sort of OT autonomous car fatality

[jimcav]

So I saw a scrolling news alert that a self driving uber hit a woman pushing a bicycle (hence the sort of OT) in AZ, apparently pushing a bike last night through an intersection. At work I can't access new sites (but can this forum--duh). A quick search on my phone said the uber had a "control driver" as back up. So I guess the conditions must have been bad for both the self-driving electronics and the human pilot to both miss seeing her. Just wondered if anyone had more details.

[tuscanyswe]

I dont but thats one of the clear problems with these cars.
They should either do it all or they should do nothing imo. Driver in the car thats not suppose to drive is a passenger imo. How would he know when to intervene unless obvious and by that time i bet its often to late anyways.

Sad that a person had to die so we can have computers driving us to work. Well that we test them in public before fool proof at least.

[fiamme red]

https://www.nytimes.com/2018/03/19/technology/uber-driverless-fatality.html

[jimcav]

https://www.nytimes.com/2018/03/19/technology/uber-driverless-fatality.html
I'm sure they will discuss it on CNN when i get home if someone else hasn't been poisoned, or fired, or resigned to completly monopolize the news

[ftf]

Sad that a person had to die so we can have computers driving us to work. Well that we test them in public before fool proof at least.

How many people do you think were killed today because we allow humans to drive cars?

Answer, almost 3300 a day, or 1,300,000 a year

[fiamme red]

Now we have self-driving trucks on the highways too: https://www.nytimes.com/2018/03/06/technology/uber-self-driving-trucks.html.

Who's responsible if a self-driving vehicle kills someone? The manufacturer?

[MattTuck]

The loss of life is tragic.

This squares with many bicycle/pedestrian deaths in that it happened at night. Article says the woman was crossing the street, outside of a cross walk. Not sure of the exact details, which I am sure will come out eventually. My dad always told me, when crossing the street, to assume that cars cannot see you. Hence, stop, look both ways, listen.

Not sure what the circumstances were that led to this, but it seems to confirm that roads are dangerous places for pedestrians and other vulnerable road users, and all caution should be taken.

As for regulations of the autonomous driving industry, policy makers have a real chance to create a system that is fair and internalizes all the costs... but they'll likely mess it up.

[saab2000]

Now we have self-driving trucks on the highways too: https://www.nytimes.com/2018/03/06/technology/uber-self-driving-trucks.html.

Who's responsible if a self-driving vehicle kills someone? The manufacturer?

If the occupant has override capabilities I would assume the operator would be responsible for accidents as the operator presumably has the possibility to avoid the accident.

It will indeed be interesting to find out the specifics of this exact case. I would assume the car has cameras that can be viewed by the developer so a real picture may eventually come to light.

[jimcav]

If the occupant has override capabilities I would assume the operator would be responsible for accidents as the operator presumably has the possibility to avoid the accident.

It will indeed be interesting to find out the specifics of this exact case. I would assume the car has cameras that can be viewed by the developer so a real picture may eventually come to light.

that would make sense. I can't decide whether to marvel or cringe at self-driving. It makes total sense that it is possible as the aviation stuff is pretty amazing these days, but then I think of the roomba i saw last year getting "stuck" and wonder what sort of redundancy and reliability is possible.

[tuscanyswe]

How many people do you think were killed today because we allow humans to drive cars?

Answer, almost 3300 a day, or 1,300,000 a year

Its not the same tho is it? Once they have concluded that the statistics are in favor of computer driven cars thats one thing but right now they are testing it! Wondering if the tech is good enough to make it work to an acceptable level.

To me, its sad someone died because of poor tech. Thats how i c it at this point in time (without more info than currently available).

And for the statistics i think without checking that if i were to do the math of hours of human driven cars in usa vs hours of computer driven cars in usa who hit a person crossing the street on foot with a bicycle the statistics would favor the human drivers. I dont know that but i would guess by a lot. And no 1 case is not a good data point i know but u brought up statistics.

[Mark McM]

Its not the same tho is it? Once they have concluded that the statistics are in favor of computer driven cars thats one thing but right now they are testing it! Wondering if the tech is good enough to make it work to an acceptable level.

To me, its sad someone died because of poor tech. Thats how i c it at this point in time (without more info than currently available).

And for the statistics i think without checking that if i were to do the math of hours of human driven cars in usa vs hours of computer driven cars in usa who hit a person crossing the street on foot with a bicycle the statistics would favor the human drivers. I dont know that but i would guess by a lot. And no 1 case is not a good data point i know but u brought up statistics.

Statistics, public perception, and regulatory actions are all completely independent.

Even after the 9/11 terrorist hijackings, commercial aviation remained the safest form of transportation by far. And yet, we now have a multi-billion dollar agency call the TSA (which even after 17 years, has yet to demonstrate that they have ever stopped a terrorist attack).

Google's statistics currently show that their self-driving cars have far lower accident rates (per mile traveled) than human driven cars. And yet, I am sure that there some people who will never feel that safe-driving cars will ever be safe enough.

[tuscanyswe]

Statistics, public perception, and regulatory actions are all completely independent.

Even after the 9/11 terrorist hijackings, commercial aviation remained the safest form of transportation by far. And yet, we now have a multi-billion dollar agency call the TSA (which even after 17 years, has yet to demonstrate that they have ever stopped a terrorist attack).

Google's statistics currently show that their self-driving cars have far lower accident rates (per mile traveled) than human driven cars. And yet, I am sure that there some people who will never feel that safe-driving cars will ever be safe enough.

Sure i get that.

What constitutes an accident in googles data? Fatality is something else. And fatalities from a person outside a car are quite rare. I think even 1 death here puts them behind on this front but again this is me with a hunch could be way wrong.

[rwsaunders]

How many people do you think were killed today because we allow humans to drive cars? Answer, almost 3300 a day, or 1,300,000 a year

Add to that stat...well over 90% of traffic deaths involve operator error. I recently read that there are over 1,000 individuals employed in the autonomous driving industry in Pittsburgh alone...the grey Volvos are everywhere here, mostly in the dense, urban neighborhoods and downtown.

PENNDOT stated in an article today that they're looking to use autonomous driving trucks in the future (as the safety bumper trucks and line striping trucks) to accompany road crews. I think that the technology is fascinating and there are predictions that the day will come where if you own a car, you can let it work while you're at work...as an autonomous driving vehicle.

[FlashUNC]

Some pretty high grade whataboutism going on in this one.

Whether this was avoidable or not in the software is just part of the larger issue that these cars will need to be programmed to make life or death decisions, including prioritizing in some cases who lives and who doesn't. Like a pedestrian outside a crosswalk.

But hey, let's use traffic deaths for cars driven by people to excuse one killed by a beta autonomous system. Because the streets can be places where we all serve as guinea pigs, right?

[MattTuck]

Some pretty high grade whataboutism going on in this one.

Whether this was avoidable or not in the software is just part of the larger issue that these cars will need to be programmed to make life or death decisions, including prioritizing in some cases who lives and who doesn't. Like a pedestrian outside a crosswalk.

But hey, let's use traffic deaths for cars driven by people to excuse one killed by a beta autonomous system. Because the streets can be places where we all serve as guinea pigs, right?

I mean, the same could be said for people who drive drunk, or drive distracted, or drive under the influence of prescription or illicit drugs... that we're just guinea pigs to see if those things are dangerous.

I'd rather like to see a tax that funds some sort of victim compensation fund. In the case of autonomous cars, there could be a 'autonomous vehicle testing registration' that would be some amount of money, with a portion of that going to fund regulatory activity, and the bulk of it going to a fund designed specifically to compensate the inevitable victims of errors, mistakes, accidents caused by autonomous cars.

I'd like to see the same for things like prescription opioids, guns and mobile phones... but I suspect the ship has sailed on all those. Atleast they have the chance to do it right with autonomous cars, but they probably won't.

[Kontact]

Unless the car swerved to hit the pedestrian, I have a hard time seeing how this was the fault of the computer system. The "driver" had full access to the brakes. I think the victim stepped out in front of a car that would have hit her no matter what was driving.

[marciero]

Some pretty high grade whataboutism going on in this one.

Whether this was avoidable or not in the software is just part of the larger issue that these cars will need to be programmed to make life or death decisions, including prioritizing in some cases who lives and who doesn't. Like a pedestrian outside a crosswalk.

But hey, let's use traffic deaths for cars driven by people to excuse one killed by a beta autonomous system. Because the streets can be places where we all serve as guinea pigs, right?

Only if the software performs as intended and still kills someone. But what the software (and the hardware) is intended to do and whether it performs as intended are two very distinct and separate issues. One is engineering and one is ethics. I dont understand why some people are freaked out about the prospect of the codifying (literally) of ethical decision-making. The alternative - what we have now- is no decision at all and leaving it to chance, since crashes all happen too fast for humans to make or act on any decision.

On the other hand I can understand people being uncomfortable with the reliability of the software, even though all the evidence indicates that humans are far more error-prone than computers.

[unterhausen]

I don't know about uber's cars, but Elon Musk is insisting they don't need lasers. No, they do need lasers.

This woman was either riding a bicycle or pushing one. I'm going with riding for now.

[goonster]

If the occupant has override capabilities I would assume the operator would be responsible for accidents as the operator presumably has the possibility to avoid the accident.

It's either a self-driving ("driverless") car, or it isn't.

Watching the car like a hawk, poised to perform some lifesaving intervention in a split second, is more exhausting than just driving.

The greatest danger happens when we're not sure what the machine will do, or how much the machine is in control. Is it in automatic, or manual? What mode is it in, and what does that do? (c.f. AF447 (http://www.vanityfair.com/business/2014/10/air-france-flight-447-crash)) Is the anti-lock/stability control halfway off, or all the way off? Is there a driver in that car, or is it fully auto?

these cars will need to be programmed to make life or death decisions, including prioritizing in some cases who lives and who doesn't.

Strongly disagree. The car doesn't make a life or death decision, it is programmed to do A or B. If programmed well, the car will respond consistently given the same set of inputs; it doesn't have a choice, per se.

Indeed, there are powerful ethical issues, with life and death consequences in play, but they are on us, not the machines.

IMHO, everybody involved in this Uber project, who shrugs their shoulders and says "the first fatality was unavoidable, just a question of time" should have their engineering degrees rescinded.

[FlashUNC]

Only if the software performs as intended and still kills someone. But what the software (and the hardware) is intended to do and whether it performs as intended are two very distinct and separate issues. One is engineering and one is ethics. I dont understand why some people are freaked out about the prospect of the codifying (literally) of ethical decision-making. The alternative - what we have now- is no decision at all and leaving it to chance, since crashes all happen too fast for humans to make or act on any decision.

On the other hand I can understand people being uncomfortable with the reliability of the software, even though all the evidence indicates that humans are far more error-prone than computers.

