there any cases where nothing bad happened, but the action taken should have been different.
The safety driver?
The lidar mechanic?
The programmer?
Uber? (after all, corporations are people too.)
Careful, I think we've got one of those "I disagree with you, therefore you are the object of my ire, ergo liberal."
Really, people are being dumb. You can't blame the automation for this. No human could have missed her. This is 100% the idiot pedestrian's fault.
She wasn't using lights on her bike like she should have been.
She wasn't wearing reflective tape/garments like she should have been.
She was wearing dark cloths at night.
If anyone with half a mind watches the video they can see that there is nothing to see until it's too late.
The fact that she walked in front of a moving car pretty much says it all. I don't get how anyone here can be blaming automation for this.
proposed model for testing an autonomous car driving amidst normal traffic conditions that does not include actually having it drive among normal road traffic?
Hire people operate vehicles on closed tracks to simulate traffic and specific scenarios.
Or, better yet, orchesate dozens of other self driving cars to move in fixed choreographed patterns that the autonomous vehicle must recognize and react appropriately to.
I'd be interested to see just what has been done by various AV developers with regard to comtrolled scenario testing. Obviously you can't re-create every possible scenario, but you could create hundreds of very likely ones. A person crossing the road in various manners would be part of that testing regimen. Seeing what happens with certain sensors failed might be another. It is certainly something worthy of a question. I assume much has been done but I'd love to see the checklist(s).
At some point though, they need to be out on the roads and get the real world testing.
To me this incident illustrates just how hard it really will be to get fully AV on the road any time soon. I know many are enthusiastic about it, but its still at least 10 years away, IMO, except maybe for very limited functions such as certain defined truck routes.
Which is exactly what happened in this case.
..unless AI can read the future.
Laws are rules for the court, but merely a bottom bar to hit for life. Think beyond laws in your actions always.
ANY DAY NOW were all going to be bitbcoin billionaires in our self-flying taxis. Generation Pleb says so!
One of the most frequent minor accident scenarios in freeway driving is when stop-and-go congested traffic comes to a sudden stop. Often we get a series of rear-end collisions at this point after drivers' attention has drifted. Does Audi have any tests on how traffic behaves when all of the cars in a congested area are on Level 3 autopilot?
Nope, horses are dangerous. People get thrown all the time. And die as a result.
If we followed that principle, we'd still be walking everywhere. Well, probably not in the Americas, since people had to cross open water to get here, and ships are really dangerous....
"I do not agree with what you say, but I will defend to the death your right to say it"
Perhaps they should focus first on not killing pedestrians on an otherwise empty road, and worry about "normal traffic conditions" later.
And will you apply the same standard to human drivers, who kill 270000 pedestrians per year on average?
No. We can't and we shouldn't. We have training and testing for human drivers. Statistically, human driving is very safe and there are only marginal ways to improve human driving. AV driving is new and in testing. We need to learn more. There is no sense in moving to autonomous if it is not an improvement, or if there are serious flaws like running over pedestrians that can be seen well in advance, or running into the side of tractor trailers.
Why should a consumer pay for that? The company is doing R&D, they should fork over the cash for the sensors and what not. Consumer should probably pay even less than for a normal car.
Can training and/or testing for fully autonomous vehicles be done in a simulation? if not, why not? You could even add pedestrian and vehicular NPCs and even human-controlled PCs to the mix. If running NPCs, you can run the simulations at faster than 1x speed and get improvements and results faster than real life.
For that matter, why is this not being done for kids learning to drive?
Isn't Arizona, like, a barren wasteland or something? I mean who cares?
One problem with Level 3 testing with a driver is that, such as in this case, if the driver had stopped the car the flaw might never have been fully exposed.
There are several different self-drive systems in beta in Arizona. The suspension only applies to Uber, and because there appears to something strikingly wrong with its implementation. The car should have seen a slow-walking pedestrian in a well lit area.
What's your proposed model for testing an autonomous car driving amidst normal traffic conditions that does not include actually having it drive among normal road traffic?
Secondly measured in terms of accidents and fatalities, autonomous cars have already caused less accidents per miles driven than your average human driver, so if that's your metric, the argument can be made that said bar has already been passed
There are entire test sites for self-driving cars. Once your car meets some standard in the test site, you could then take it on the road. https://www.freep.com/story/mo... Autonomous vehicles, in general, may be safer than ones operated by the average human driver. But that doesn't mean that the *Uber* vehicles are safer than ones operated by human drivers.
It appears that Waymo and such have done a great job getting their vehicles as ready as possible before putting them on the road and, as a result, have a great safety record. Uber just slapped a bunch of components together probably using technology that they "borrowed" from competitor without understanding, put it on the road, and watched the failures pile up. Their failure rate is one per 13 miles vs Gogole's one per 5k miles. Now there are some thing that make those not a perfect comparison but even giving Uber an order-of-magnitude benefit of the doubt, their cars are still awful dangerous things that don't belong on the road. That's different than saying no autonomous cars belong on the road.
"We all knew this was going to happen because it always happens with new transportation technologies that are in the process of being perfected."
So people dying during testing is considered normal??? What kind of a fucking douche bag are you....
For the brain damaged out there, no people dying during testing is not considered normal. As others have pointed out, testing is done in controlled conditions before they are allowed in live environments. THIS WAS NOT DONE.
This kind of scenario is a rather fucking simple one to test!!
So fuck off and die.
A better solution is to do like airplanes and allow SAE3 Level automation but require *two* human backup drivers.. One to continuously monitor the self-driving mechanisms for things like are the lane lines being detected properly, are all objects being sensed, et cetera. This person should be constantly communicating with the one who can actually operate the controls. If the two are constantly engaged in evaluating the situation and confirming with each other verbally that they are assessing things the same, they won't be going from "cold" to fully engaged.
Sorry ass-hole, responsibility is on the driver. Read the traffic laws.
This is not a person who suddenly jumped out onto the road here.... while she was jaywalking, she was also *WALKING*... I've seen an overhead view of the section of road where the incident occurred, and there's no significant occlusions there; ordinarily, vision seems that it would be pretty good there in daylight conditions. It's my understanding that self-driving cars use lidar sensors, and should even be able to detect a person in an absence of any visible light at all, so the fact that it was night should be immaterial. Reasonably, the car should have seen that she was on the sidewalk long before she stepped out into the road, and the very *instant* that she started to go off of the sidewalk should have been detected by the car, and the car should immediately begin to slow down.
