Uber's Self-Driving Cars Were Struggling Before Arizona Crash (nytimes.com)

An anonymous reader quotes a report from The New York Times: Uber's robotic vehicle project was not living up to expectations months before a self-driving car operated by the company struck and killed a woman in Tempe, Ariz. The cars were having trouble driving through construction zones and next to tall vehicles, like big rigs. And Uber's human drivers had to intervene far more frequently than the drivers of competing autonomous car projects. Waymo, formerly the self-driving car project of Google, said that in tests on roads in California last year, its cars went an average of nearly 5,600 miles before the driver had to take control from the computer to steer out of trouble. As of March, Uber was struggling to meet its target of 13 miles per "intervention" in Arizona (Warning: source may be paywalled; alternative source), according to 100 pages of company documents obtained by The New York Times and two people familiar with the company's operations in the Phoenix area but not permitted to speak publicly about it. Yet Uber's test drivers were being asked to do more -- going on solo runs when they had worked in pairs. And there also was pressure to live up to a goal to offer a driverless car service by the end of the year and to impress top executives.

Comments

Pressured to proceed despite poor test results....

[hey!]

Why does that sound so familiar?

Oh, wait. I'm a software developer.

Self driving car hype

[Teckla]

Self driving cars are mostly hype. They're primarily self driving on very good, very clean, very well mapped roads only. Take them out of perfect conditions, and they fail miserably.

That being said, the technology is still cool, even though it has a long, long way to go. A lot of the technology could eventually be incorporated into normal everyday cars to help human drivers avoid accidents.

But the hype, at this point, is kind of out of control.

Re:Self driving car hype

[Rei]

Right. So it's perfectly okay for a "self driving car" to have a LIDAR that doesn't work, a radar that doesn't work, cameras that can't see at night and/or a neural net that doesn't work, and ultrasonic sensors don't work, and to have the "self driving car" rely on a person who's not been driving being suddenly instantly able to hop into "driving mode" during each of the once-in-every-1500-mile occurrences where the car tries to crash itself without warning. Got it! This is all totally okay.

Re:Self driving car hype

[toonces33]

A human driver would have seen her and not hit her.

Re:Self driving car hype

[Anonymous Coward]

This has already been known. To quote an excerpt from a book published on June 19, 1973...

“Stupidity cannot be cured. Stupidity is the only universal capital crime; the sentence is death. There is no appeal, and execution is carried out automatically and without pity.” -- Robert A. Heinlein

Re:Self driving car hype

[aaarrrgggh]

Victim was already crossing the road, in the adjacent lane, as the car approached. The condition was caused by a bad self driving car that should apparently never been on the road.

That the root cause was someone jaywalking doesn't change the fact that jaywalking happens and is a predictable event. It also suggests that the car is inadequately prepared for avoiding hazards. (It also doesn't change the fact that when you don't have crosswalks at reasonable intervals people will improvise.)

Re: Self driving car hype

[bill_mcgonigle]

The problem with the hype is that high early expectations of perfection will drive fear which may lead to regulations that will ultimately cause more deaths. Give it ten years and this stuff will probably outperform human drivers, but watch one kid chase a ball out in front of a robot car and e.g. Utah will ban the technology.

Re:Self driving car hype

[AC-x]

And what if the obstacle was a fallen tree, or a large animal, or anything else that could be fatal to the car's occupants? I guess they shouldn't be so entitled to expect a self-driving car not to kill them either right?

Re:Self driving car hype

[fluffernutter]

It's not up to Uber or any self driving company to make stupidity a capital offense.

Re:Self driving car hype

[JonnyCalcutta]

Jaywalking doesn't even exist in the rest of the world. It is some weird concept spoken about in movies and TV shows from the US.

For example, in the UK we would call it 'crossing the road' and 'crossing the road' is not illegal (it is expected that pedestrians will use common sense). The idea that 'crossing the road' could be illegal is very strange to someone from the UK.

Re:Self driving car hype

[Bert64]

As things in the world are made safer, people seem to lose the ability to look out for their own safety and become totally reliant on others...
Not to mention arrogance, many cyclists have an attitude problem and disdain for both pedestrians and vehicles.

So yes stupidity should be punished, why should society bear the burden of protecting people too stupid to look out for themselves?

Re: Self driving car hype

[Anonymous Coward]

You apparently don't work in the industry. I do. The "hype" is to attract investors. It's purely for money. The actual timelines I've seen from the companies developing the sensors don't even suggest autonomous driving is possible in 10 years. There aren't any sensors that can detect common road obstacles, and there are no known solutions. A child is the road? Can't see it. This is by their own admission. Remember the time when a car crashed head on into a semi-truck? Can't see them either - they admitted that too. Couple that with the idea of human test drivers that think that they can take their eyes off the road, there will be accidents. The only reason why there aren't more accidents is very likely because humans prevent them by watching the road. The whole "testing" and stat reporting is very flawed, because it is still mainly humans preventing accidents. The way to improve "autonomous" stats, is to drive on the nice, straight, wide roads in Tempe. It's obvious why they drive on the roads here. I know the valley well. But these are over-glorified power-steering systems. Anybody today has a car that can do that. Hype indeed.