The braking system fails and the car has to determine whether to kill a pedestrian to save the drivers life, or vice versa. Who programs that? Who's responsible? If the pedestrian, is the car company prioritizing car owners' lives over the lives of the public? If the public, who's going to buy a car that doesn't want to keep them safe?

This kind of programming is Pandora's box man.

[rwsaunders]

MIT's Technology Review published these articles a few years back regarding the ethics of programming a driverless car...making split second decisions of valuing the occupant more than the non-occupant, adult vs child, squirrels vs oncoming traffic, etc. Interesting questions indeed.

Here is an excerpt which is food for thought...Others believe the situation is a little more complicated. For example, Bryant Walker-Smith, an assistant professor at the University of South Carolina who studies the legal and social implications of self-driving vehicles, says plenty of ethical decisions are already made in automotive engineering. “Ethics, philosophy, law: all of these assumptions underpin so many decisions,” he says. “If you look at airbags, for example, inherent in that technology is the assumption that you’re going to save a lot of lives, and only kill a few.”

https://www.technologyreview.com/s/542626/why-self-driving-cars-must-be-programmed-to-kill/

https://www.technologyreview.com/news/539731/how-to-help-self-driving-cars-make-ethical-decisions/

[jimcav]

you'd think that 3 years later they would have actually had to do some programming for certain scenarios. even if just in "the lab" because they need to be to have the 3laws strong difference engine ready


MIT's Technology Review published these articles a few years back regarding the ethics of programming a driverless car...making split second decisions of valuing the occupant more than the non-occupant, adult vs child, squirrels vs oncoming traffic, etc. Interesting questions indeed.

Here is an excerpt which is food for thought...Others believe the situation is a little more complicated. For example, Bryant Walker-Smith, an assistant professor at the University of South Carolina who studies the legal and social implications of self-driving vehicles, says plenty of ethical decisions are already made in automotive engineering. “Ethics, philosophy, law: all of these assumptions underpin so many decisions,” he says. “If you look at airbags, for example, inherent in that technology is the assumption that you’re going to save a lot of lives, and only kill a few.”

https://www.technologyreview.com/s/542626/why-self-driving-cars-must-be-programmed-to-kill/

https://www.technologyreview.com/news/539731/how-to-help-self-driving-cars-make-ethical-decisions/

[binxnyrwarrsoul]

I'm sure they will discuss it on CNN when i get home if someone else hasn't been poisoned, or fired, or resigned to completly monopolize the news

Or there's even a dusting of snow.

[Kontact]

IMHO, everybody involved in this Uber project, who shrugs their shoulders and says "the first fatality was unavoidable, just a question of time" should have their engineering degrees rescinded.

What should they do or say instead?


Unless a robot driver is able to break the laws of physics, there are always going to be accidents. The reaction time can be as fast as the speed of light, but brakes don't stop cars instantaneously, so some version of this accident was preordained.

[marciero]

The braking system fails and the car has to determine whether to kill a pedestrian to save the drivers life, or vice versa. Who programs that? Who's responsible? If the pedestrian, is the car company prioritizing car owners' lives over the lives of the public? If the public, who's going to buy a car that doesn't want to keep them safe?

This kind of programming is Pandora's box man.

I'm pretty sure Mercedes-Benz, for example, has decided to prioritize passenger safety over public safety, with the rationale that the public would in fact end up being safer in the end. They reason that with more people buying cars that prioritize their own safety, more of these cars safer cars would be on the road, making the roads safer for the public.
But yes, a Pandora's box in the sense that, as rwsaunders points out, it takes all of these interesting and sometimes disturbing Trolley Problem-type ethical dilemmas out of the purely academic realm.

[soulspinner]

Unless the car swerved to hit the pedestrian, I have a hard time seeing how this was the fault of the computer system. The "driver" had full access to the brakes. I think the victim stepped out in front of a car that would have hit her no matter what was driving.

This. I'm in these cars every day. Unless the user disa
bled the system, something isn't right.

[oldpotatoe]

So I saw a scrolling news alert that a self driving uber hit a woman pushing a bicycle (hence the sort of OT) in AZ, apparently pushing a bike last night through an intersection. At work I can't access new sites (but can this forum--duh). A quick search on my phone said the uber had a "control driver" as back up. So I guess the conditions must have been bad for both the self-driving electronics and the human pilot to both miss seeing her. Just wondered if anyone had more details.

She wasn't in the intersection, is what I read, but walking across the road between intersections. Gotta wonder why she didn't see the car.

[jimcav]

Or there's even a dusting of snow.

I haven't bought a HD antenaa and the only thing I've figured out how to add besides amazon prime and netflix was cnn2go, so yes, that is what has been on
and of course i was wrong, it was the cambridge analytica whisteblower last night: so i swtiched over to prime and watched "swiss army man" which i enjoyed

[Mark McM]

IMHO, everybody involved in this Uber project, who shrugs their shoulders and says "the first fatality was unavoidable, just a question of time" should have their engineering degrees rescinded.

Pretty much what Kontact said. The only vehicle that has zero possibility of hitting something (or someone) is one that never moves. We as a society have to decide on what cost/risk vs. benefit we are willing to accept for autonomous vehicles, but no matter how hard engineers try, the risks of autonomous vehicles can never be zero - although they may be made much less than for human drivers.

[Tony T]

I'm sure they will discuss it on CNN when i get home if someone else hasn't been poisoned, or fired, or resigned to completly monopolize the news

CNN?
The 24/7 "Stormy Daniels" channel?
They'll spend 10 minutes on the news and then get back to "Stormy"

[Big Dan]

CNN?
The 24/7 "Stormy Daniels" channel?
They'll spend 10 minutes on the news and then get back to "Stormy"

I feel bad for you dude.

[Tony T]

Why? Because I've lost faith in CNN to report the news?

[MattTuck]

We as a society have to decide on what cost/risk vs. benefit we are willing to accept for autonomous vehicles, but no matter how hard engineers try, the risks of autonomous vehicles can never be zero - although they may be made much less than for human drivers.

This is precisely it. If you lose someone because of a drunk driver, you blame the fact that he was drinking, not the precise situation that led to the actual accident. It is easier to parse that way. If an autonomous car kills someone you love, even if in aggregate they are safer for everyone, there is still a question of what caused it. Oh, our neural network algorithm, told the car it was safe to merge. Why? well, because our training data set that the algorithm learned from didn't have this particular situation, and did not recognize the danger...

[merckxman]

The following from LAB:

Act Now: Require Automated Vehicles to See Bicyclists and Pedestrians
On Sunday night, an autonomous vehicle hit and killed a pedestrian in Tempe, AZ, as she was walking her bicycle across the street. While the details of the crash are still forthcoming, the League of American Bicyclists is concerned that these vehicles are being deployed without having to first prove their ability to recognize and respond to people biking and walking in our streets.

When human drivers apply for a driver’s license we have to pass a vision test. The League believes that all automated driving systems should first have to pass a “vision test” as well — requiring a safety performance standard — proving their ability to recognize and respond to people bicycling and walking, before they are on community streets.

Right now the Senate is considering S. 1885: AV START Act to set guidelines for automated vehicle manufacturers to test their vehicles on our streets. Please join the League of American Bicyclists in asking Senators to require automated vehicles to pass a vision test by going to : http://p2a.co/Oy2UjKv

[Kontact]

This is precisely it. If you lose someone because of a drunk driver, you blame the fact that he was drinking, not the precise situation that led to the actual accident. It is easier to parse that way. If an autonomous car kills someone you love, even if in aggregate they are safer for everyone, there is still a question of what caused it. Oh, our neural network algorithm, told the car it was safe to merge. Why? well, because our training data set that the algorithm learned from didn't have this particular situation, and did not recognize the danger...

Every self driving accident should be reviewed and learned from - something that is uniquely possible with a driver that has a hard drive.

But every accident involving a robot isn't because the robot's programming is deficient.

[MattTuck]

The video has been released. I don't really feel like posting it here, because it feels a bit gratuitous and tragic.

Three things jump out. First, the human sitting in the driver seat that is supposed to be supervising the driving appears to be texting or looking at his phone. Second, the person crossing the street does seem to come into view pretty suddenly. I have no idea if this is a function of the camera's resolution and low light abilities, and that perhaps a human eye would have identified the pedestrian sooner. and Third, the pedestrian appears to be walking at an angle to the road, away from the oncoming car. Which certainly doesn't help with situational awareness.

[Kontact]

I saw the video, and she looked reasonably perpendicular to the road to me.

The video probably doesn't have the same night vision as a human driver as it is just a lowish resolution dash cam.

Regardless of what was visible in video, the Uber vehicle uses radar, lidar and a video camera that is probably able to see past the visible spectrum and intensify low light images (like most camcorders do nowadays). It is inconceivable that the victim and her metal bike could have been invisible to such a layered system crossing left to right on a straight 3 lane road.


It appears Uber has really screwed the pooch.

[Louis]

Lots of blame to go around on this one:

1) Pedestrian crossing in an unsafe manner. (color of clothing, low light, right in front of an oncoming vehicle etc)

2) "Safety monitor" clearly distracted by smart-device, probably texting. Why bother having a monitor there if they aren't going to watch the road?

3) Vehicle systems (hardware + software) - it sure as heck looks as if they were not able to sense the pedestrian and/or react in time. And surely that's an expected and typical situation (where something suddenly crosses the road in front of the moving vehicle). I don't know if here the sensors were visual only, but if they weren't, the non-visual (e.g. radar) didn't do a good job.

Finally, there are plenty of cases where neither the best drivers nor the best machines could possibly avoid an accident. It's probably too early to say if this was one of them. Driving can be hard and dangerous, depending on the conditions.

[ripvanrando]

The run over cyclist was walking perpendicular to the road with no parked vehicles and no vehicles coming the other direction. The throw on the Uber vehicle lights looks well short of my cars. Looks misadjusted but that does not excuse or explain this manslaughter.

What? How?


The ex-con Rafaela Vasquez should go back to prison.

Uber is screwed.

Poor lady who was killed. It was probably a prius and she relied upon sound to cross. Pedestrians walk out on my all the time when riding my bike.

[Louis]

Poor lady who was killed. It was probably a prius and she relied upon sound to cross.

I'm pretty sure it was a Volvo.

https://www.telegraph.co.uk/content/dam/technology/2018/03/19/TELEMMGLPICT000157872703_trans_NvBQzQNjv4BqNJjoeBT 78QIaYdkJdEY4CnGTJFJS74MYhNY6w3GNbO8.jpeg?imwidth= 1400

[fiamme red]

Poor lady who was killed. It was probably a prius and she relied upon sound to cross.It was a Volvo XC90 SUV.

[ripvanrando]

It was a Volvo XC90 SUV.