Yet, by all reports I've heard, the car did not even see this pedestrian at all, and had not even tried to slow down until after the collision. Why? What the fuck happened?
File under 'M' for 'Manic ranting'
Oh it's worse than that. These algorithms aren't ready; all this testing leads to tuning and adjustment of the code.
Testing them in a closed environment will build a learning algorithm more-refined to work on the fundamental assumptions of that closed environment.
Support my political activism on Patreon.
Found the Uber shill
How the hell do you"know" that?
Fuckwad
What's your proposed model for testing an autonomous car driving amidst normal traffic conditions that does not include actually having it drive among normal road traffic?
Simulations, no reason someone can't simply drive around recording sensor data which you then feed to your driving algorithms later to see if it behaves properly.
We should also know all know that alpha or not, there's no bug-free software. You can do all the simulations and all the testing you want, bugs and accidents will still happen. However, once they do and are fixed, the vehicles will not do the same mistakes again, which is not true for most human drivers. Human drivers also do not learn from the mistakes of other human drivers that they've never met. Autonomous cars do.
We've seen numbers that Uber's software needed human intervention every 13 mi, Google's on the other hand is 5600 mi. If we consider how far people drive Google's still needs help a few times a year, Uber's needs it a few times an hour.
They need two safety drivers in each test vehicle.
Now you're not even safe on the sidewalk!
Confucius say, "Find worm in apple - bad. Find half a worm - worse."
SAE Level 3 automation should be illegal. Period. The backup driver simply cannot be expected to go from "no interaction with the vehicle for 1500 miles on average" to "rescuing the vehicle from an emergency".
I agree for normal usage. For testing however it should be possible, however the driver should probably be an engineer working on the program, they're much more likely to pay attention and understand what the system needs.
What's your proposed model for testing an autonomous car driving amidst normal traffic conditions that does not include actually having it drive among normal road traffic?
You seriously have to ask this question? Why are you even on this site?
A corporation killed someone - at the absolute minimum their entire board should be charged with manslaughter and the company shut down.
Putting the autonomous car on the road can be justified when the car doesn't need human supervision
Yeah why don't we hold humans to that standard, no supervised driving on public roads until you have your driver's license. You can have a professional driving instructor giving you reasonable challenges building up your skill step by step making sure to catch your mistakes or you can have a teen with ADD and a dad who'll throw you out there to either sink or swim with a legal babysitter. And the last guy just mowed down somebody, clearly supervised driving on public roads is the problem... not. Uber is the problem and they were probably lucky they weren't more obviously to blame, because they're simply a reckless company in everything they do. Waymo has driven 5+ million miles now without fucking up anywhere near this bad, clearly it's possible if your priorities are right.
Live today, because you never know what tomorrow brings
Test tracks are useless here. That's the problem with machine learning - if you test too much in the same environment, the algorithms will "memorize" the test data and still be useless in the real world.
What these companies need to do is put all the needed sensors on millions of cars and collect the data, without actually being self-driving. Problem is, what consumer would pay for that?
You are aware that test tracks can be modified for different scenarios, yes?
Oh, and why would the consumer pay to put sensors on their car. If Uber wants to get these on the road, then they should pay for the testing.
The AC is right. You area shill.
What's your proposed model for testing an autonomous car driving amidst normal traffic conditions that does not include actually having it drive among normal road traffic?
Drive around in real traffic with a human driver in a car kitted out with all the sensors that the autonomous car would have. Record the sensor data. Use that sensor data to build a simulation to train the AI. I know an actual solution would be more complicated, but it could be done. It's just cheaper to put real cars on real roads and endanger real people.
Secondly measured in terms of accidents and fatalities, autonomous cars have already caused less accidents per miles driven than your average human driver, so if that's your metric, the argument can be made that said bar has already been passed, although that obviously does not mean that the safety cannot and should not be further improved until the fatalities drop to zero.
You can't just lump all autonomous vehicles together under one rubric when the issue is whether Uber should be allowed to continue road testing. Uber has a significantly worse safety record that its peers when taken their statistics are considered in isolation. Their technology is much less mature with more human interventions required per mile, creating more mode switches for the human operator and more opportunities for failure.
It might also explain why we haven't seen a self driving Alfa yet. It's stuck in early testing :)
Good plan, I support your promotion to head of Uber's driving department.
"First they came for the slanderers and i said nothing."
We do.
We don't allow them on the road without a qualified driving teacher before they demonstrate that they are out of "beta stage".
And we don't even allow them on the road WITH a driving teacher before demonstrating that they aren't blind.
Judging from the video, this car was so blind that it didn't even realize it didn't have enough light to see where it was going, and the person in the car wasn't a qualified driving teacher.
Exactly, this was quite a basic scenario. Not to mention total ignorance by the operator, who was there to deal with dangerous situations.
1. Test software in countless virtual reality based on real data.
2. Build a semi-virtual room with display walls, ceiling and floor and test the software and sensors there with countless hours of real recordings (it will test only passive sensors in visual spectrum - but it's a start)
In both cases the testing recordings cannot be the training one
3. Only then take the car to the road, but with a human driver and working computer in a passive mode - to check its reactions, any event when computer makes a decision after human or makes a wrong decision go to point 1.
4. Let independent authority test the car in environment as point 3.
5. Let the car drive in closed safe environments.
6. Let the car drive on public roads not with an operator but with a professional driver NOT watching his cellphone or tablet.
It seems obvious, but after the accident it does not seem like it was obvious for everybody.
Evidently, you are the butthurt libturd n1gger lesbian at the top of the list.
this is the computer equivalent of that.
Hi! I make Firefox Plug-ins. Check 'em out @ https://addons.mozilla.org/en-US/firefox/addon/youtube-mp3-podcaster/
Why? What the fuck happened?
Obvious. Uber rushed their half-baked "autonomous" car onto city streets without the proper testing.
Level 3 is fine if there is plenty of warning to take over, say a minimum of 30 seconds.