Re:Self driving car hype

[burtosis]

I agree. Look at how well lit the stretch of road is at night The camera view uber showed us was either purposefully darkened (just look at a histogram of a random frame or two), or the camera had an insufficient light sensitivity. The driver, had she not been texting but paying attention, would have had 5-6 seconds to do something. That said, I agree putting a human safety driver in the seat for 200 hours doing nothing except take over on a seconds notice is safety theater.

Re:Self driving car hype

[burtosis]

Worse, the scene was actually well lit to the point this should be vehicular homicide

Re:Self driving car hype

[Kjella]

Jaywalking doesn't even exist in the rest of the world. It is some weird concept spoken about in movies and TV shows from the US. For example, in the UK we would call it 'crossing the road' and 'crossing the road' is not illegal (it is expected that pedestrians will use common sense). The idea that 'crossing the road' could be illegal is very strange to someone from the UK.

Germany has the same rule, I think. Came as a surprise to us from Norway, but we crossed so far from traffic nobody fined us or anything. But then Germany is notorious for having rules for everything and actually sticking to them. It's kinda nice and incredibly frustrating at the same time, depending on what side of the stick you're on.

Re:Self driving car hype

[Bing Tsher E]

Or a toddler.

Which won't work here. We hate children on Slashdot.

Re:Self driving car hype

[Anonymous Coward]

As things in the world are made safer, people seem to lose the ability to look out for their own safety and become totally reliant on others...
Not to mention arrogance, many cyclists have an attitude problem and disdain for both pedestrians and vehicles.

So yes stupidity should be punished, why should society bear the burden of protecting people too stupid to look out for themselves?

Social mores are strange contradictory things. If someone achieves something, the narrative that this was the result of their skill and hard work is generally supported. If someone does something stupid and suffers a predictable outcome, well suddenly they're a poor helpless victim of random chance. Can't have this one both ways.

Re:Self driving car hype

[Anonymous Coward]

Jaywalking doesn't even exist in the rest of the world. It is some weird concept spoken about in movies and TV shows from the US.

That's because it's an invented concept created by car companies to trick people into thinking that roads are only for cars and not for walking. No, seriously: the origin of the term comes from "jay drivers" and has to do with the fact that "jay" was 1900s American slang for effectively "idiot." Drivers of the time were frequently called "jay drivers" and were seen as a menace. After all, originally, there were very few cars (as they were expensive) and quite a lot of pedestrians and horses (as that's how people had moved around for millennia).

To combat this negative view of drivers, car companies invented "jay walker" and started heavily promoting it, trying to convince Americans that they were seriously stupid if they thought that roads were for people.

And, as you may be able to guess, they succeeded.

Re:Self driving car hype

[amiga3D]

And yet if a human had hit her they wouldn't have been judged to be at fault. Over 4000 pedestrians get killed and another 50,000 injured every year and many of those are because of their own error. The real question here is would it have made any difference if it had been a human driving the car. Most likely not. I personally have barely avoided killing an individual running on the side of the road in the dark. I managed to miss them by 2 or 3 feet and spent the next few miles running 20 mph below the speed limit looking for more fools running on a highway before dawn. Then I realized that I was going to kill someone else as impatient rush hour traffic blew by me in unsafe passing areas. Some people seem to have death wishes.

Re:Self driving car hype

[fluffernutter]

I think that's obvious, but it doesn't change the fact that self driving should be held to a standard where they can accommodate people doing stupid things, otherwise the whole line about saving lines is just blowing smoke. Defensive driving is a 'thing' for good reason and very little of it changes if a computer is driving. If you see something curious, you slow down. Simple?

Re:Self driving car hype

[goose-incarnated]

And yet if a human had hit her they wouldn't have been judged to be at fault.

Yes, they would. Drivers are routinely held to be responsible for hitting almost stationary objects in the middle of the road. That is irrelevant anyway. What's relevant is that the car did not see an almost stationary object in the middle of the road. That the object happened to be an old lady pushing a bike is irrelevant.

Re:Self driving car hype

[Knuckles]

Germany has the same rule, I think. Came as a surprise to us from Norway, but we crossed so far from traffic nobody fined us or anything. But then Germany is notorious for having rules for everything and actually sticking to them. It's kinda nice and incredibly frustrating at the same time, depending on what side of the stick you're on.