Senseless death.

Uber should have stopped

[Louis]

Uber should have stopped

I've hit deer crossing in front of me at night. I've tried to stop, but in both cases wasn't able to in time.

It's easy to say "should have stopped" but in reality it doesn't always happen.

[Kontact]

I've hit deer crossing in front of me at night. I've tried to stop, but in both cases wasn't able to in time.

It's easy to say "should have stopped" but in reality it doesn't always happen.

Do you have radar, lidar and electric nerves?

The Uber didn't even attempt to stop or swerve.

[ripvanrando]

I've hit deer crossing in front of me at night. I've tried to stop, but in both cases wasn't able to in time.

It's easy to say "should have stopped" but in reality it doesn't always happen.

Uber did not even react. The ex-con was not even looking

It looks to me that there was at least 100-120 feet, which was plenty of distance to stop from 50-60 mph in a car with good brakes. From 30-40 mph should have been routine.

[Louis]

Do you have radar, lidar and electric nerves?

Yes, in fact, I do. Don't you?

The Uber didn't even attempt to stop or swerve.

As one might expect, if the sensors didn't "see" the pedestrian.

Machines aren't magic. I don't understand why people expect them to work in all situations.

[ripvanrando]

Yes, in fact, I do. Don't you?



As one might expect, if the sensors didn't "see" the pedestrian.

Machines aren't magic. I don't understand why people expect them to work in all situations.

Someone walking across the road with no other traffic present and you excuse the manslaughter? Recognizing a human on the road and braking to a stop is not magic-in fact, the technology has been employed for several years already.

[Louis]

Someone walking across the road with no other traffic present and you excuse the manslaughter?

There's a difference between "excusing" something and saying we don't have access to all the information to know why it happened. Don't you think it's a bit early to pass judgement? I'm perfectly happy to wait for more information based on technical analysis of the situation before deciding.

[ripvanrando]

Don't you think it's a bit early to pass judgement?

No.

[Kontact]

Yes, in fact, I do. Don't you?



As one might expect, if the sensors didn't "see" the pedestrian.

Machines aren't magic. I don't understand why people expect them to work in all situations.

Do you get in elevators and expect there to be a decent chance you will plummet to your death?

We don't expect machines to never fail, but we do expect them to work as designed except under extenuating circumstances. There don't appear to be any extenuating circumstances to this incident. Straight road, no weather, the victim had a metal bike to help reflect radar.

I can't understand how at least one of the three "visual" systems didn't detect the woman and/or bike and make some speed or directional change before barreling into her. I'm betting that part of the car did "see" her and just didn't make up its "mind" to do something. Yet it saw the lane markers and managed to track its lane perfectly.

[Louis]

Do you get in elevators and expect there to be a decent chance you will plummet to your death?

The conditions in which an elevator is expected to operate are way more predictable than an autonomous vehicle, so the machines that are to do the job are vastly different in complexity and this really isn't a legitimate comparison.

But to answer your question, I bet if you compared the injury and death rate of elevator users to car and truck users, I think you'd find that passengers in cars and truck are involved in serious incidents far more often than those in elevators.

[Davist]

How would he know when to intervene unless obvious and by that time i bet its often to late anyways.


This is what happened with that Air France flight that went down between Brazil and Paris in the South Atlantic, they didn't know what to do.

https://www.vanityfair.com/news/business/2014/10/air-france-flight-447-crash

As to the ethics issues, funny how it doesn't really stray too far from Asimov's 3 robot laws, from I, Robot (in fact a semi-autonomous vehicle crash was the basis for the Will Smith movie of the same name..)

[verticaldoug]

Lots of blame to go around on this one:

1) Pedestrian crossing in an unsafe manner. (color of clothing, low light, right in front of an oncoming vehicle etc)

2) "Safety monitor" clearly distracted by smart-device, probably texting. Why bother having a monitor there if they aren't going to watch the road?

3) Vehicle systems (hardware + software) - it sure as heck looks as if they were not able to sense the pedestrian and/or react in time. And surely that's an expected and typical situation (where something suddenly crosses the road in front of the moving vehicle). I don't know if here the sensors were visual only, but if they weren't, the non-visual (e.g. radar) didn't do a good job.

Finally, there are plenty of cases where neither the best drivers nor the best machines could possibly avoid an accident. It's probably too early to say if this was one of them. Driving can be hard and dangerous, depending on the conditions.

I think you are missing the most important part of the blame- Uber's history of flouting rules and cavalier attitude regarding product deployment. It seems pretty obvious that the system is not ready for prime time. And here they are driving around public streets.

Yeah, people do stupid things, but anyone who has ever driven a car knows that.

I think software developers can be pretty cavalier about new version rollouts and the debugging process. I bet the techies even jokingly refer to this stuff as a feature. It is one thing when a system gets hung and you can't close your screen, it is another when a system gets hung and kills someone.

This is total bull***** on Uber. You can say a system failure occur, but more important, an organizational faillure enabled this.

[ripvanrando]

For those who feel that the killed women ""jumped out" in front of the Uber, consider how far ahead of the Uber she was and whether any of us on our bicycles give more "notice" when we take the lane.

By my estimation the deceased was least 200-300 feet ahead of the Uber (in the opposing lane) when she entered the roadway. Why didn't Uber recognize the bicycle then? Why didn't Uber register the cyclist at all?

What if you were riding your bike crossing a street and there was an unseen pothole taking you down, would Uber ignore your corpse?

[MattTuck]

For those who feel that the killed women ""jumped out" in front of the Uber, consider how far ahead of the Uber she was and whether any of us on our bicycles give more "notice" when we take the lane.

By my estimation the deceased was least 200-300 feet ahead of the Uber (in the opposing lane) when she entered the roadway. Why didn't Uber recognize the bicycle then? Why didn't Uber register the cyclist at all?

What if you were riding your bike crossing a street and there was an unseen pothole taking you down, would Uber ignore your corpse?

I get what you're saying. The software obviously did not work, and the human back-up system was insufficient.

"It absolutely should have been able to pick her up," said Sam Abuelsmaid, an analyst for Navigant Research who follows autonomous vehicles. "From what I see in the video it sure looks like the car is at fault, not the pedestrian."

At the same time, the pedestrian could have easily avoided the fate that she did. To say she was in the right, but dead, is little consolation to her family and friends. It could have been avoided simply by waiting to cross the street until there were no cars coming.

Her death wasn't inevitable. If it had been a drunk driver, she'd have met the same fate, and you'd blame the driver, but in this case you blame the computer, not the human driver in the car. In both cases, the pedestrian would have been not at fault, but still tragically dead.

Do you really live your life like everyone is going to get out of your way? Maybe I am too careful, but when I cross the street, I never assume that a car is going to stop for me, even if they flash their lights, I wait to make sure they've slowed down enough and clearly see me. Why? I'd rather be alive and slightly inconvenienced than technically in the right and run over.

[Tony T]

In their tests, does Uber (Google too) review that the "safety drivers" have there hands "hovering above the steering wheel (which is what most backup drivers are instructed to do because it allows them to take control of the car quickly in the case of an emergency) — NY Times"?
Easy enough to have sensors in the car to do this.

[goonster]

the pedestrian could have easily avoided the fate that she did.

We hold people and machines to different standards. We can't reprogram all pedestrians to not be absent-minded sometimes. We do expect to revise the programs on these cars so that they are not occasionally blind.

Easy enough to have sensors in the car to do this.

https://upload.wikimedia.org/wikipedia/en/thumb/5/55/King-Size_Homer.png/225px-King-Size_Homer.png

[MattTuck]

We hold people and machines to different standards. We can't reprogram all pedestrians to not be absent-minded sometimes. We do expect to revise the programs on these cars so that they are not occasionally blind.


I think you actually make a good point. Different people are held to different standards. If it were a person driving and they were absent mindedly fiddling with the radio, and not paying attention to the road when they hit a pedestrian, you wouldn't absolve them because that is somehow part of the human condition. But if it is a pedestrian, it is ok to be absent minded on our public roads?

The woman who was killed did not know it was an autonomous car at the time. She just knew it was a car driving at night, that was not slowing down.

Trying to assign blame only goes so far. The more important question for me is to ask how we can prevent the next accident. To me, the choice is between A.) waiting for autonomous cars to achieve 100% safety and B.) looking both ways and waiting to cross the street until it is safe to do so. B seems like the more practical thing to do right now, even if it is taxing on the average road user.

[ripvanrando]

I get what you're saying. The software obviously did not work, and the human back-up system was insufficient.



At the same time, the pedestrian could have easily avoided the fate that she did. To say she was in the right, but dead, is little consolation to her family and friends. It could have been avoided simply by waiting to cross the street until there were no cars coming.

Her death wasn't inevitable. If it had been a drunk driver, she'd have met the same fate, and you'd blame the driver, but in this case you blame the computer, not the human driver in the car. In both cases, the pedestrian would have been not at fault, but still tragically dead.

Do you really live your life like everyone is going to get out of your way? Maybe I am too careful, but when I cross the street, I never assume that a car is going to stop for me, even if they flash their lights, I wait to make sure they've slowed down enough and clearly see me. Why? I'd rather be alive and slightly inconvenienced than technically in the right and run over.

Me neither. That is why I now use a mirror on all bikes and if a car does not move over, I am diving off the road. Riding legally with the right of way and an Uber mows you down, would you think the same way?

[Kontact]

Do you really live your life like everyone is going to get out of your way? Maybe I am too careful, but when I cross the street, I never assume that a car is going to stop for me, even if they flash their lights, I wait to make sure they've slowed down enough and clearly see me. Why? I'd rather be alive and slightly inconvenienced than technically in the right and run over.

As cyclists, we absolutely do. The vicitim was hit on the right side of the far right lane by a vehicle that should have seen her and gone around. If she was riding in the direction of traffic at 10 mph on the right of the right lane she would be just as dead.

What do you do when you ride to make sure that overtaking motorists observe you and don't run you over from behind? We ride on faith that vehicles won't simply ram into us from behind because they are blindly staying in the lane markings. All respect to mirror users, but I actually doubt there is time in a 45 mph zone to tell the difference between "haven't changed lanes yet" and "I don't have time to get out of way" when the oncoming car is approaching at 4 car lengths a second and only needs to move 6 feet to the left to pass.

[MattTuck]

As cyclists, we absolutely do. The vicitim was hit on the right side of the far right lane by a vehicle that should have seen her and gone around. If she was riding in the direction of traffic at 10 mph on the right of the right lane she would be just as dead.

What do you do when you ride to make sure that overtaking motorists observe you and don't run you over from behind? We ride on faith that vehicles won't simply ram into us from behind because they are blindly staying in the lane markings.