Was that a brain freeze? 30 seconds, if your going at 60 km/h, it's 500 meters. If you know 500 meters ahead that there's an issue, it's not an emergency. Emergency is when you're doing 120 on the highway and the car 20 meters in front of you brakes hard. Emergency is when someone cuts you off at the last second. 30 seconds of warning, isn't a challenge even for novice pilots...
What's your proposed model for testing an autonomous car driving amidst normal traffic conditions that does not include actually having it drive among normal road traffic?
I thought Uber had an entire town set up for this sort of thing. On a related note, someone should develop mechanized crashtest dummies that can be programmed to walk into the path of an autonomous car; when the cars are capable of avoiding people who are actively trying to get hit, then (and only then) have they achieved enough superiority over humans that we can justify (maybe and only maybe) thinking about sharing our roads with them
That's a dangerous slope.
He didn't watch well so another Hawtch-Hawtcher had to come in as a watch-watcher-watcher! And now all the Hawtchers who live in Hawtch-Hawtch are watching on watch watcher watchering watch...
He's getting rather old, but he's a good mouse.
The traffic condition could be as simple as "exit ramp a mile away".
There's plenty of the country where a car that allowed for totally distracted driving more than one mile from an exit ramp would be awesome.
Yeah. Makes sense.
Damping absorbs vibrations. Dampening is caused by moisture.
One obvious problem with your "logic" is that while the SAE levels may have been the "best guesstimate" of the engineers involved, without any commercial or pro-technology bias (dubious), it wasn't and isn't based on what the real competencies and INcompetencies of the technology - once they are (note the plural) developed. It is like a 6 or 8 year old setting up criteria for a (romantic) boy-friend/girl-friend for themselves when they are 30 (and sexually and romantically experienced). The SAE ought to be embarrassed but it seems they think that any publicity is good publicity. Anyway...Uber's technology clearly failed and failed in multiple ways. If they didn't know, 1 day after they killed that woman, WHY their device killed that woman, the obvious conclusion is they are NOT approaching the experiment with any professionalism. In R&D the optimum experiments tell you just as much (or more) when they fail as when they succeed. If they weren't monitoring the "interior" state of their equipment (including the software, but especially the sensors and response system) then they were playing at actual professional conduct in a Cargo Cult Science (Programing?) mode.
Everything you say is mostly true. However, I would think that autonomous vehicles should pass some type of 'safety' course before being allowed on public roads. Same as a human driver.
Yet we don't really know for sure if anyone else is better than Uber, just that the flaw isn't showing up in testing. Whether that testing is designed to expose that flaw if it exists, we don't know.
Laws are rules for the court, but merely a bottom bar to hit for life. Think beyond laws in your actions always.
Frist problem they found - the backup driver wasn't engaged.
Since this 'driver' was being monitored, albeit perhaps incidentally to video recording, they will be going back and determining how many other drivers were not 'engaged' during testing.
Fix that, and they we can talk about software, sensors, and features of the autonomous mode vehicle...
deleting the extra space after periods so i can stay relevant, yeah.
What's your proposed model for testing an autonomous car driving amidst normal traffic conditions that does not include actually having it drive among normal road traffic?
As far as I've read, Google (Waymo) created an entire fake town to test different traffic situations out before letting their autonomous cars onto the actual roads in actual towns...I have not heard of Uber doing such a thing.
Secondly measured in terms of accidents and fatalities, autonomous cars have already caused less accidents per miles driven than your average human driver, so if that's your metric, the argument can be made that said bar has already been passed, although that obviously does not mean that the safety cannot and should not be further improved until the fatalities drop to zero.
This is a completely irrelevant statistic at this point. You're comparing accidents per miles driven of regular vs. experimental self-driving cars (all of which, btw, have human drivers which are supposed to - and often do - intervene to prevent accidents)...the two "sample sizes" so to speak are so vastly different that no valid comparison is possible. How many miles have self-driven cars driven? Several orders of magnitudes less than regular cars. How many self-driving cars are on the roads at any point? Tens? Hundreds? On the other hand there are hundreds of millions of cars in the US alone, which means tens of millions of cars are on the road every day, i.e. at least a couple of million driving at any one moment. Those regular cars drive in all conditions, on all roads, during all times of day, in every imaginable traffic situation, and, apart from beginner drivers riding around with instructors (who have a second brake to use at the passenger seat), none of those drivers are being constantly supervised by someone who can take control of the vehicle if they make a mistake. So comparing their accident record to self-driving cars which are being tested in a handful of cities and which, by admition of the manufacturers of said self-driving cars, cannot yet handle all weather conditions for example, is just...nonsense.
Bottom line: put 5 million self-driving cars on the road (using current technology), let them drive randomly around (i.e. no careful selection of driving days based on weather, routes based on street suitability for current technology, etc.), remove the human driver supervising them, and I'm pretty sure that accident rate will be higher than with normal cars.
Finally, the link you provided talks only about Google cars. No data about Uber or the other players in this game.
Plane safety has gone up dramatically as a result of tens and hundreds of accidents in both software and hardware, and the same's true with 'regular' cars. Your standard seems to me to be an illogical 'unless it's completely failsafe it should not be used at all', and if we followed that principle we'd still be using horses.
You're missing a crucial point here. When cars first came about, many cities would not allow them...many others allowed them if a person would walk (yes, walk) in front with a red flag, a horn, or shouting "danger!" or something like that (yes, this is true...). Early cars were, in modern terms, painfully slow - not just due to lack of engine power but also due to lack of infrastructure (no paved roads) and regulation ("jaywalking" had to be invented as a crime to allow car traffic to move more smoothly and more safely, on the back of a lot of lobbying by the car companies). There was a gradual ramp up of cars and the necessary infrastructure to support them. Similar things happen with airplanes - airports were placed away from built up areas, flight routes restricted, and this meant that potential harm from a failure was limited to the passengers - whom no one had forced to fly. Testing of self-driving cars on infrastructure not built for them (but built for huma
I heard that Alfa is back, in Pog form!
I can imagine some lawmaker somewhere declaring a halt to driverless cars after this accident.