In Germany it depends, but is is not illegal to cross the road. The law is:

(3) Wer zu Fuß geht, hat Fahrbahnen unter Beachtung des Fahrzeugverkehrs zügig auf dem kürzesten Weg quer zur Fahrtrichtung zu überschreiten. Wenn die Verkehrsdichte, Fahrgeschwindigkeit, Sichtverhältnisse oder der Verkehrsablauf es erfordern, ist eine Fahrbahn nur an Kreuzungen oder Einmündungen, an Lichtzeichenanlagen innerhalb von Markierungen, an Fußgängerquerungshilfen oder auf Fußgängerüberwegen (Zeichen 293) zu überschreiten. Wird die Fahrbahn an Kreuzungen oder Einmündungen überschritten, sind dort vorhandene Fußgängerüberwege oder Markierungen an Lichtzeichenanlagen stets zu benutzen.

-- https://www.gesetze-im-interne...

Translation:
Someone walking on foot has to cross roadways, while heeding vehicle traffic, speedily on the shortest path perpendicular to driving direction. If the density of traffic, speed of traffic, visibility conditions, or the flow of traffic require it, a roadway must only be crossed at road intersections, at traffic lights, on the inside of markings, or at pedestrian crosswalks. If crossing the roadway at road intersections, any available pedestrian crossings or markings at traffic lights must be used.

I.e., you cannot simply cross just anywhere on a high-level road with dense, fast traffic (think Autobahn) or in really bad visibility like dense fog. But a road like in the accident video, even if it's dark, is just fine if there is no dedicated crossing nearby

Re:Self driving car hype

[Knuckles]

What's relevant is that the car did not see an almost stationary object in the middle of the road. That the object happened to be an old lady pushing a bike is irrelevant.

Exactly. And any modern non-autonomous car with "simple" collision avoidance system would have noticed. Dunno what Uber is doing, but it does not come as a surprise that it's Uber who are the most irresponsible and reckless

Re:Self driving car hype

[aaarrrgggh]

I can think of a few different kinds of Jaywalking: indifference by walker, convenience to the walker, indifference by planners, and just crossing the street when there is a reasonable break in traffic.

The indifference kind (think San Francisco) is really inexcusable. The convenience kind is a shame, but kind of reality-- going a couple blocks out of your way to cross at a marked intersection rather than an unmarked intersection can be a bit of a pain. In an area where there should be an expectation of walkability, I am happy to give it a pass.

The planning type of indifference -- especially pisses me off though. This is the type of crap that makes it nearly impossible to walk anywhere. I remember a planning commission actually *banning* sidewalks for some insane reason. The same concern holds true if you ignore natural pedestrian routes. If people need to go significantly out of their way, they will either not walk, or they will cross "illegally."

Re: Self driving car hype

[Lanthanide]

So Waymo being able to drive 5,600 miles without human intervention is just a lie then?

Re:Self driving car hype

[fluffernutter]

This is about saving lives, and is no place for a tit for tat argument. Why can't both sides just try be as safe as FUCKING possible so if one side makes a mistake it is covered by the other side? Is it really that hard to comprehend? Are people really that insensitive that they can't inconvenience themselves by grazing a brake pedal if something unexpected happens up the road?

Re: Self driving car hype

[fluffernutter]

80% of driving is easy, 20% of it is full of a million edge cases like this one. Furthermore, edge cases like this are easily avoidable if you have a vested interest in choosing where and when to drive. Waymo could easily rack up 5600 miles and never hit an edge case. We only got elucidated on this one because of a bad test driver and it gave us a window into how bad the situation really is.

Re:Self driving car hype

[plopez]

How is not being hit by a car having the world revolve around you?

Re:Self driving car hype

[thegarbz]

Self driving cars are mostly hype. They're primarily self driving on very good, very clean, very well mapped roads only. Take them out of perfect conditions, and they fail miserably.

I have a solution to this. Let's develop some technology and put it through an iterative improvement process combined with years of trials in the real world in increasingly complex scenarios. It's a shame no one has thought of doing this.

By the way, hype has two possible meanings:
a) It's publicised a lot to which I say So? That's kind of the point of all new technology that many groups are working on.
b) It's benefits are exaggerated: No they're not.

So yes, Self driving cars are currently hyped up. That is not a bad thing.

Re: Self driving car hype

[plopez]

I'm in that older demographic and I never got any. You won or lost and 2nd place was the first loser.

Re:Self driving car hype

[unknown_user_name]

I have a cheap $50 dashcam that provides better night recording than the Uber video. Given Uber’s insane market valuation, a reasonable person would think they would spend the minimal amount of money needed to have multiple high resolution cameras recording every test drive. Doesn’t a test AV exist solely as a platform for evaluating vehicle performance? How do you do that without referencing the basic visual data humans use in everyday driving?

Re:Self driving car hype

[Phillip2]

Jaywalking has an interesting history actually. During the early days of the car manufacture, there were quite a few fatalities as a result of cars hitting pedestrians, to the point that the car industry got a lot of bad press.