Well, for one, I don't typically ride at night. And when I do, I have several old lights that I would mount, and wear a reflective vest. In fact, I just got a new light for day time use. Two, I turn around and look at cars to let them know that I see them, and that also allows me to see if they are traveling fast, giving me a wide clearance, etc.
I have ridden with a mirror in the past, but I found that relying on that made some drivers feel like they could come closer to me, and I prefer to look backwards now even if I have a mirror. I think that when a driver sees that you know they're there, their behavior improves.

In my view, there is no way to eliminate all risk from these environments. You're dealing with heavy high velocity objects with huge amounts of kinetic energy, and potentially unpredictable behaviors due to the operators. The best you can do is take actions to minimize the risks you think are the most dangerous.

[Kontact]

Well, for one, I don't typically ride at night. And when I do, I have several old lights that I would mount, and wear a reflective vest. In fact, I just got a new light for day time use. Two, I turn around and look at cars to let them know that I see them, and that also allows me to see if they are traveling fast, giving me a wide clearance, etc.
I have ridden with a mirror in the past, but I found that relying on that made some drivers feel like they could come closer to me, and I prefer to look backwards now even if I have a mirror. I think that when a driver sees that you know they're there, their behavior improves.

In my view, there is no way to eliminate all risk from these environments. You're dealing with heavy high velocity objects with huge amounts of kinetic energy, and potentially unpredictable behaviors due to the operators. The best you can do is take actions to minimize the risks you think are the most dangerous.

I am taking the point of view that a machine that sees in lidar, radar and amplified video can see at night, so this isn't a failure to be seen but a failure to take action. The Uber never reacted to the pedestrian and struck her at full speed centered in its lane.


We act on faith that cars aren't going to jerk the wheel into us when we ride, walk or drive around other cars. We can game that to an extent, but the closing rates of vehicles is beyond see and avoid reaction times. Especially since bicycles require a countersteer to make sudden maneuvers - we have slow emergency maneuver speed.

[alancw3]

arizona police have released the video of the accident both inside and outside of the vehicle:

https://www.youtube.com/watch?v=hthyTh_fopo

looking at the video i am actually surprised that the autonomous car did not stop or even attempt to slow down. perhaps i am excepting to much from self driving! i certainly would have attempted to stop as a driver. just saying.

[tuscanyswe]

Edit: Its so terrible to watch stuff like this. Just sad all way around regardless of fault

[merckxman]

Scientific American article about the accident:
https://www.scientificamerican.com/article/uber-self-driving-car-fatality-reveals-the-technologys-blind-spots1/

[alancw3]

Scientific American article about the accident:
https://www.scientificamerican.com/article/uber-self-driving-car-fatality-reveals-the-technologys-blind-spots1/

interesting article. looks like self driving cars still have a ways to go to truly be autonomous.

[Kontact]

Scientific American article about the accident:
https://www.scientificamerican.com/article/uber-self-driving-car-fatality-reveals-the-technologys-blind-spots1/

The article has "blind spots" in the title, but then goes on to say that there is no good reason the layered senses of the car didn't see the victim.


I'll be very interested when they tell whether the victim was "visible", whether she was recognized as an object in the road and whether the system attempted to make a control input.

[ripvanrando]

The real challenge, says Pratt, a former academic and government program manager in robotics and intelligent systems, is developing a vehicle that can drive in "very difficult domains," such as rainy weather or crowded roads. That's level five, and Shladover, for one, says he wouldn't be surprised if it's 2075 before we get there

Shladover believes AV companies need to be much clearer about the "operational design" of their vehicles—in other words, the specific set of conditions under which the cars can function without a driver's assistance. "But most of the time they won't say, or they don't even know themselves," he says.

Congress is now considering legislation that would allow AV-makers to deploy the cars so long as they are deemed as safe as current vehicles. Even that is a high bar, Shladover notes

How much less safe than a driver operated vehicle is good enough?

http://www.sciencemag.org/news/2017/12/are-we-going-too-fast-driverless-cars

[72gmc]

I've been quietly hoping that non-human sensors and automated responses might increase cyclist safety. That we could add something to our bikes or clothes that is sensed by vehicle systems much more reliably than we are sensed by humans who are staring at their phones.

This Uber incident doesn't necessarily eliminate that hope, it just tells me they aren't there yet. And it tells me Uber doesn't have an entirely professional approach to this effort, which is not surprising given that it's Uber.

[Climb01742]

There's a profound question at the heart of this tragic situation:

We humans are imperfect (like when we drive). Technology offers the possibility of correcting some of our flaws (like when we drive). Getting there will have its price and tragic consequences.

No one should die because any company pushes technology forward too fast, or too sloppily. But equally, people should not continue to die if, someday, technology could prevent it. (Think of medicine. Medicine's path to progress costs lives, but now it saves lives.)

Staying where we are isn't a great option. Pushing technology forward too fast isn't a great option. Question is, how do we progress humanely? If anything good could come out of the awful situation it might be trying to find that answer.

As an aside, it's hard not to see some of these same issues involved with Facebook. When profit is your only goal, technology can f' over a lot of people and potentially democracy. Who's watching the technology? Not the folks who make it, that's for sure.

[OldCrank]

I just saw that video-
anybody else think it looks ummm... enhanced?

In the critical moment just before she was visible, the woman's reflective sneaker looked like it was sliding towards the car, Very Fast.

And then, there she was- totally lit up. Very focused streetlights?
Headlights that stopped at 30' from the car?

Curious.

Also I feel bad for that paid rider. Instant pariah.

Finally, she was Walking the bike. A pedestrian.
Just a legal nit.

[ripvanrando]

There's a profound question at the heart of this tragic situation:


No one should die because any company pushes technology forward too fast, or too sloppily. But equally, people should not continue to die if, someday, technology could prevent it. (Think of medicine. Medicine's path to progress costs lives, but now it saves lives.)

Staying where we are isn't a great option. Pushing technology forward too fast isn't a great option. Question is, how do we progress humanely? If anything good could come out of the awful situation it might be trying to find that answer.


I thought of medical device development and pharmaceutical science and trials, too. I spent 30+ years in this field and am considered an expert in design and development. Aside from regulatory controls during design and development, there are phases trials to minimize risks and unintended consequences for medicinal products. The risks and benefits are carefully evaluated in a step by step fashion. What did the development process and risk analysis look like for Uber's AD car? What were the test cases before they made Tempe population their guinea pigs? What approval process did they have to go thru with any government agency to take this product from the lab to the streets? Who decided the risks were acceptable? How? Upon what data?

This failure is not even at the ethical level. What would the AD vehicle do if being left crossed by a truck and there was a bicycle to the right? Take the left cross or kill the bicyclist? What if the oncoming left cross was a bicycle,

[FlashUNC]

The video is galling. The radar/lasers and all the other bells and whistles are supposed to avoid this kind of thing. That's essentially the entirety of why we're being sold on this kind of autonomous car future, so stuff like this doesn't happen.

I can understand this if this were a human driver given all the limitations of our senses. But it ain't like radar cares how dark it is out.

[MattTuck]

I thought of medical device development and pharmaceutical science and trials, too. I spent 30+ years in this field and am considered an expert in design and development. Aside from regulatory controls during design and development, there are phases trials to minimize risks and unintended consequences for medicinal products. The risks and benefits are carefully evaluated in a step by step fashion. What did the development process and risk analysis look like for Uber's AD car? What were the test cases before they made Tempe population their guinea pigs? What approval process did they have to go thru with any government agency to take this product from the lab to the streets? Who decided the risks were acceptable? How? Upon what data?

This failure is not even at the ethical level. What would the AD vehicle do if being left crossed by a truck and there was a bicycle to the right? Take the left cross or kill the bicyclist? What if the oncoming left cross was a bicycle,


Yeah, I'm with you on most of this. Medicine has this idea of informed consent. That, hey, something might be potentially dangerous, but if you understand the risks, we consider it fair because there may be some benefit to you or society.

This woman is essentially the victim of a 21st century Tuskeegee Study. That said, you can only test this technology so much in the lab before it needs real life experience to get better, and in that context, accidents are sure to happen. Even with more regulation, there is no guarantee that this stuff will work -- and even if it does work in general and make the streets safer than with human drivers, they may not drive accidents to zero... so there will still be costs.

That's why I think the best thing other than clear communication is a regulatory structure that makes companies pay up front for the eventual cost that is born by the victims. For each mile of road that your autonomous vehicle drives, it is $x (not sure of the exact amount) would be one approach. Another would be a flat fee for a license to test an autonomous vehicle within your city. I prefer the first, because it makes companies carefully consider how they will deploy, monitor and examine their test data. If you allow them to costlessly roam the city, each incremental mile matters very little, and so you have much more lax safety measures.

[Kontact]

Yeah, I'm with you on most of this. Medicine has this idea of informed consent. That, hey, something might be potentially dangerous, but if you understand the risks, we consider it fair because there may be some benefit to you or society.

This woman is essentially the victim of a 21st century Tuskeegee Study. That said, you can only test this technology so much in the lab before it needs real life experience to get better, and in that context, accidents are sure to happen. Even with more regulation, there is no guarantee that this stuff will work -- and even if it does work in general and make the streets safer than with human drivers, they may not drive accidents to zero... so there will still be costs.

That's why I think the best thing other than clear communication is a regulatory structure that makes companies pay up front for the eventual cost that is born by the victims. For each mile of road that your autonomous vehicle drives, it is $x (not sure of the exact amount) would be one approach. Another would be a flat fee for a license to test an autonomous vehicle within your city. I prefer the first, because it makes companies carefully consider how they will deploy, monitor and examine their test data. If you allow them to costlessly roam the city, each incremental mile matters very little, and so you have much more lax safety measures.

Maybe they should be street testing the technology with vehicles built out of pillows with bright flashing lights rather than 5000 pound Volvos.

[goonster]

you can only test this technology so much in the lab before it needs real life experience to get better, and in that context, accidents are sure to happen.

I can understand this attitude in the general public, but this is not an acceptable stance on engineering and project teams for this kind of technology.

We don't have the full story yet, but as an automation engineer who programs machines that could hurt people, this "failure to detect and respond" issue looks to me very much like something that simply cannot slip through review and testing.

[Climb01742]

I thought of medical device development and pharmaceutical science and trials, too. I spent 30+ years in this field and am considered an expert in design and development. Aside from regulatory controls during design and development, there are phases trials to minimize risks and unintended consequences for medicinal products. The risks and benefits are carefully evaluated in a step by step fashion. What did the development process and risk analysis look like for Uber's AD car? What were the test cases before they made Tempe population their guinea pigs? What approval process did they have to go thru with any government agency to take this product from the lab to the streets? Who decided the risks were acceptable? How? Upon what data?