I have already several articles suggesting that this should not be done because only more and more refinement of such a complex product will cause it to become viable. Also even with a few bugs, driverless cars are possibly already less accident-prone than humans.
As a software developer, I naturally side with continuing development.
Looking at the FAA gives a good model on how to proceed.
When an airplane crashes, the FAA sometimes grounds all models of that plane until the cause of the crash is determined and, if it was a technology error, will not allow the planes to fly again until the problem is satisfactorily resolved.
That would appear to be a measured response to this type of problem.
Don't halt all development. Don't proceed, ignoring the death(s).
Prohibit the specific driverless system from using the public roads until the problem is determined and an acceptable fix is made.
Just as cars have model years that receive approval, so should specific versions of driverless systems.
Then we can have official patches deployed on an as-needed basis, not just when a software engineer declares a bug has been fixed.
Very strict controls need to be in place to allow/deny a software/hardware update to a driverless system.
I don't want my car to be hacked and used as a killer weapon.
That link (the one that supposedly shows autonomous cars are safer) has a date of January 2016. Any data studied would obviously have to be older than that. Now, exactly how many fully autonomous cars were there, driving in real-world conditions, more than 2.5 years ago? Maybe 0?
Let's not pretend they are even close to as safe as human driven vehicles. If you look at the real metrics, cars killed 1.18 people per 100 million miles. Autonomous vehicles are on track to do much worse, but we don't have clear statistics yet.
I'm also not confident that software engineers will have the same level of respect for life that other engineers have shown.
Cheap storage VM.
Your response has nothing to do with the question.... I asked why the car didn't see her. By all rights, it should have, so why didn't it? If they don't know yet, what are they doing to find out?
Was it a sensor malfunction? Did the software in the car fail to recognize her as something on the road? If not, what did it see her as? Can this situation be recreated in simulation to figure out what the car did wrong, and corrected in the future so that there can be some assurance it won't happen again if or when autonomous cars are allowed again?
File under 'M' for 'Manic ranting'
Testing is also done (or should be done) in controlled environments until you get way past the alfa and beta stages. Putting the autonomous car on the road can be justified when the car doesn't need human supervision and it can deal with normal traffic conditions in day and night with the same performance as that of a human driver.
I seriously have no idea why autonomous cars in pre-alfa stage are on the roads.
Unfortunately most people are treating autonomous cars as software. And we know how software engineers think. Throw the alfa software to the public and fix mistakes afterwards. Oh and we're not responsabile for anything the software might do that brings down your house, empties your bank account etc....
Speaking of testing in a controlled environment. Run that video in front of 100 human drivers in a simulator with some variable amount of the video taken in the few minutes before hand with a control group that cuts out a few seconds before the collision and see if they can react in time to avoid the collision. I would be willing to bet that many won't be able to react in time to avoid that collision either.
To be realistic give a few of them a beer, make a few not get enough sleep the night before and choose a group of people representative of the old, the young and people with poor driving records as well as your perfect drivers without a collision or speeding ticket.
Even the ideal driver expecting a collision would have had trouble with that one.
I seriously have no idea why autonomous cars in pre-alfa stage are on the roads.
They're SUPPOSED to be safe with a single safety driver being present to takeover the moment the self-driver fails to react or does something wrong. That's a bad approach because of how a person fatigues when they're JUST WATCHING for long periods of the time.
There may be other viable solutions but the one-safety-driver solution doesn't work.... maybe they'd be better off with 2 safety drivers holding a big red "Stop Now" button.
Fuck you Uber!!
Governments should have their own review and testing processes, which involve both code audits (in the case of neural nets, audits of the net core and how the nets are trained) and real-world testing of simulated hazard scenarios.
Government already has a test for driver readiness. It involves a guy in the passenger seat telling a teenage kid to perform some simple maneuvers on some side street while following the rules of the road and making notes on a notepad while nodding or shaking his head... then hopefully that kid only gets into a few accidents before he or she actually learns how to drive. And re-certifications require showing up to prove you are still alive.
I think before you start idealizing human drivers and raising the bar impossibly high you should have a proper control group of human drivers to compare against. Because right now it appears to me that human drivers are not a heck of a lot better than a brick glued to the gas peddle and rope tied on the steering wheel.
"What's your proposed model for testing an autonomous car driving amidst normal traffic conditions that does not include actually having it drive among normal road traffic?"
The obvious answer is to install the hardware and software on vehicles that are driven by humans performing normal driving and not to enable the autonomous control of the vehicle until millions of miles are accumulated with no false negatives (failure to detect a problem) and minimal false positives. In practice, that's not going to work perfectly. But it's clearly not what Uber is doing.
You can't see ANYTHING from a car, You've got to get out of the goddamned contraption and walk...Edward Abbey
By the way, you're drooling...
Yeah, that's not the law or the reality. But if it makes you feel safer in a scary world that is beyond your control, go on believing it.
responsibility is on the driver. Read the traffic laws.
There was no driver: my arguments
1. No eyes on road.
2. No hands on wheel
3. No feet on pedals.
LITERALLY no driver in the car.
In response to the request to have all of our autonomous vehicles removed from public roadways:
1. A request has been made to have them all removed.
2. Except for Christine. We've lost contact with Christine. And we really don't know where this car is.
I think that was Waymo that set up the town.
"Failure" is defined as the backup driver having to take over. That says nothing about the overall safety of the car (and driver) on the road, just that the cars are not ready to be left on their own.
Google's relative caution may have less to do with caring about public safety and more about fearing public and regulatory backlash.
Yes, people dying is part of the calculus in testing new vehicles just as it is in testing new drugs. Taking no risk results in far more deaths and only a whiny name calling child thinks otherwise.
Oh dear, it killed a pedestrian, how inconvenient, that'll play hell with our testing schedule!
WTF, are we China now? Human life is cheap, throw-away? Someone being KILLED can't stand in the way of progress? Fuck that. Good on you, Arizona, ACTUALLY looking out for public safety.
It's just cheaper to put real cars on real roads and endanger real people.
That's all that matters, really. If it would have taken $100 million in extra testing to save this one life, it would not have been worth it. When you also consider that such rigorous testing would delay the deployment of safe AVs (even if only by days), it makes real world testing even more imperative.