So they got some heavy duty PR, invented a word "jaywalking", pushed this into the press, and lobbied intensely that it was all really the fault of the pedestrians. They managed to get it passed into law with one or two years effort. Very effective. Blame the victim.

The Uber video was a classic example.

Re:Self driving car hype

[Xenographic]

Not likely. A.R.S. says the pedestrian has a duty to yield outside of crossings. It's damned hard to see pedestrians crossing in black at night. And yes, I have seen that before.

Re:Self driving car hype

[thegarbz]

Jaywalking doesn't even exist in the rest of the world

It does in many places actually, but it is usually applied with some manner of common sense. E.g. Australia, It's only Jaywalking if you're within 10m of a traffic light that is red for pedestrians.

Re: Self driving car hype

[HiThere]

Depends on the event. IIRC in track there were often 1st, 2nd, and 3rd place trophies. Probably copied from the Olympics. Or perhaps from horse racing. (I'm not really sure what "Win, place, and show" means.)

Re: Self driving car hype

[Bing Tsher E]

Pedestrians can easily step over broken chunks of cement. Even if there are bits of metal protruding out of it. Cars going 20-25 mph can often avoid them, too. Clearly there is a need for a lot more random 'calming chunks' in the road bed in urban areas where pedestrians are common.

Re:Self driving car hype

[amiga3D]

No, they would not have. These types of accidents are routinely written off as just that. People tend to look straight ahead and see what they expect to see. This is why I taught my children to look both directions before crossing the road. I taught them that drivers most likely wouldn't see them because they usually don't. It's unfortunate that autonomous cars are just as bad as humans.

Re: Self driving car hype

[fluffernutter]

I'm not talking about accidents though, I'm talking about potentially accident-causing situations. It's a different set of events, because many possible accidents are avoided by humans.

Re:Self driving car hype

[Calydor]

Lidar works a little differently from your eyes, though.

WHEN it works.

Re: Self driving car hype

[aaarrrgggh]

What about this makes it an edge case? I appreciate most of driving (on a per-mile base) is keeping lane, following the speed limit, and not running into any other cars; and that this part is relatively trivial to manage. Detecting traffic signs/signals is the next increment, and defensive driving follows.

But, saying that a pedestrian at an unmarked crossing is an "edge case" I think belies what driving is.

I will accept a beach ball blowing across a highway as an edge case though.

Re: Self driving car hype

[fluffernutter]

I call it an edge case because a person dressed in dark standing in a darkly lit section of road with an outcropping behind and choosing to walk into the path of the car is not a common event. Yet it is an event that must be considered, as well as the other million events that contain some novel circumstance.

Re:Self driving car hype

[KingMotley]

Well, uber is fixing that my lowering the once-in-every-1500-mile occurrences to once every 11 miles. That'll keep the drivers engaged. LOL.

Re:Self driving car hype

[g01d4]

putting a human safety driver in the seat for 200 hours doing nothing except take over on a seconds notice is safety theater

I think the engineers or, God forbid, management should do stints as safety drivers. Not just to eat their own dog food but to give them more performance insight than the basic data recordings likely provide. One would think they'd also be more motivated to pay attention (well the engineers at least). Any engineer not paying attention as a safety driver isn't likely paying the necessary attention to their code/design.

Re: Self driving car hype

[Anonymous Coward]

What a load of Bull.

Good behaviour is a precondition of luck.

Re: Self driving car hype

[fluffernutter]

The first thing regulation should enforce is a set of sensors that are capable of handing these types of events then. Uber obviously isn't, who knows if Waymo is?

Re: Self driving car hype

[Anonymous Coward]

Darkness doesn't qualify as an edge case, lidar and radar don't need light - and (IR-)cameras should be at least as good as the human eye (contrary to the crappy dashcam footage from the accident).

Re: Self driving car hype

[aaarrrgggh]

Exactly.

Re:Self driving car hype

[edtice1559]

Well maybe. But in the case of the Uber car, it's once every 13 miles, so two orders of magnitude different. With the Waymo car and a problem ever 1500 miles it may still be safer than a human driver.

Re:Self driving car hype

[Namarrgon]

While that's true, it doesn't change the duty that drivers also have to not hit pedestrians. In a case where it should be relatively easy to avoid that, they can certainly be judged at fault as well.

Re:Self driving car hype

[swillden]

So they got some heavy duty PR, invented a word "jaywalking", pushed this into the press, and lobbied intensely that it was all really the fault of the pedestrians.

Not really, at least with respect to the origin of the term. And it seems it was more automobile clubs than the auto industry that pushed it. https://en.wikipedia.org/wiki/...

Re: Self driving car hype

[Type44Q]

LOL! The self-driving industry has no need for enemies with friends like you!

Re: Self driving car hype

[Brockmire]

The homeless in my town is really, really bad for that. Motherfuckers refuse to even look or acknowledge the cars they interfere with.