As you rightly illustrate, medicine has a good template for clinical trials. A humane template. It's a template that technology companies could learn a lot from. Technology companies abhor regulation. They like being the wild west because that's how money is made fastest and biggest. But it's clear that the intersection of humanity and technology needs some regulation. Clinical trials might be a good place to look for guidance. That said, as this famous painting by Thomas Eakins, The Clinic of Dr. Agnew, shows, medicine had its pre-regulation trials too.

[ripvanrando]

Yeah, I'm with you on most of this. Medicine has this idea of informed consent. That, hey, something might be potentially dangerous, but if you understand the risks, we consider it fair because there may be some benefit to you or society.

This woman is essentially the victim of a 21st century Tuskeegee Study. That said, you can only test this technology so much in the lab before it needs real life experience to get better, and in that context, accidents are sure to happen. Even with more regulation, there is no guarantee that this stuff will work -- and even if it does work in general and make the streets safer than with human drivers, they may not drive accidents to zero... so there will still be costs.

That's why I think the best thing other than clear communication is a regulatory structure that makes companies pay up front for the eventual cost that is born by the victims. For each mile of road that your autonomous vehicle drives, it is $x (not sure of the exact amount) would be one approach. Another would be a flat fee for a license to test an autonomous vehicle within your city. I prefer the first, because it makes companies carefully consider how they will deploy, monitor and examine their test data. If you allow them to costlessly roam the city, each incremental mile matters very little, and so you have much more lax safety measures.

I mean no disrespect but that's the attitude from the marketing department. Without seeing what type of simulated texting they conducted, it is hard to speculate. But this sort of death is not just an improve the breed acceptable consequence. This is the second AD death.

[Kontact]

I totally expect future deaths from autonomous vehicles entering complex situations or conditions that cause the computer to make the wrong choice.


That isn't what happened here. Uber drove right into a human being on a clear, open road without hesitation. The Uber system, at least, is not ready for public street testing.

[Seramount]

just curious why the pedestrian exhibited no urgency to get out of the path of oncoming headlights.

[Tony T]

This woman is essentially the victim of a 21st century Tuskeegee Study. That said, you can only test this technology so much in the lab before it needs real life experience to get better, and in that context, accidents are sure to happen. Even with more regulation, there is no guarantee that this stuff will work -- and even if it does work in general and make the streets safer than with human drivers, they may not drive accidents to zero... so there will still be costs.

There is absolutely no excuse that Uber did not have controls in place to prevent the safety driver from keeping his hands from the wheel.
I really don't understand that at this stage in real-world testing that only one safety driver is used.

[Kontact]

just curious why the pedestrian exhibited no urgency to get out of the path of oncoming headlights.

How can you tell she didn't? She appears in the video to move 6 feet left to right in 1 second.

[fiamme red]

This video of the location shows that the road is actually pretty well lit: https://www.youtube.com/watch?v=CRW0q8i3u6E.

[FlashUNC]

Juxtapose Uber's approach here with Volvo's rather measured approach:

https://www.theverge.com/2017/12/14/16776466/volvo-drive-me-self-driving-car-sweden-delay

[Louis]

This video of the location shows that the road is actually pretty well lit: https://www.youtube.com/watch?v=CRW0q8i3u6E.

So I wonder why the Uber video appears to show a very dark road.?

Different types of cameras?

Editing of one or both videos?

Different lighting?

[Kontact]

So I wonder why the Uber video appears to show a very dark road.?

Different types of cameras?

Editing of one or both videos?

Different lighting?

Because it is a cheap little dash cam with a fisheye lens, not the system the vehicle uses to drive itself.

[Louis]

Because it is a cheap little dash cam with a fisheye lens, not the system the vehicle uses to drive itself.

I understand that, but I'm comparing the two "dash-cam" videos.

(there's been some speculation that Uber doctored the video from the Volvo, making it appear darker than it really was)

[Kontact]

I understand that, but I'm comparing the two "dash-cam" videos.

(there's been some speculation that Uber doctored the video from the Volvo, making it appear darker than it really was)

Is the second one a dash cam? Maybe the second one has better gain. But if you watch both, the Uber one is so dark it makes streetlights look like moons.


BTW, the 33 second mark in the second video is a clear view of the accident location, about the same spot as second 3 in the Uber video.

[Mr. Pink]

A lot seem to be fooled that all this is a sacrifice to making cars safer, which is a cruel irony in itself, but, consider this the first sacrifice to the capitalists and profiteers of the commercial transportation industry. It's not about safety, it's all about eliminating billions of truck and taxi driver jobs, and the trillions of dollars those jobs cost to the industry. It's also about monopoly power and profits, too. The company that wins first in this race, and, it is a ferverish race at the moment, wins. What's a few cyclists and pedestrians when you're talking about that kind of glory and money? This death will be far from the last. They'll find excuses to cover themselves.

[ripvanrando]

what was the purpose of a vehicle occupant?

One would suppose:

1. to observe autonomous driving patterns and to learn>>>to the lab

2. intercede if necessary

Wouldn't an engineer or technician be indicated rather than a Felon behind the wheel?

A quick read shows that Uber autonomous driving machines have had a lot of accidents (Tempe and Pittsburgh). Who monitors these trials? Who tests the vehicle's AD systems and sensors routinely during these trials?

[Mr. Pink]

what was the purpose of a vehicle occupant?

One would suppose:

1. to observe autonomous driving patterns and to learn>>>to the lab

2. intercede if necessary

Wouldn't an engineer or technician be indicated rather than a Felon behind the wheel?

A quick read shows that Uber autonomous driving machines have had a lot of accidents (Tempe and Pittsburgh). Who monitors these trials? Who tests the vehicle's AD systems and sensors routinely during these trials?


There's someone in the driver's seat because that seems to be the only requirement some freedom loving towns have put on these companies to allow them to test there. Otherwise, I highly suspect that, well, in this day and age of "regulation is bad", and, "free market good", there is no oversight. Now, if you have been following the sad tale of Uber as a company, and the sociopath who was its CEO, and still is its major shareholder, that does not bode well for the rest of us, especially cyclists.

[AngryScientist]

it's all about eliminating billions of truck and taxi driver jobs, and the trillions of dollars those jobs cost to the industry.

it's true.

we're in a desperate race to the bottom in a LOT of industries right now.

since i travel for work a bunch, i (obviously) need to get to the airport regularly. i used to use a car service - nice guy in a suit would come pick me up and drive me to airport. average $80 bucks a ride.

these days, i hit the uber or lyft button. average 18 - 22 bucks for the same ride.

better for me the customer, but that industry is getting bled dry. how the guy behind the wheel is making any real money driving me to the airport for 18 bucks is beyond me.

now we're about to cut the driver out of the loop entirely.

eliminating lower level jobs is not going to be good long term for civil unrest. people need to work. for the most part - people want to work and earn a living. i'm sure it wont be long before college kids cant find any work because robots flip hamburgers and deep fry potatoes.

i dont know what the answers are either, but the mounting problems are pretty clear.

[Mr. Pink]

it's true.

we're in a desperate race to the bottom in a LOT of industries right now.

since i travel for work a bunch, i (obviously) need to get to the airport regularly. i used to use a car service - nice guy in a suit would come pick me up and drive me to airport. average $80 bucks a ride.

these days, i hit the uber or lyft button. average 18 - 22 bucks for the same ride.

better for me the customer, but that industry is getting bled dry. how the guy behind the wheel is making any real money driving me to the airport for 18 bucks is beyond me.

now we're about to cut the driver out of the loop entirely.

eliminating lower level jobs is not going to be good long term for civil unrest. people need to work. for the most part - people want to work and earn a living. i'm sure it wont be long before college kids cant find any work because robots flip hamburgers and deep fry potatoes.

i dont know what the answers are either, but the mounting problems are pretty clear.


Most people don't know, because the mass media and even financial news outlets don't cover it, is that Uber is bleeding billions of dollars, and that low pricing is a major reason. It's practically predatory, and the expectation was/still is that their business model will succeed in eliminating licensed taxi drivers first, and eventually, with this tech, eliminate drivers entirely. And, there are a lot of very wealthy venture investors who have taken that bet. I think we all know who wins in the end when that kind of money is involved, although it didn't turn out too well for Theranos and Elizabeth Holmes, but, she can rest easy that she avoided jail time, which is a crime n itself.

Yes, truck driving is the last job a high school grad can do and live a middle class existence. We're talking millions of jobs, just poof. It's not going to be pretty when they're gone.

[AngryScientist]

Most people don't know, because the mass media and even financial news outlets don't cover it, is that Uber is bleeding billions of dollars, and that low pricing is a major reason. It's practically predatory, and the expectation was/still is that their business model will succeed in eliminating licensed taxi drivers first, and eventually, with this tech, eliminate drivers entirely. And, there are a lot of very wealthy venture investors who have taken that bet. I think we all know who wins in the end when that kind of money is involved, although it didn't turn out too well for Theranos and Elizabeth Holmes, but, she can rest easy that she avoided jail time, which is a crime n itself.

Yes, truck driving is the last job a high school grad can do and live a middle class existence. We're talking millions of jobs, just poof. It's not going to be pretty when they're gone.

how would you have likes to have bought a NYC taxi medallion in 2013-ish?

https://cbi-blog.s3.amazonaws.com/blog/wp-content/uploads/2015/09/taxi-medallion-prices-fall.jpg

[ripvanrando]

I pay the old guy in a suit for my rides to the airport. I'll never take an Uber for many reasons. Getting into a motor vehicle is the riskiest thing I do in my life and I'm not trusting it to someone taking 18 bucks because something is seriously wrong with that inequality.

[AngryScientist]

I pay the old guy in a suit for my rides to the airport. I'll never take an Uber for many reasons. Getting into a motor vehicle is the riskiest thing I do in my life and I'm not trusting it to someone taking 18 bucks because something is seriously wrong with that inequality.

on the flip side, and according to my experience - the amount of drunk driving that cheap, immediately available uber rides have eliminated - is really fantastic.

[fiamme red]

how would you have likes to have bought a NYC taxi medallion in 2013-ish?The value of a taxi medallion is now $175,000.

http://www.nydailynews.com/new-york/queens/cabbie-hangs-metal-cable-queens-garage-article-1.3888237

A Queens cabbie who drove a yellow taxi for three decades hanged himself in his garage after suffering massive financial woes in the era of Uber, officials and friends said Wednesday.

Nicanor Ochisor, 65, was found hanging from a wooden beam in his garage on 58th Road near 69th Lane in Maspeth Friday morning, police said.

Taxi advocates quickly blamed the Romanian immigrant’s suicide on the glut of drivers working for app-driven, for-hire companies like Uber and Lyft taking money from medallion drivers.

“He could no longer bear the strain of the impending loss of everything he had worked for in his life in America,” the Taxi Medallion Owner and Driver Association said in a statement.