Thank you for injecting some actual common sense and decency into this conversation.
This is a completely irrelevant statistic at this point. You're comparing accidents per miles driven of regular vs. experimental self-driving cars...the two "sample sizes" so to speak are so vastly different that no valid comparison is possible.
You are quite mistaken. The comparison is fair in deciding whether AVs are safe enough for real world testing in limited environments, not whether they are ready to be left unsupervised in all conditions.
So it wouldn't matter if such a system saved thousands of lives if it has even a single failure that kills one?
It most certainly answered your question "Why didn't the car see her?"
Because this system was not properly tested on a closed course. Period.
Any failings of the sensors should be found and fixed before moving to public roads.
Stop being obtuse.
Secondly measured in terms of accidents and fatalities, autonomous cars have already caused less accidents per miles driven than your average human driver, so if that's your metric, the argument can be made that said bar has already been passed, although that obviously does not mean that the safety cannot and should not be further improved until the fatalities drop to zero.
Two points:
1) This comparison is of automated vehicles driving only in the safest conditions to humans driving in all conditions. That introduces a huge bias in favor of automated vehicles.
2) From your link:
Low exposure for self-driving vehicles (about 1.3 million miles in this study) increases the uncertainty in Self-Driving Car crash rates compared to the SHRP 2 NDS (over 34 million miles) and nearly 3 trillion vehicle miles driven nationally in 2013 (2,965,600,000,000). ...
Current data suggest that self-driving cars may have low rates of more-severe crashes (Level 1 and Level 2 crashes) when compared to national rates or to rates from naturalistic data sets, but there is currently too much uncertainty in self-driving rates to draw this conclusion with strong confidence.[Emphasis mine]
This isn't an apples-to-apples comparison. It's an apple-seed-to-orange-tree comparison, and should be taken with a whole lot of salt.
If the driver is having to take over every 13 miles, it means that Uber's "self-driving" system performs worse than the adaptive cruise control available on garden variety economy cars today. I have no idea what they're even testing at this point. The performance is abysmal. The safety of the "self-driving" car will be about the same as if there were no "self-driving" features at all since it will basically be under human control all the time.
Governor Ducey, you disappoint me. Everyone knows that regulation is Tey Bad, and now you've regulated. Ergo, you are bad. Bad Governor Ducey, bad!
We can all count on corporations to do the right thing. When they do the wrong thing, Nothing Can Be Done, so just live with it.
Level 3 is fine if there is plenty of warning to take over, say a minimum of 30 seconds.
Emergencies develop in a fraction of a second, not half a minute. If a human is needed, a human is needed NOW. But you are right that people can't realistically take over in such a small amount of time. That's why level 3 is not fine. There's really no reason why a car should need a human to take a car over in 30 seconds, so at that point, why even have that level of automation?
1) This comparison is of automated vehicles driving only in the safest conditions to humans driving in all conditions. That introduces a huge bias in favor of automated vehicles.
Fair point, but as a society we've decided to tolerate a certain amount of carnage to get where we need to go. IMO, as long as AV testing stays at about that level they are doing little harm. That human drivers put themselves in far more dangerous situations is not necessarily a point in their favor.
Except it's not 13 miles on a well marked highway. It's 13 miles driving at least partially in a complicated city environment, something the ACCs can't do. If that really means it could do 5 full Uber pickup/dropoffs without intervention that is pretty impressive technology. Not ready for primetime, but still unimaginable from what was possible 18 years ago.
You are making a good case that Level 3 will never prevent all traffic accidents. You are not making a case that Level 3 isn't useful or that it won't save lives.
Even the "safety driver" probably expected the vehicle to see this pedestrian. If the car is only expected to react at the last possible moment, then there is literally no time for the safety driver to evaluate and take action. As you are indicating, this points to a flaw in the entire concept of a safety driver. This is also why the driver is not necessarily at fault here. If the LIDAR were malfunctioning so badly that it did not see this pedestrian, that car should not have been self-driving at all. If the system failed to detect this internal fault, then it was homicidally unsafe even with a safety driver. Either someone made the decision to run this car in a broken state and they should go to jail, or Uber's self-driving car division should be put out of business because they are recklessly endangering the public. I think even Telsa autopilot (Level 2) would have avoided killing this person even if the driver wasn't paying attention. This appears to be exactly the kind of situation that auto-braking commercials show. Is that level 1?
We proactively suspended self-driving operations in all cities immediately following the tragic incident last week.
proactive - adjective
(of a person, policy, or action) creating or controlling a situation by causing something to happen rather than responding to it after it has happened.
I disagree, it IS a point in their favor. Life goes on even in bad weather, etc. The point of a car is to get you from point A to point B when you want to go. If the car can't/won't do that it is a failure. Now, certainly there are times when you just should not drive, but ordinary rain and snow are not those times.
Because Uber doesn't know what they're doing.
The car is supposed to have LIDAR sensors. So either Uber wasn't using them or the software failed catastrophically. (Or both.)
I'm sure they've been busy doctoring logs and sensor data.
Yes that is literally the use case for emergency braking assistance that has been a feature of upper market cars for many years.
Correction, you can't walk everywhere either: http://www.newsweek.com/apple-...
Until people learn how to walk, we should revoke their walking license.
Tesla buyers.
No, the question you answered was more along the lines of "how could this have happened?", a far more general question and has nothing to do with why the software in the car did not work in the expected manner.
That you'd find it more straightforward to suggest that I am being obtuse or asking an idiotic question and try to make yourself sound superior than it is to ignore a question that you clearly don't know the real answer to is quite beyond me.
File under 'M' for 'Manic ranting'
It depends whether the family is able to sue for more than $100 million in damages, doesn't it? Hard to put a price on what a life is worth.
Laws are rules for the court, but merely a bottom bar to hit for life. Think beyond laws in your actions always.
You're defending the comparison that allowed these Uber cars on the road? Who is regulating what environments these cars are tested in once they are on the road?
Laws are rules for the court, but merely a bottom bar to hit for life. Think beyond laws in your actions always.
To be clear what you are saying is that it's okay for people to double or triple the chance of killing someone because life goes on and they have shit to do. Got to get to your job at 7/11, got to pick up that pepperoni pizza, got to get to that Buffalo Bills game in the snow, ...