Re: Self driving car hype

[Jeremi]

I call it an edge case because a person dressed in dark standing in a darkly lit section of road with an outcropping behind and choosing to walk into the path of the car is not a common event

Note that while the above is a difficult case for a human, there is really no excuse for LIDAR/Radar/Sonar sensors not to have detected this particular human. My suspicion is that we'll find out that the either some or all of those sensors weren't working properly on that car, or the car's software was inadequately tested and some sort of bug prevented it from reacting appropriately to its sensors' input. Really, "don't hit pedestrians, no matter what" should be the First Law of Automotive Robotics, and this car failed it badly.

Re: Self driving car hype

[Brockmire]

It exists in Canada. There's deaths in Vancouver all the time. Aggregious jaywalking that can cause accidents needs to be fined like a motherfucker. I don't give a shit if people jaywalk respecting traffic and the vehicles. But we have dark and rainy conditions for a good chunk of the year and don't need fuckers walking out of nowhere.

Re: Self driving car hype

[sonamchauhan]

Or virtually do as the Greeks did. Have the architect stand underneath his monument as the supports are taken away. Have them cross poorly lit streets... In a simulation.

Afterthought: why are these bloody cars even on the streets? Simulate the conditions in software first... poorly lit streets, pedestrians, kids, police redirecting traffic...

Re:Self driving car hype

[phantomfive]

Hype means the capabilities are exaggerated.

Re:Self driving car hype

[kubajz]

Here (one small European country) you can cross the street, but only if the nearest "zebra crossing" is farther than 100 meters. Outside the crossing, though, (understandably) pedestrians do not have right of way.

Re: Self driving car hype

[fluffernutter]

Darkness can be difficult for humans though.

Re:Self driving car hype

[vandamme]

She would have been texting, so, no.

Re: Self driving car hype

[corydoras]

Right. I'm under the impression that the only reason self driving cars are getting good numbers is that they're cherry picking their routes.

Re: Self driving car hype

[JonnyCalcutta]

Good for you. I live in Scotland where I'm pretty sure we have equally dark and rainy conditions for a good chunk of the year and its still weird to me that you can be punished for 'crossing the road'.

Re:Self driving car hype

[thegarbz]

No, hype means exaggerated or over publicised, nothing more. It sure as heck doesn't change the subject in a sentence.

If you wanted to talk about capabilities then you should have mentioned that "Self driving cars *capabilities* are mostly hype."

Re: Self driving car hype

[phantomfive]

The name "self driving car" itself is hype, because no car can drive itself. The hype that it will be here by this year, or by 2020 or 2025is hype. Calling "lane assist" self driving is *definitely* hype.

Re:Pressured to proceed despite poor test results.

[Anonymous Coward]

What I'm shocked about is that, of all companies, Uber would make morally dubious decisions in its race to profit off of a new market. I mean, when have they ever acted like that before?

Nice company

[tsa]

Really nice company, Uber. Their 19th century attitude towards its employees surely will make them do their very best.

Re: Nice company

[bill_mcgonigle]

Their mutual rating system actually does do just that. Get used to the gig economy - your 19th century factory model is going away.

Re: Nice company

[Anonymous Coward]

What? The gig economy IS the 19th century factory model. Back then workers assembled at places and the bosses pointed out which got to work that day. No job security, bad wages, no insurance.

The gig economy is just a scam trying to fool people into believing that they are "freelancers" or "consultants" when they in fact are making slave wages working more hours than is legally allowed. A freelancer or consultant can set their own wage and negotiate on it. In the gig economy you have to accept whatever wage is offered, no negotiating allowed. And if someone made an app to act as a intermediary they get to rake in 30% or so off the top because you are their slave now.

I will never get used to the gig economy and hope that any civilized economy outlaws it. We already allows freelancers/independent contractors in todays economy, but that wouldn't make Uber and its' ilk any money, so off course they don't want that.

Re: Nice company

[thesupraman]

Shhhh, don't disturb the millennial, they are busy disrupting their way to success.

At least until daddy stops paying the bills.

Corporate death penalty...

[aaarrrgggh]

Ok, so forcing them to liquidate might be extreme, but clearly there is some kind of regulatory framework missing here!

I hope the victim has some relatives that want to get rich though.

Re:Corporate death penalty...

[phantomfive]

The regulatory framework is there.......Arizona made regulations allowing them to do this. Clearly they didn't think deeply about the problem, they just trusted tech companies.

Re:Corporate death penalty...

[novakyu]

Two words: tort liability.

Re:Pressured to proceed despite poor test results.

[Anonymous Coward]

What I'm shocked about is that, of all companies, Uber would make morally dubious decisions in its race to profit off of a new market. I mean, when have they ever acted like that before?

Yes, I am shocked. SHOCKED!

You want good? Or cheap?

[BeerCat]

Clearly, not all autonomous vehicles are the same.