The organization pointed out that in 2014, medallions were selling for more than $1 million. Now, the value has dropped to about $175,000.

The group said Ochisor is the fourth cab driver, and first medallion owner, to take his own life over the last few months...

[fiamme red]

So I wonder why the Uber video appears to show a very dark road.?

Different types of cameras?

Editing of one or both videos?

Different lighting?https://arstechnica.com/cars/2018/03/police-chief-said-uber-victim-came-from-the-shadows-dont-believe-it/

It's not surprising that the road was actually more brightly lit than the Uber video makes out. Think about it: the Uber car was going 38 miles per hour (61km/h), and people on pitch-black country roads drive faster than that all the time. That would be extremely reckless if—as the video implies—headlights can't illuminate the road two seconds ahead at that speed.One comment: "I lived in Tempe. It's not such a dark stretch of road that you would've missed her stepping onto the road, if you paid attention. Not only that but video footage is kind of deceptive - it's a pretty wide piece of road. Not like one of those European streets where you truly can surprise a driver by stepping out of an alley or parked cars."

[AngryScientist]

The value of a taxi medallion is now $175,000.

http://www.nydailynews.com/new-york/queens/cabbie-hangs-metal-cable-queens-garage-article-1.3888237

that's really sad, and i can not even fathom the despair one would have after paying over 1M for something (that you cant live in) now worth so little.

[MattTuck]

Taxi medallions were kept purposely scarce by the taxi industry, to keep supply constrained and prices high. Now, Uber is subsidizing to keep prices artificially low in an attempt to gain market share, change ingrained behaviors and cripple the cab industry.

In either case, it is a market failure.

[oldpotatoe]

on the flip side, and according to my experience - the amount of drunk driving that cheap, immediately available uber rides have eliminated - is really fantastic.

NOT trying to argue, honest but.

It's a claim the company has made repeatedly. On its website it says cities with Uber have "fewer drunk drivers on the streets." It recently used the argument in a heated battle over background checks in Austin.

But a new study in the American Journal of Epidemiology has found no noticeable impact on the number of drunk driving fatalities in cities where Uber runs.

http://money.cnn.com/2016/07/29/technology/uber-drunk-driving/index.html

and

http://fortune.com/2017/04/13/uber-drunk-drivers/

[Bradford]

that's really sad, and i can not even fathom the despair one would have after paying over 1M for something (that you cant live in) now worth so little.

I also travel a lot for business and have for 30 years. New York City cabs have always been awful and represent the worst transportation experience in the country. The value of a medallion has plummeted because they have been gouging customers for years with an inferior service at full price.

I usually rent a car when I travel, in large part because of some of the awful cab experiences I have had in New York and other big cities. However, the times I've take Uber have all been very nice experiences. Cost aside, which I'm indifferent to since I expense all my travel, Uber is a much better experience than big city Yellow Cabs.

The New York Cab industry had a predatory business model that lost in the marketplace as soon as someone figured out a way around their monopoly. I'm sure that there were some individual drivers with medallions, but my impression from talking to cabbies over the years is that most medallions were owned by companies and the drivers were not making a fortune. If that is the case, I'm actually pretty happy to see them get slaughtered.

[livingminimal]

I'm sure that there were some individual drivers with medallions, but my impression from talking to cabbies over the years is that most medallions were owned by companies and the drivers were not making a fortune. If that is the case, I'm actually pretty happy to see them get slaughtered.

I agree that if a company is exploiting the conditions and making things ****ty for their drivers, holding all the leverage of a medallion, etc, the ****ty experience will trickle down to the person using the service, and that sucks.

The problem (for me) is the second part of your statement; if were happy to see a piggish crappy exploitive industry with bad service be gutted, what happens when the replacement is basically the same thing (Uber as a company? no better) exploiting (The average Uber wage is pretty damn ****ty) people and providing highly-marginal service, with a series of news stories on ride sharing rapes, assaults, etc?

Both ****ty sides of the same parasitic, vulture capitalistic coin. Nationalize more transportation services, please.

[ripvanrando]

https://arstechnica.com/cars/2018/03/police-chief-said-uber-victim-came-from-the-shadows-dont-believe-it/

One comment: "I lived in Tempe. It's not such a dark stretch of road that you would've missed her stepping onto the road, if you paid attention. Not only that but video footage is kind of deceptive - it's a pretty wide piece of road. Not like one of those European streets where you truly can surprise a driver by stepping out of an alley or parked cars."

Thanks for that link. I found the Volvo lighting from the clip to be bizarre.....it made no sense......almost as much as that Uber speeding. 10% over the limit, why? Police report 40 mph, Uber says 38mph. 35 mph zone.

I was going to say the dashcam clip looked doctored but something is clearly wrong because my bike lights have 450 feet of throw and this dashcam makes it look like the bicycyist appeared out of nowhere that is not true although the Volvo light throw looks like 100-150 feet, this clearly can't be the case. Something does not add up.

[David Tollefson]

Why does the lighting in the videos even matter? Wasn't the car equipped with detection systems that don't rely on the visible light spectrum?

[tuscanyswe]

Why does the lighting in the videos even matter? Wasn't the car equipped with detection systems that don't rely on the visible light spectrum?

Yes for the accident its does not matter like you say.
But what it implies if tampered is that someone is allrdy trying to push public opinion by making it look like a very possible accident for any driver while infact it could look very different in reality. Perhaps even like an accident that would not have happened to a human driver?

[MattTuck]

Why does the lighting in the videos even matter? Wasn't the car equipped with detection systems that don't rely on the visible light spectrum?

It depends on what your thinking is. I think most people are trying to suggest that the autonomous driving system made a mistake, and the person was not recognized as an object in the road, and thus no actions were taken to avoid her.

And they're using the difference in video lighting to indicate that Uber either 1) deliberately made the video darker so as not to make it look so bad as hitting someone on a well lit road or 2) to indicate that the road was actually quite well lit, regardless of what you see in the video.

[ripvanrando]

Why does the lighting in the videos even matter? Wasn't the car equipped with detection systems that don't rely on the visible light spectrum?

LIDAR and Radar can identify an object but they can't see a red light or identify objects with precision. It would seem to me that LIDAR should have alerted and slowed the vehicle and the optical vision system should have confirmed that a human being was crossing the street and avoidance and braking algorithms engaged. Optical systems need visible light, which is clearly missing from the dashcam rendition. That is why light is important or at least in my opinion. If Uber doctored the dashcam video to make Volvo's lights look crappier than they already are, shame on them.

[ultraman6970]

I have rear almost nothing about the accident in question and the little i read gave me the impression that besides having a safety driver inside of the car that is/was a felon, the driver was distracted but at the same time if you see the video it was super dark and the lady pretty much was nowhere to be seen (need to check details of the area, speed of the car and stuff) aswell... dont take me wrong ok but im under the impression that uber wants to throw the felon under the bus for non stopping or looking at the road????

I have to drive a bus at night a couple of days a week and depending on the conditions you can;t see jack; in the video im under the impression the car is going way too fast for my taste and looked like it was a foggy night? Dunno man... one thing is to have accidents like this that could have happened to anybody but when you have like zillions of sensors and craaaaaaaaapp around the car that did not work how you can throw the driver (felon or not) under the bus?

The idea of autonomous cars sucks in practice IMO.

[Mark McM]

LIDAR and Radar can identify an object but they can't see a red light or identify objects with precision. It would seem to me that LIDAR should have alerted and slowed the vehicle and the optical vision system should have confirmed that a human being was crossing the street and avoidance and braking algorithms engaged. Optical systems need visible light, which is clearly missing from the dashcam rendition. That is why light is important or at least in my opinion. If Uber doctored the dashcam video to make Volvo's lights look crappier than they already are, shame on them.

I won't argue that the optical system iss needed to actually identify the object in the middle of the road - but shouldn't the LIDAR and/or Radar detecting that there is something in the road be enough to stop or go around the object? Are there any types of objects (of the general size of a human) that there is no need to avoid? As stated before, lighting conditions are no excuse for this crash.

[MattTuck]

I won't argue that the optical system iss needed to actually identify the object in the middle of the road - but shouldn't the LIDAR and/or Radar detecting that there is something in the road be enough to stop or go around the object? Are there any types of objects (of the general size of a human) that there is no need to avoid? As stated before, lighting conditions are no excuse for this crash.

I think you're saying two different things. To go around an object requires complex algorithms and intelligence to actually do. To just detect something in front of the car and just 'dumb' apply the brakes, that is much easier, and is probably something the Volvo should have done automatically... but perhaps the Volvo system is over-ridden by the autonomous driving system that Uber uses.

[72gmc]

Today's NYT looks into the state of this project prior to the fatality.

Everything uber does should be questioned.

[ripvanrando]

I won't argue that the optical system iss needed to actually identify the object in the middle of the road - but shouldn't the LIDAR and/or Radar detecting that there is something in the road be enough to stop or go around the object? Are there any types of objects (of the general size of a human) that there is no need to avoid? As stated before, lighting conditions are no excuse for this crash.

Motor cycle in opposite lane.

I can only guess that so much information needs to be processed and prioritized that like all/most failures, they are multifactorial and poor lighting could have been one chain in the failure link. Truly speculation on my part. I used to know something about imaging systems in the stone age.

[Kontact]

Motor cycle in opposite lane.

I can only guess that so much information needs to be processed and prioritized that like all/most failures, they are multifactorial and poor lighting could have been one chain in the failure link. Truly speculation on my part. I used to know something about imaging systems in the stone age.

Something in the opposite lane is already avoided.

The point here is that there are no man size objects that a car could drive into without concern. No ghosts, no radar reflecting columns of fog, no free roaming towers of whipped cream. Anything in the lane that big is something that needs to be avoided. The AI doesn't need to know whether it is a tree limb, a refrigerator or a person to know that hitting it isn't acceptable.

[ripvanrando]

Something in the opposite lane is already avoided.

The point here is that there are no man size objects that a car could drive into without concern. No ghosts, no radar reflecting columns of fog, no free roaming towers of whipped cream. Anything in the lane that big is something that needs to be avoided. The AI doesn't need to know whether it is a tree limb, a refrigerator or a person to know that hitting it isn't acceptable.

Until it isn't. How are objects prioritized. What is the computational delay especially if the optical system is compromised. Another example would be a car door opening. Parked cars would have to be de-prioritized.

That is the point

[Kontact]

Until it isn't. How are objects prioritized. What is the computational delay especially if the optical system is compromised. Another example would be a car door opening. Parked cars would have to be de-prioritized.

That is the point

Parked cars in the path of the vehicle would totally be a priority. I don't follow the point you're making - a correctly designed system isn't going to choose to drive into something. The pedestrian was moving slowly enough that there is no reason the Uber shouldn't have regarded it as an object in its path without even needing to guess where she was going.