What you're saying is that driving is a binary choice. There are certain conditions that it is just absolutely wrong to drive, but for every other time, every other situation, every other driver then it's just fine. It's not binary.
What bugs me most is the complete conflict, the cognitive dissonance that this opinion entails (not necessarily by you, but you can see it all over this thread). People are okay with other drivers, sometimes terrible drivers, driving in poor conditions for dumb reasons, but just can't tolerate an AV on the road until it has been proven to be completely safe.
Nope, smart people do it all the time. It's only hard for people who are irrational or have an agenda.
It's self regulated by their performance, of course. This is actually how every human driver operates too. Start out on safe local streets in daylight and clear weather and keep expanding the test conditions if everything goes well. Next question.
I thought it was ARPA
I think we've pushed this "anyone can grow up to be president" thing too far.
The problem here is that analysis makes it clear that management was pushing things ahead even when problems were reported.
There are various things that should have happened if the car were working as designed, but which didn't happen. It didn't slow, it didn't attempt to dodge. It didn't use lidar (I believe that's built into this model of car). Etc. Possibly those wouldn't have prevented the accident, but they should have reduced it's impact. But more to the point, even if this is an edge case where they wouldn't have helped, they should have been tried, because they *might* have helped.
Management had been warned that things weren't progressing satisfactorily, and their response was to push ahead with testing in order to meet their deadline.
I'm not sure they're at fault in this accident. I not only haven't studied it, I'm not an expert in the appropriate fields. But it's clear that the reactions of the car were not acceptable. There's a large gap between "at fault" and "good performance".
I think we've pushed this "anyone can grow up to be president" thing too far.
If the software failed catastrophically, can't they figure out why it did?
It's a computer... they should be able to run simulations to figure out what went wrong, and then be able to to alter the software to account for the possibility so that at least there can be some assurance that it won't happen again.
File under 'M' for 'Manic ranting'
I see what you're saying. The governor knew pedestrians and bicyclists are more difficult to detect and avoid hitting, and because they are more likely to be liberal, he allowed self-driving cars to be tested in his state with few protections. Crafty!
Any sufficiently unpopular but cohesive argument is indistinguishable from trolling.
Which means that it needs to be able to handle those situations on it's own. I'm not sure that 30 seconds is sufficient, as I've seen people take longer than that to be able to disengage.
But there are cases where it would be extremely useful. The one mentioned above is "The car can handle freeway driving, but the exit you need is 30 seconds ahead, so get ready to take over.", and that seems like a quite plausible one with a multitude of use cases.
Also, I'm not clear as to the relevance of your mention of "novice pilots". I think you've got a use case in mind that isn't occurring to me.
I think we've pushed this "anyone can grow up to be president" thing too far.
They aren't useless, but you need a LOT of different ones. And actually, just saying "test track" is going to put your imagery in the wrong scenario. Testing mazes is a better image. Each maze designed by a different group, and required to be significantly different from all the others. And you need each different maze to come in a lot of variations. One of them, e.g., might have a random selection of man hole covers that could pop open in front of it, or might not. Another random intrusions from the side. Etc. Look at video games for inspiration of what might happen, the ones where people have to dodge. Also look at shopping malls. Etc.
No, it's not simple. But it should result in improved results. And be sure to redo old mazes occasionally to ensure that new learning doesn't override old.
I think we've pushed this "anyone can grow up to be president" thing too far.
Speak for yourself, that titty cleavage and booty cheeks is all I need.
Ah, 'double or triple the chance of killing someone'. You mean, in exactly the same way as you double your chances of winning the lottery by buying two tickets. Double the chances! It's practically a sure thing! Per your logic, the 'best' car (with the most points) is one that doesn't move at all.
By the way, the biggest cause of weather related accidents isn't snow, or ice, or even rain. It is wet roads. Do you know how often the roads are wet around here? Damn near every day. Guess we should all just be hermits so we are safe.
I never said autonomous cars had to be 'perfect', but I do think they need to be better than at least AVERAGE human drivers. And that includes getting people where they want to be when they want to go there, whether you think it is a 'dumb' reason or not., and whether or not that involves driving in a thunderstorm (around here known as 'afternoon'), or through a construction zone, or in an inch of snow, or in a congested area, or anywhere that is not 'safe'.
And I don't see anything dumb about any of those you listed.
And yet nobody ever says that a kid just starting to drive on safe local streets in daylight is a superior driver. Quite the opposite in fact. Yet, the autonomous driving proponents make that exact claim. Weird.
Duh... Most software developers aren't actual people. Actual people don't publish shitty, buggy, unoptimized code and call it a finished product. This is why even though Uber's sensors were able to detect this, the car still kept going anyways. They literally stole Google's tech, only they still can't figure out how to implement it.
WTF is your garbled response even in response to? No one said that alpha software means its ready for user testing. And how is your misspelling of Alpha as Alfa a pun?
So all Waymo needs to do is drive up and down an open highway in daylight with no intersections for a million miles and they'll get a golden pass to drive anywhere one day? That's not right. Eventually they'll have to demonstrate they can deal with anything a random environment can throw at them.
Laws are rules for the court, but merely a bottom bar to hit for life. Think beyond laws in your actions always.
That pedestrian was crossing the street illegally at night with dark clothing. I'm not saying she deserved to die - it would always be preferable to avoid injury to anyone. However, the reality is that even a human controlled car would have had a high likelihood of striking and killing her.
The accident rate will NEVER be zero. It's not possible. All we can hope for as an accident rate that is as good or better than that of a human driver.
"People who think they know everything are very annoying to those of us who do."-Mark Twain
The bug that caused this death was probably hiding another bug...
If Uber wasn't using lidar sensors yeah... that's entirely on Uber.
Yes.
If the software failed catastrophically, can't they figure out why it did?
If Windows has a bug, can't MS figure out how to fix it?
It's a computer... they should be able to run simulations to figure out what went wrong, and then be able to to alter the software to account for the possibility so that at least there can be some assurance that it won't happen again.
See all computers ever.