It's very like camera - a cheap one and an expensive one will both offer "autofocus" and "zoom lens"

The cheap one will have 3 or 4 focus settings, while the expensive one will be continuous. The cheap one will have 2 or 3 zoom settings, while the expensive one will, again, be continuous.

So, Uber's cars look to be at the "what is the minimum that can make a car steer itself" end of the scale, and the Google ones are "have we missed anything off the long list of things that will help a car steer itself" end

Re:You want good? Or cheap?

[fluffernutter]

The problem is:
1) It is damn expensive to do it right, and
2) There is no standard method to prove if someone is doing it right before allowing them on public roads

Re:You want good? Or cheap?

[mikael]

What they are doing at the moment is like teaching someone to drive a car, where the instructor takes over when something goes wrong. What they also need to be doing is being like a flight school where they train the pilots to handle things when they go wrong. Otherwise, there are always going to be high profile accidents in the news, like the Viola Group accident:

https://www.telegraph.co.uk/ne...

I've seen these kinds of bridges in Norway. One end of the road at the bridge becomes a solid concrete wall. The other end becomes an empty void with no barriers.

In may places they paint murals on the outside walls of buildings. Those extended from the ground to the top of the building and can be anything from crowds of people to fake roads, staircases and windows.
https://farm8.staticflickr.com...
http://www.theparisblog.com/wp...
http://assets.design.cultuurpl...

The UK allows buses to have advertising on them. Those can be anything from pictures of other vehicles to a collage of road signs. Sometimes even an aircraft. For a simple vision systems, those are going to be confusing.

Re:You want good? Or cheap?

[Applehu Akbar]

1) This is Silicon Valley. They have lots of money to pursue difficult technology like this.
2) There is a standard measure already, average distance between interventions by the human monitor. Waymo is clearly ahead here at 5100 miles to Uber's 13. When that average distance gets to some agreed value of very high, self-drive will be ready for general use.

Re:You want good? Or cheap?

[fluffernutter]

1) But they have to make something marketable. Then why didn't Uber have a LIDAR array?
2) That's a poor measure. We need to know exactly how many and what type of obstacles the cars are dealing with in those miles. The biggest problem is, self-driving is easy 80% of the time. It only gets hard for the 20% of edge cases like this one. It doesn't really matter how many successful miles there are if we don't know how well they deal with difficult edge cases.

Re:You want good? Or cheap?

[fluffernutter]

Another thing I should have added for 2) was the fact that Uber *was* making the requirement for interventions and someone died. So obviously the measurement is bull.

Re:You want good? Or cheap?

[thegarbz]

So, Uber's cars look to be at the "what is the minimum that can make a car steer itself" end of the scale

You say this based on an accident, because it sure as hell has nothing to do with reality. Uber's cars are currently among the most expensive on the road (high end Volvo SUV, fully kitted out with about every instrument you can think of). Just because it doesn't work doesn't make it cheap. By comparison most of Waymo's fleet is significantly cheaper.

Re:You want good? Or cheap?

[Applehu Akbar]

I have already opined that I think Uber's failure was in the sensor array. Because this is exactly where so many human drivers fail, self-drive has to field sensors that are a lot better than human eyes, including being able to see in multiple directions at once.

I don't see a real problem with the distance-between-interventions measure. Whenever a human intervention occurs, it represents a situation the self-drive system didn't see. Logging these situations and meticulously cross-checking each one with what the sensor array saw and what the software was doing is how we debug the system. That in the Tempe pedestrian case the Uber monitor failed to intervene successfully is beside the point. The accident obviously represents a moment where intervention should have taken place.

Re:You want good? Or cheap?

[fluffernutter]

So we're going to keep doing this for a million situations? That's a million fatalities. How many edge cases will there be in the real world? Equipment failing is a valid edge case, which brings up my next point. What distance-between interventions misses is the ability of the sensors to operate in all conditions. The camera feeds I saw from the car were terribly dark. Lidar is easily fooled by textured backgrounds. Uber could have avoided this becoming public simply by driving in the day. This tells me not only that it was a weakness that didn't get caught, it was a weakness they didn't even know about.

Re:You want good? Or cheap?

[BeerCat]

So we're going to keep doing this for a million situations? That's a million fatalities

And yet, fatalities happen on meatware driven vehicles, and we haven't stopped allowing people to drive. Since every system that involves effectively controlling a 2 ton exoskeleton has to make general assumptions ("is this a good idea" or "does the convenience outweigh the pitfalls"), then there will come a "good enough risk/reward balance" that will say either "good to go" or "not ready for primetime". That level might be set at 10 situations. Or 100. Or "never". But setting it at 1 million will never work, given that that would eliminate every single human driver from now until the end of time

Lidar is easily fooled by textured backgrounds. Uber could have avoided this becoming public simply by driving in the day. This tells me not only that it was a weakness that didn't get caught, it was a weakness they didn't even know about.