[FlashUNC]

Interesting fact for comparison I saw on Jalopnik: Waymo, Google's autonomous car effort, has traveled 5,600 miles and counting without human intervention needed. Ubers program has failed to meet it's own stated goal of traveling more than 13 miles without requiring human intervention.

Their system is a joke and shouldn't be on public roads.

[HenryA]

An autonomous vehicle is going to do things a human won’t do, yet it is expected to interact with humans on and around the road. Drive the posted speed limit using various sensors that allow the car to “see” better than human? Why, yes. But the humans around it still can’t see any better than before such a device hit the road. Is it legal to drive the posted limit when conditions might dictate to a human to slow down? Questionable, perhaps. Will the machine make such a decision?

Until roads are redesigned to accommodate autonomous vehicles and eliminate human interaction with said AVs I don’t want the damned things around me. Not at all.

As for liability for an AV, I look at them like a “set-gun”. The owner/user should be strictly liable for any and all harm done by the device they operate in public.

Someone will no doubt gain great wealth from the production of AVs but the general public will suffer and suffer badly. They’re coming, but you won’t like it.

[pasadena]

poor lighting could have been one chain in the failure link.

Uber's uses Lidar

The president of Velodyne Lidar (makes the sensors in the Uber car) already said the system should have 'seen' the victim crossing the street and avoided them.
The system is designed specifically to avoid this exact type of situation.
They do not understand why the Uber car did not stop.

I have no faith in current technology. They fail even in cars now, that are not even close to autonomous... and they use those same sensors for autonomous cars.
Snow, freezing rain, debris, dirt that covers sensors, objects that "confuse" systems...
I can point to two specific
1. The Mercedes caravan cars that self-slammed on the brakes riding near the peloton, and Mark Cavendish slammed into the back of (forget the race, a few weeks back)

2. auto-braking sensors that slam on the brakes in curving, tight roads when it 'sees' the guardrail or lumpy road and tight turn (Mulholland Drive)

those are just ones I know about but i have read of cars slamming on brakes on the freeway for no reason among a host of other issues.

[Louis]

As drivers and cyclists, how high do we think the bar should be set for this type of vehicle?

1) Very close to perfection? (Assumes that 100% perfect isn't possible)
2) Maybe not 100% perfect, but still pretty darn good?
3) At least as good as the average human?
4) As good as a so-so human?
5) Same bar as humans, so as good as the very worst humans who are currently allowed to drive?

[ripvanrando]

Velodyne Lidar's CEO does not know how Uber has integrated Radar, Lidar, and optical cameras nor do any of us. There is now way possible that he or she knows how the Lidar images are integrated into the entire system or the overall decision making process therein. Catastrophic failures usually require more than one system breakdown.

That is my point. Yes, I am speculating.

How did Uber's system categorize the cyclist/ped initially and were any one of the three sensors malfunctioning or did poor vehicle headlights contribute. Was the pedestrian mis-categorized and then deprioritized until it was too late. Uber's AV could have initially classified the pedestrian as say a bush or even parked shopping cart and then because it was speeding (using old Google maps with earlier higher speed limit) and due to the massive computational workload either did not re-classify or needed to confirm using the optical system. How does Uber's AV system avoid false positive results when there is ambiguity? Or does it just slam on the brakes for every snowflake or tree branch falling from the sky? Why misclassification? Does Uber's catalogue of imagery include homeless person's bike riddled with shopping bags (looks kind of like a bush)? How about a tandem? A velomobile? Recumbent? Toddler's bigwheel? Tesla killed a driver by misclassifying a semi-truck as an overpass.

Or the developers could have simply put the risk to the passengers ahead of the pedestrian and the Volvo took no action. This has been my fear as a cyclist, AV systems will just run us over by design.

Why wasn't a driver fatigue and alertness monitoring system installed? To verify that the Felon was looking? I know German cars have them.

I don't want AV on the road, either. Not a fan.

[binxnyrwarrsoul]

I keep reading that the "accident" would have happened regardless, even if there was a driver with no "nannies" driving. That, imo is a flawed opinion. As someone who has driven some times for 12+ hrs a day, commuting and driving a 33K lb, 40 ft. truck (22 yrs) on narrow streets in NYC, in all weather conditions, starting at 1AM, there has to be an account for experience, and of the natural impetus to be constantly looking/scanning/ the road (sometimes unconsciously) and a small area on both sides of the hypothetical lane. And always scanning mirrors. Would someone with that kind of experience and way of driving have avoided the accident? No one can answer that. I have driven countless miles/hrs in the last 30 yrs (pro and leisure), and more often than not I have had to deal with something on the street that is new and requires a skillset a machine can't replicate, unless that particular situation and dealing with it is programmed into the computer. And there's millions of those particular situations. Computers can be programmed to replicate human behavior, reactions etc. almost to the letter. How much "almost" is the unanswerable question.

[tuscanyswe]

I keep reading that the accident would have happened regardless, even if there was a driver with no "nannies" driving. That, imo is a flawed opinion. As someone who drives some times for 12 hrs a day, commuting and driving a 33K lb, 40 ft. truck (22 yrs) on narrow streets in NYC, in all weather conditions, starting at 1AM, there has to be an account for experience, and of the natural impetus to be constantly looking/scanning/ the road (sometimes unconsciously) and a small area on both sides of the hypothetical lane. And always scanning mirrors. Would someone with that kind of experience and way of driving avoided the accident? No one can answer that. I have driven countless miles/hrs in the last 30 yrs (pro and leisure), and more often than not I have had to deal with something on the street that is new and requires a skillset a machine can't replicate, unless that particular situation and dealing with it is programmed into the computer. And there's millions of those particular situations. Computers can be programmed to replicate human behavior, reactions etc. almost to the letter. How much "almost" is the unanswerable question.

I dont for one second believe that this "accident" was something that could not have been avoided by a normal driver. No way!

[ripvanrando]

I dont for one second believe that this "accident" was something that could not have been avoided by a normal driver. No way!

Ditto.

Even with Volvo's lousy lights, there was 2-3 seconds (120-180 feet) to react and from 40 mph, most cars can be stopped.

[ultraman6970]

What bothers is that all over the press they are trying to blame the felon inside of the car. Their systems did not work, there's no grey line in here... as binks says... there's some stuff a machine can't replicate and thats the reason of the driver inside the car. Both failed...

Self driving cars works only in the movies man...

[binxnyrwarrsoul]

....It's not about safety, it's all about eliminating billions of truck and taxi driver jobs, and the trillions of dollars those jobs cost to the industry. It's also about monopoly power and profits, too. The company that wins first in this race, and, it is a ferverish race at the moment, wins...........

This, a thousand times this.

[Kontact]

I'm still wondering why Uber's dash video is so dark. How did it get that way?

[fiamme red]

This article argues that it's unreasonable to expect safety drivers in self-driving cars to stay vigilant at all times.

https://www.wired.com/story/uber-crash-arizona-human-train-self-driving-cars/

We are good drivers when we’re vigilant. But we’re terrible at being vigilant. We get distracted and tired. We drink and do drugs. We kill 40,000 people on US roads every year and more than a million worldwide. Self-driving cars are supposed to fix that. But if we can’t be trusted to watch the road when we’re actually driving, how did anyone think we’d be good at it when the robot’s doing nearly all the work?

[Mr. Pink]

This, a thousand times this.

It's kinda remarkable to me that many think that all of these tech companies and most of the large car companies are investing billions of dollars so that the roads will be safer. If it was the insurance industry, that would make some sense. But, really? C'mon.

[Kontact]

It's kinda remarkable to me that many think that all of these tech companies and most of the large car companies are investing billions of dollars so that the roads will be safer. If it was the insurance industry, that would make some sense. But, really? C'mon.

The tech companies are doing it to break into the auto industry (or to eliminate paying drivers). The car companies are doing it to compete with the tech industry.

Everyone probably realizes that when the technology is proven, legislation mandating safer, self driving cars will follow. It is all just a matter of time at this point.

[rwsaunders]

Well they haven't built cars totally by hand since what, 1961, when robots started doing spot welding? Over half of the industrial robotic purchases are made by the automotive industry, so it's only natural that the industry would work to perfect improved driving technology.

[ripvanrando]

Uber originally had two occupants during these trials. One monitored data and systems. The other watched the road and took control over from the AV when necessary (about once every 13 miles). Uber eliminated one of the occupants. It is possible that the Felon was monitoring systems as required by Uber but this will all come out.

[ripvanrando]

Well they haven't built cars totally by hand since what, 1961, when robots started doing spot welding? Over half of the industrial robotic purchases are made by the automotive industry, so it's only natural that the industry would work to perfect improved driving technology.

That is like saying we have i-Phones because the ENIAC was up and running in what 1950 or 1948? The first welding prototypes were 1962 but it was at least 2 decades later until the technology was fully ripe to be implemented whereas Uber is shipping green bananas and shoving unproven, unregulated, dangerous technology upon us.

[HenryA]

A robot welding cell doesn’t have people walking through its work space and its certainly not open to the general public. Every piece of it is planned, engineered and guarded. If a person has to enter the area its shut down and locked out by industry policy and government regulation. Compare/contrast with public roadways.

Thought experiment:
Imagine you are driving along a city street and notice over the top of the trunk of one of the cars parked on the side of the road, a child’s hat bobbing up and down with a pom-pom on top. Just enough for you to recognize a child near the road and not much else. A normally vigilant person would put that together and slow down so they don’t run over a kid who darts out between parked cars. Does the AV do that? Not likely. And a kid dies.

[tuscanyswe]

A robot welding cell doesn’t have people walking through its work space and its certainly not open to the general public. Every piece of it is planned, engineered and guarded. If a person has to enter the area its shut down and locked out by industry policy and government regulation. Compare/contrast with public roadways.

Thought experiment:
Imagine you are driving along a city street and notice over the top of the trunk of one of the cars parked on the side of the road, a child’s hat bobbing up and down with a pom-pom on top. Just enough for you to recognize a child near the road and not much else. A normally vigilant person would put that together and slow down so they don’t run over a kid who darts out between parked cars. Does the AV do that? Not likely. And a kid dies.

But if there wasent a crossing the kid was in the wrong! doh..
Same with drunks etc

There are so many scenarios every day when i ride in traffic where i can see where a car or person is going that has very little to do with their previous movement which as far as i can tell is what ai bases their projectories on alone (thus far)? There are so many things we pick up on about where n what ppl are going to go or to do next that we are not even aware we are processing.

What about sound btw. do ai use sound in any way thus far?

[rwsaunders]

This is a good read and I appreciate the segment which talks about the basic human fear of the unknown. An excerpt...

So why are we so afraid? I asked Prof. Luciano Floridi, Professor of Philosophy and Ethics of Information at the University of Oxford, that very question.