I mean one self driving car has one accident, and they ban all self driving cars. If an illegal immigrant had run over a family and the state decided to ban all illegal immigrants from the roads, everyone would be up in arms and calling for the execution of the governor Arizona. However since it is a self driving car, nobody thinks twice about this discriminatory action. Self driving cars have right too.
Lets show compassion for the machines that are going to be replacing us, just as we show compassion for illegal immigrants that take our jobs. Humans have no more right to the roads than artificially intelligent vehicles do.
Lets make the streets safe for self driving cars, and end the inequal and unjust domination of the highways by human beings.
Your point is well made, although with lives on the line, the stakes are a bit higher than just being annoying or inconvenient or even economically costly.
File under 'M' for 'Manic ranting'
Yeah, that's exactly what it means and it means they can rape your small children, of course. Once again you are demonstrating you have no knowledge of autonomous vehicle technology or even simple logic.
Yet nobody ever says that a kid should never be allowed anywhere near a real world driving condition until he has proven himself 100% safe. Yet the autonomous driving critics make that exact claim. Weird.
Or we could, you know, ignore the extremists on both ends and look for a reasonable middle ground? Nah, where's the fun in talking about that?
Yes, but has there been any testing on a stream of traffic made up entirely of these cars?
This question arises because many of the minor collisions in stop-and-go traffic occur when one inattentive driver slams on his brakes at the last instant. Even if he avoids hitting the car in front, this action can create a ripple through the line of traffic that leads to everyone suddenly braking, with a larger number of collisions. So let's see what happens when all cars use the same kind of automation as Audi in stalled traffic.
Ah, 'double or triple the chance of killing someone'. You mean, in exactly the same way as you double your chances of winning the lottery by buying two tickets. Double the chances! It's practically a sure thing!
Wait, so do you want to talk about the odds or not? You can't have it both ways. Either you care about that 1 in 10,000 chance of killing someone or you don't.
Per your logic, the 'best' car (with the most points) is one that doesn't move at all.
No, by my logic we should take reasonable risks for reasonable benefits. Why is that hard to understand?
By the way, the biggest cause of weather related accidents isn't snow, or ice, or even rain. It is wet roads. Do you know how often the roads are wet around here? Damn near every day. Guess we should all just be hermits so we are safe.
Or, again, we make reasonable choices. You are an advocate for just saying "fuck it, we're all going to die anyway." Which means you have absolutely nothing constructive to say on the safety of AVs.
I never said autonomous cars had to be 'perfect', but I do think they need to be better than at least AVERAGE human drivers. And that includes getting people where they want to be when they want to go there, whether you think it is a 'dumb' reason or not., and whether or not that involves driving in a thunderstorm (around here known as 'afternoon'), or through a construction zone, or in an inch of snow, or in a congested area, or anywhere that is not 'safe'.
And I don't see anything dumb about any of those you listed.
Of course you don't. Everything is the same and nothing matters. You are the perfect stooge for the Trump era of false equivalence.
Lives on the line doesn't mean it won't be a buggy computer. It just means that the basic premise is bad idea.
In principle, self driving cars could drastically cut down on vehicle accidents simply because of their faster reaction time alone.
File under 'M' for 'Manic ranting'
Should be, yes.
Will be? Hell no. See medical devices. Drug trials. Imported steel for military craft.
An experimental car with many thousands of dollars worth of equipment should not be fitted with a $2 dash cam. That much is pretty obvious.
None of the other companies have killed anyone yet. That's better in my book.
Hey, it's tough work, not for the faint of heart. Can't even attend a fraction of those furnerals either!
That "safety driver" was watching cartoons or the home shopping network, and would have had precisely as much positive impact on the situation at high noon. It doesn't matter why this situation occurred. It matters that this situation exposed a glaring lack of attention to safety in the testing process.
Some people refuse dictates of logic, just because of all that longer painful experiences.
And will you apply the same standard to human drivers
Yes. If a human driver kills a pedestrian in an avoidable accident, they should have their license revoked until they can demonstrate their competence.
Waymo and Tesla have killed zero pedestrians despite driving far more miles than Uber. Uber's road privileges should be revoked until they can explain what caused this accident, and show that it has been fixed.
Well, they could test on a closed course and have the executives walk across the road and see how well the car responds.
Probably because Uber Disabled Volvo's SUV Safety System
This is a completely irrelevant statistic at this point. You're comparing accidents per miles driven of regular vs. experimental self-driving cars...the two "sample sizes" so to speak are so vastly different that no valid comparison is possible.
You are quite mistaken. The comparison is fair in deciding whether AVs are safe enough for real world testing in limited environments, not whether they are ready to be left unsupervised in all conditions.
No it isn't. It isn't a fair comparison is any way, because we are not comparing the same type of data. That linked study analyzes the country-wide crash rate to the Google autonomous vehicle crash rate. Google autonomous vehicles do not operate in every environment. Your assertion would be OK if the comparison was autonomous vehicles vs. human drivers driving on the same streets, in the same conditions. So if Google is only testing in 3 cities in California or whatever, then only the crash data from those three cities. Don't use the crash data from Vermont, where there are accidents due to snow or ice, while the Google cars were initially confied only to sunny and dry places.
Furthermore the linked study attempts to compensate for unreported crashed using data from "naturalistic studies" of crashes...which means we are no longer dealing with actual national crash data (and using all of it was a flawed approach to begin with), but with estimates. Furthermore from reading the article I am left unsure whether for the autonomous cars, they have counted only the actual crashes, or the potential crashes as well (the car would have been in a crash but the human driver intervened to stop it), because those should be counted too for a fair comparison.
Finally, as I already said, the study only applies to Google cars, not Uber cars - and an Uber car killed someone. Even we accept your argument, then this study should only be used to decide whether to allow Google to test its cars on the road in real conditions, and not for anything else.
No it isn't. It isn't a fair comparison is any way, because we are not comparing the same type of data.
You listed a bunch of reasons that are irrelevant to the comparison, number of miles, number of cars etc. That's all bullshit. You could say the data set is not large enough to be statistically proven, but that does not mean it's not useful. This is pure FUD on your part.
You also talk about how it's not fair because of the backup driver. More FUD. The point is to compare whether these cars are safe enough to test WITH the backup driver.