Now we're into hubris (sadly). Gaming a system means avoiding the times when your playbook doesn't work (to make the times when it does work seem to be down to your playbook). Hubris is when you think you have all options covered, even when you don't

Or, as I once saw about aviation: A superior pilot uses his superior judgement to avoid situations that would require the use of his superior skill

In this case, Uber management didn't use good judgement

Re:You want good? Or cheap?

[djinn6]

The biggest problem is, self-driving is easy 80% of the time. It only gets hard for the 20% of edge cases like this one.

Where the heck do you get numbers like 80% vs 20%? Are you counting avoiding a nearly stationary obstacle in the lane an "edge case"?

It doesn't really matter how many successful miles there are if we don't know how well they deal with difficult edge cases.

So tell me, how well do humans deal with difficult edge cases? Can you quantify it? If so, why would that not be a sufficient measure for self-driving cars? If not, how do you know humans are better at those edge cases? And no, gut feelings don't count.

Re:You want good? Or cheap?

[fluffernutter]

Human drivers are very safe compared to Uber, that we know. About other self-driving technologies we know nothing, but I really don't think they are that far ahead of Uber.

Re:You want good? Or cheap?

[fluffernutter]

We can know how well humans deal with edge cases by tracking self driving cars and after modelling driving that resembles human driving (meaning mostly diveway to work to grocery store to driveway type stuff). Come up with a criteria for 'difficult event' and log how many occur during a day of normal driving. Compare that statistic to real human accident statistics, extract the gap and there you have the number of accidents prevented by humans.

Re:You want good? Or cheap?

[bingoUV]

So tell me, how well do humans deal with difficult edge cases? Can you quantify it?

Measure pedestrians killed by cars where a professional was not overseeing the "driver". Self-driving cars : nearly zero miles driven in this scenario (on public roads). Humans kill about one pedestrian in every 600 million miles driven.

This metric works perfectly for self-driving cars. Except that we have very little data.

Similarly you can measure for other problems - like driving off a cliff, or causing squished car pile-up on a motorways with fast driving cars etc.

What is a Jay?

[nsaspook]

Contrary to popular belief, the term jaywalking does not derive from the shape of the letter âoeJâ (referencing the path a jaywalker might travel when crossing a road). Rather, it comes from the fact that âoeJayâ used to be a generic term for someone who was an idiot, dull, rube, unsophisticated, poor, or simpleton. More precisely, it was once a common term for âoecountry bumpkinsâ or âoehicksâ, usually seen incorrectly as inherently stupid by âoecityâ folk.

http://www.todayifoundout.com/...

Yes, Uber has not solved the "moron' human Jaywalking problem.

Re:What is a Jay?

[vandamme]

I nearly nailed one this morning, crossing against my green light. He had on camo clothing.
Then there's the idiots who who walk in the road with dark clothes at night. Where is Darwin when you need him?

You too hard on Uber

[Anonymous Coward]

Let's look at the data. You people all luv data. One human intervention every 13 miles. Well if a human were driving there would be a human intervention every .5 seconds. Win for uber. Second uber have been driving continuously without coffee cigarette or bathroom breaks. How would you holier than though human drivers perform if you were not allowed bathroom breaks. Win for uber. Last no-one cars about uber lives. That human that tragically lost her life getting in Ubers way was not broadcasting on any kind of transponder device. Her location was not represented on any server. How can uber be expected to know where every human is when they Inisist on sneaking around serruptitiously trying to scare Ubers.

Re:Uber hatred turned political a long time ago

[Anonymous Coward]

There are plenty of reasons to dislike Uber; serving on Donald Trump's advisory board is insignificant compared to those.

Everything else you write is just in your head and your echo chamber.

I am satisfied

[DaMattster]

Call this flamebait if you want, but it is my opinion that autonomous vehicles are a bad idea. I am cheering their failure. But, I feel deeply sorry for the person that was killed by the self-driving vehicle in Tempe. That is an extremely unfortunate side effect of this experiment. Autonomous vehicles are an all around bad idea from putting people out of work to safety. There are some things that should remain in control of a human being and one of those is driving.

Re:I am satisfied

[Applehu Akbar]

Most of the driving we do is routine, boring and done from sheer necessity. This is exactly the kind of task at which humans tend to slack off and do poorly. Because of this, decades of car safety improvements have been unable to reduce the annual death rate below about thirty thousand in the US. This is impressive in terms of the increasing number of vehicle-miles, but still an unforgivably large number.

Re:I am satisfied

[Anonymous Coward]

I am on the opposite end of the spectrum. I believe that, one day, autonomous cars will be much safer than human operated ones. Also, they could use infrastructure much more efficiently (driving closer to each other in "trains"), not needing parking (since they can drive off and pick up someone else) lessening the need for parking spaces in crowded cities, not needing a drivers license to be able to go by car (for example allowing children or sick/elderly to get a car to take them somewhere). I see a lot of advantages. I think this is one think that humans definitely should let control of, or even not be allowed to do (except for perhaps special "playgrounds", kind of like a shooting range).