“I think it’s an ancestral fear of the unknown, fear of machines. It was the Golem and Frankenstein, and now it’s a technology that we don’t fully control in the case of driverless cars,” Professor Floridi tells me. “It also seems to be, perhaps, a mis-selling on the side of the industry, where total safety is sold as a possibility, which we know is not the case. Nothing is 100% safe, you can only be safer.”

So even though driverless cars are safer and reduce a person’s chance of death, drivers struggle to accept the idea that they could be killed by a machine, rather than by human error.

http://www.trustedreviews.com/news/driverless-car-deaths-insurance-blame-uber-tesla-2945865

[verticaldoug]

Uber originally had two occupants during these trials. One monitored data and systems. The other watched the road and took control over from the AV when necessary (about once every 13 miles). Uber eliminated one of the occupants. It is possible that the Felon was monitoring systems as required by Uber but this will all come out.

Waymo's control AV incidents are 1 per 5000 miles. I really think this is another case of UBER pushing with no regard to safety.

[ripvanrando]

Waymo's control AV incidents are 1 per 5000 miles. I really think this is another case of UBER pushing with no regard to safety.

Agree.

I struggled thinking why people think LIDAR "sees".

A LIDAR image is not like a picture or portrait off an optical camera. LIDAR produces a bunch of dots spaced kinda fair apart. The processing and decision algorithm and the extent to which UBER relies upon either radar or optical cameras is unknown. Classifying and prioritizing all those dots is a hugely challenging. Uber failed under the most ideal conditions. LIDAR should work better at night than day AND there were no other vehicles and the background was not busy and nothing moving.

1 intervention every 13 miles? Wow. Total fail. Get it off the road.

[Kontact]

Agree.

I struggled thinking why people think LIDAR "sees".

A LIDAR image is not like a picture or portrait off an optical camera. LIDAR produces a bunch of dots spaced kinda fair apart. The processing and decision algorithm and the extent to which UBER relies upon either radar or optical cameras is unknown. Classifying and prioritizing all those dots is a hugely challenging. Uber failed under the most ideal conditions. LIDAR should work better at night than day AND there were no other vehicles and the background was not busy and nothing moving.

1 intervention every 13 miles? Wow. Total fail. Get it off the road.

I don't understand why you would say that LIDAR doesn't "see" but a computer scanning a video camera does. In neither case does the AI "understand" what those pixels or dots mean in the way an animal does, and neither one paints a more realistic version of reality than the other.

What LIDAR does do is provide more accurate range to object information than video does, even binocular video. If all you needed to accomplish was to avoid hitting static objects, LIDAR would be sufficient.

[merckxman]

Tesla Says Crashed Vehicle Had Been on Autopilot Before Fatal Accident https://nyti.ms/2GoO7Ss

[ripvanrando]

I don't understand why you would say that LIDAR doesn't "see" but a computer scanning a video camera does. In neither case does the AI "understand" what those pixels or dots mean in the way an animal does, and neither one paints a more realistic version of reality than the other.

What LIDAR does do is provide more accurate range to object information than video does, even binocular video. If all you needed to accomplish was to avoid hitting static objects, LIDAR would be sufficient.


Because LIDAR doesn't see per se. I don't remember ever saying a computer scanning a video does nor is it clear how that was construed.

I explained it well enough for the lay person.

Aside from a potential sensor issue (doubtful), the series of dots from LIDAR was processed and the object was mis-identifed by the "computer"......the question I raised was how does the Uber system handle ambiguity? Does the video images come into the decision making and as I had said, none of us knows how that system works. A hunched over pedestrian pushing a bike with shopping bags was probably not classified. Uber system either thought it was a bush or something or just said, **** it, protect occupants.

[ripvanrando]

Tesla Says Crashed Vehicle Had Been on Autopilot Before Fatal Accident https://nyti.ms/2GoO7Ss

Is that three now for Tesla?

[merckxman]

Yes, just read:

"Concern has grown over the safety of Autopilot, its driver-assistance system, after a fatal crash on March 23 that occurred while the system was engaged. It was at least the third fatal crash that has taken place while a driver was using Autopilot."

Is that three now for Tesla?

[azrider]

Settlement issued. Experts saying $1-$3M expected payout.

https://www.azcentral.com/story/news/local/tempe/2018/03/29/uber-settlement-self-driving-car-death-arizona/469278002/

[Kontact]

Because LIDAR doesn't see per se. I don't remember ever saying a computer scanning a video does nor is it clear how that was construed.

I explained it well enough for the lay person.

Aside from a potential sensor issue (doubtful), the series of dots from LIDAR was processed and the object was mis-identifed by the "computer"......the question I raised was how does the Uber system handle ambiguity? Does the video images come into the decision making and as I had said, none of us knows how that system works. A hunched over pedestrian pushing a bike with shopping bags was probably not classified. Uber system either thought it was a bush or something or just said, **** it, protect occupants.

I understand the difference between the two pieces of gear, but in neither case does the AI "understand" the output of either in a way that is like what we call "seeing". Your comparison appears to give the AI credit for being able to "read" a video image like we do. Some groups of human beings and most animals can't understand video images as representing reality, and AI certainly abstract it differently than we do.

There is nothing about a machine looking at video that is implicitly better than a machine scanning radar or lidar returns.

[ripvanrando]

I understand the difference between the two pieces of gear, but in neither case does the AI "understand" the output of either in a way that is like what we call "seeing". Your comparison appears to give the AI credit for being able to "read" a video image like we do. Some groups of human beings and most animals can't understand video images as representing reality, and AI certainly abstract it differently than we do.

There is nothing about a machine looking at video that is implicitly better than a machine scanning radar or lidar returns.

If that is how you interpreted what I said, I am sorry but not what I meant.

[ripvanrando]

Tesla crash recreated.

A Tesla owner drove the same route under the same conditions, his Tesla would have driven him to his death into the barrier, too.


https://www.youtube.com/watch?v=6QCF8tVqM3I

[tuscanyswe]

Tesla crash recreated.

A Tesla owner drove the same route under the same conditions, his Tesla would have driven him to his death into the barrier, too.


https://www.youtube.com/watch?v=6QCF8tVqM3I

Wow thats interesting. And we have these things riding around among us all the time in city traffic.

[Kontact]

If that is how you interpreted what I said, I am sorry but not what I meant.

I was just trying to understand the line you were drawing between two systems that process received light to build a 3D model of objects. You seemed to be saying that their receipt by the AI was different in an important way, and that "seeing" is implicitly different between received LIDAR bounces and received ambient light in terms of how an AI would be able to use them.

[pasadena]

Settlement issued. Experts saying $1-$3M expected payout.

https://www.azcentral.com/story/news/local/tempe/2018/03/29/uber-settlement-self-driving-car-death-arizona/469278002/

Recall coordinators formula already applied
AxBxC=X if X is < than a recall, carry on killing people.

https://www.youtube.com/watch?v=SiB8GVMNJkE

[unterhausen]

Wow thats interesting. And we have these things riding around among us all the time in city traffic.

the problem is the people that believe the name "autopilot." Tesla should rename it "keep your hands on the wheel and your eyes out the windshield because you are driving"

Tesla says the crash was the highway department's fault for not putting the barrels back after a previous accident. I say they failed to pick up on a faint line on the road and line following is a stupid way to drive a car in the first place. Uber is lucky one of their autonomous cars didn't get to that intersection first.

[ripvanrando]

I was just trying to understand the line you were drawing between two systems that process received light to build a 3D model of objects. You seemed to be saying that their receipt by the AI was different in an important way, and that "seeing" is implicitly different between received LIDAR bounces and received ambient light in terms of how an AI would be able to use them.

No. No line being drawn.

ambiguity or classification could be due to several reasons.

What if LIDAR system failed, do we know whether the video system doesn't or could act as a backup. How Uber uses both systems is unknown......that is what I said.

These systems are not ready for prime time and the lack of oversight over us guinea pigs is disgusting. Not really interesting in discussing the technology any further. As I said, I'm an old dinosaur on this stuff. Prefer to be a living one.

[merckxman]

I used the LAB link *posted on page 3 of this thread* to contact my NJ US Senators. This is the reply I received today from Senator Bob Menendez:

"Thank you for contacting me to express your concerns regarding the American Vision for Safer Transportation through Advancement of Revolutionary Technologies (AV START) Act. Your opinion is very important to me, and I appreciate the opportunity to respond to you on this critical issue.

As you may know, this legislation aims to modernize federal motor vehicle regulations in order to facilitate the deployment of autonomous passenger vehicles onto our roadways. As traffic fatalities, including pedestrian and cyclist deaths, continue to rise, many have lauded the ability of autonomous vehicles to improve safety and eliminate human error. However, I have also heard concerns from some that the AV START Act could compromise safety by providing exemptions to federal motor vehicle safety standards to some autonomous vehicles. However, it should be noted that under current law, exempted vehicles must demonstrate equivalent levels of safety to non-exempt vehicles. The AV START Act was passed by voice vote by the Senate Commerce Committee, of which I am not a member, but has not yet come to the floor of the Senate for consideration.

Throughout my tenure in Congress I have been a staunch advocate for motor vehicle safety. I have fought hard to ensure that commercial drivers get the proper amount of rest before operating motor vehicles, called for improved data collection to better inform best vehicle and pedestrian safety practices, and have called on the National Highway Traffic Safety Administration to expeditiously implement the impaired driving prevention provisions of the Fixing America’s Surface Transportation (FAST) Act.

Please be assured that the safety of New Jerseyans and all Americans remains my top priority. I will continue to support policies that make our roadways safer for all users, and will keep your thoughts squarely in mind should the AV START Act come before the Senate for a vote."

[azrider]

Woman facing vehicular manslaughter charges after it was determined driver was streaming "The Voice" on Hulu at the time of the accident. :crap:



https://www.reuters.com/article/us-uber-selfdriving-crash/uber-driver-was-streaming-hulu-show-just-before-self-driving-car-crash-police-report-idUSKBN1JI0LB

[unterhausen]

I have doubts Uber told her she was responsible for stopping the car if anything out of the ordinary happened. They are culpable as well

[azrider]

I have doubts Uber told her she was responsible for stopping the car if anything out of the ordinary happened. They are culpable as well

Hmmmm......not sure about that. I'm positive Uber had an army of attorneys writing the contracts these 'operators' signed. In fact i found this excerpt from the article. I'm no attorney but I would think this would absolve Uber from a lot of the blame........no??

Last month, an Uber spokeswoman said the
company was undergoing a "top-to-bottom safety
review," and had brought on a former federal
transportation official to help improve the company's
safety culture. The company prohibits the use of any
mobile device by safety drivers while the self-driving
cars are on a public road, and drivers are told
they can be fired for violating this rule.