Then you talk about how it's not fair because there are no stats on AVs for difficult conditions. Again, that would be relevant if we wanted to see whether it was safe to test in those environments, but that is not the question.
In summary, everything in your first post was pure luddite fear mongering.
Your assertion would be OK if the comparison was autonomous vehicles vs. human drivers driving on the same streets, in the same conditions. So if Google is only testing in 3 cities in California or whatever, then only the crash data from those three cities.
Only now do you bring up a truly relevant point, however it is one I addressed in a different post. Sure, it would be interesting to compare the latest Waymo cars against the current drivers on the same streets and conditions (with the backup drivers, of course). If you have that data, please provide it.
Don't use the crash data from Vermont, where there are accidents due to snow or ice, while the Google cars were initially confied only to sunny and dry places.
You would absolutely use that data in deciding whether an AV is safe enough to test on a public road as that data tells you what level of risk we are willing to accept for road transportation. If we're willing to accept 1 accident for every 100k miles then it is irrelevant where and when. This comparison answers the simple question "do these AVs cause more harm than human drivers." The answer is no. It is irrelevant whether these AVs are safer than the average driver on the exact same road. Interesting, sure, but not a measure of whether we should allow testing.
they have counted only the actual crashes, or the potential crashes as well (the car would have been in a crash but the human driver intervened to stop it), because those should be counted too for a fair comparison.
It's clear why you keep getting the wrong answers, because you still don't even understand the question.
:
Multiple choice, are we using this data to decide
A) is it safe to test an AV on certain public streets in certain conditions with a backup driver, or
B) is it safe to deploy these AVs everywhere under every conditions with no backup driver
You keep arguing why B is not true. A point that absolutely nobody here is arguing with.
That is an emotional response. It doesn't necessarily save lives, it may actually lead to far more traffic fatalities.
That response has everything to do with the question.
Über is a company run by and full of arrogant, self-absorbed brogrammer douche bags who know just enough to be dangerous. They are shining examples of the Dunning-Kruger effect.
Their car didn't see the pedestrian because the "engineers" they have working on it came straight off the set of "Sillicon Valley" and know more about "raw water" trends than they do about engineering.
Driving at night is a PRIMARY USE CASE. There is only one reason the car didn't see the pedestrian -- extreme negligence.
God, it used to be that you had to hold a P.E. license to be an engineer. Now, you just have to talk the talk.
What you assholes have done to the tech industry over the last decade makes me sick.
Shouldn't that be whether specific AVs are safe enough? It may well be that Waymo AVs are safe enough, and Uber AVs are not.
"When you have eliminated the unacceptable, whatever is left, however improbable, must be the truthiness" - Holmes
People keep changing the parameters, use whichever you like.
If you are deciding whether a specific AV is "safe enough" you would probably want to be as specific as possible. However, there is probably not enough data on any specific vehicle and program to know for sure. Using expanded data might be useful.
What is more important is whether the company is willing and able to pay for damages their program causes.
The data on the Uber cars suggest they're a lot less safe than Waymo cars. Assuming a government will allow self-driving cars, it's reasonable to look at which ones look safe enough.
Damages don't necessarily apply. No amount of money will bring back the victim of the last Uber accident. We don't want rich people and well-funded corporations to go out and kill people.
"When you have eliminated the unacceptable, whatever is left, however improbable, must be the truthiness" - Holmes
The data on the Uber cars suggest they're a lot less safe than Waymo cars. Assuming a government will allow self-driving cars, it's reasonable to look at which ones look safe enough.
I agree the data "suggest" Waymo is safer, especially for the tasks they've been put to. However, it's not certain what this really means as there are too many variables. Uber takes more chances so it's unclear Waymo would handle those situations a lot better.
Damages don't necessarily apply. No amount of money will bring back the victim of the last Uber accident. We don't want rich people and well-funded corporations to go out and kill people.
Of course damages apply. Yours is a terrible philosophy which ends up costing many lives.
I hate that so many people just don't get this, and many of them are the politicians that are making terrible choices.
Pretending that we don't put a $ amount on human lives is just silly and hypocritical. If lives were as precious as you believe then this accident never would have happened because the road would have been better lit, there would have been a pedestrian overpass, there would have been a safety officer to escort homeless people across the road... Any number of safety improvements that we decided were not worth the cost.
Demanding financial compensation isn't a license to go out and murder individuals, it is a value to risk calculation that a responsible government does and responsible corporations also do. Irresponsible governments and corporations also do it, just very badly.
Delaying AV development unnecessarily will cost lives. It may be that Uber will never have a vehicle safer than a human, but the government has no clue how to evaluate that (obviously). As long as they pay the external costs for their testing it should be the company's decision.
If they want to pay $100m for every fatality, I'd gladly let them test in my town and we'd be much safer for it.
It may be that Uber cars are not as good as Waymo's, or that Uber does far more demanding tests. In either case, there's reason to allow Waymo to continue testing and stop Uber until their cars are reasonably safe to use in the tests they are running, by some combination of improving their cars or making their tests less ambitious.
I was obviously insufficiently clear about the damages. What you said was "What is more important is whether the company is willing and able to pay for damages their program causes." My point is that no damages are going to make the victims whole in the case of death, and so not killing people is more important than paying weregild (which was dropped from the legal system a long time ago). If what's important is that the company is prepared to pay $X million if its experiments kill someone, then they have a license to kill that's limited only by their cash supply. I'd rather see them not kill people if they can avoid it.
The street was apparently considered adequately safe beforehand, but that was without self-driving cars not noticing pedestrians while the safety driver plays Candy Crush or whatever. That looks a lot like criminal negligence, and we don't what that happening no matter how much money we can soak the company for.
"When you have eliminated the unacceptable, whatever is left, however improbable, must be the truthiness" - Holmes
I had a long response, but fuck it. Life's too short. Nothing I say can convince you that being a better person is better than feeling like you're a better person.
I'd be real interested in finding out how you got your conclusion by reading what I wrote.,
"When you have eliminated the unacceptable, whatever is left, however improbable, must be the truthiness" - Holmes
it would apparently take you years of introspection. Better get started. (I suggest you first try reading what you wrote.)