Where I live the authorities have a zero vision when it comes to traffic fatalities. Building better roads etc has helped getting towards that goal, but it is an utopia. Humans makes mistakes, so we will never get there with human limitations.

Autonomous vehicles have the potential of making less mistakes than human by significantly faster reaction times, not prone to emotional stress or affected by external factors like someone talking to you, having potentially better senses (like not relying only on vision, but a combination of systems like radar, lidar, termal imaging). They could also be programmed not to break laws and take weather/road conditions into consideration better. I know of a lot of drivers that insist on keeping the speed limit of a road, even if there has been a sudden cold spell and the previous day left the roads wet. Not saying that autonomous vehicles will be perfect, nothing is. But I still think they can make a better job of driving than humans, eventually.

So, I am sad when they fail, but we shouldn't give up. Eventually they will get there. If you would have asked me 10 years ago if I believed in self driving cars, I would have said that you are crazy. Now I am a firm believer that we will see them sooner than most of us think. But that doesn't mean that the companies can cut corners to be the first one to put it on the market.

Re:I am satisfied

[fluffernutter]

That sounds great. Why can't corporations get there without killing people on public roads. It's not like there aren't huge testing facilities all over the place. It's not like odd test cases are hard to think up.

Re:I am satisfied

[djinn6]

Why can't corporations get there without killing people on public roads.

Because testing under controlled conditions will never cover all the edge cases.

It's not like there aren't huge testing facilities all over the place.

Name 10.

It's not like odd test cases are hard to think up.

Weren't you the one who said there's a million edge cases? Can you make up your mind please?

Re: I am satisfied

[Brockmire]

None of your suggestions would change the daily shitty drivers I encounter everyday. I look forward to a world where the signal lights are actually used fucking properly and vehicles enter and exit simple roundabouts without doing two things wrong. There's a left turn off a highway I need to take. The response time from the first driver to move on a green left light is 3 seconds at best up to 8 seconds (5 is about average) by the time someone honks. That's a difference of 5 to 13 cars that can take the turn on one green left. I look forward to the day when the first car drives within 1 second and all the cars behind it efficiently get through, saving minutes of time and traffic frustration.

Re: I am satisfied

[Brockmire]

I was a QA manager in another industry. I could find bugs in software within hours of testing, often getting a "never thought a user could/would do that". Then with real world testing, you run into other issues where *I* didn't possibly conceive or weren't repeatable under ideal lab conditions. Only real world testing can find not-thought-of use cases. You get a lot more noise in real world vs lab.

Re:I am satisfied

[fluffernutter]

It's very simple. You think up all the edge cases you can and test. When you're out of ideas, then you put a much safer car on public roads.

Re:I am satisfied

[bingoUV]

It's not like there aren't huge testing facilities all over the place.

Name 10.

If there are such facilities, this "Uber" belonged there, not on public roads. If there aren't, Uber is all the more evil for not even testing in controlled conditions before unleashing their dangerous vehicles on public roads.

Why is Uber's ability to test "all the edge cases" worth the lives of human beings ?

Re: I am satisfied

[bingoUV]

Awesome! How many human beings did you kill to get this "real world" testing ?

13 miles per intervention

[Anonymous Coward]

13 miles per intervention is not a self-driving car. That's alpha-level quality and should not be allowed on public roads.

Re: 13 miles per intervention

[Type44Q]

Not even a 12.9mi road?

Not just a jaywalker....

[Anonymous Coward]

Agreed. Also, how many of us have had kids run out in front of us? I have. And if I weren't paying attention, the kid, his parents and I would have a very bad year. Who cares who fault it is: I was part of hurting someone.

And so what if someone was an "airhead" and stepped out in front of the car? Doesn't make it OK. The purpose of self-driving cars is to make the roads safer because the machines are supposed to be better than humans.

Re: Uber hatred turned political a long time ago

[Anonymous Coward]

That's a baldface fucking lie, and you know it. Uber has come under massive criticism from day 1 for shamelessly and egregiously breaking livery and employment laws in nearly every jurisdiction in which they have established themselves.

Now their half-baked AI implementation has killed someone, and you want to beg off criticism of blatant criminality as merely political grandstanding? These mobsters deserve every ounce of criticism they get regardless of who is in office.

Re:Pressured to proceed despite poor test results.

[K. S. Kyosuke]

As Germans would say, Uber Uber uber alles!

Re:Uber hatred turned political a long time ago

[Junta]

When Uber started as a 'ride sharing app', ostensibly helping people coordinate carpooling where they were going to be going anyway, it was fine.

When it became "a taxi, but paying drivers less and trying to get out of the same regulations for no other reason than somehow being 'cooler' than taxi companies", a lot of deserved criticism came about.

Re:That's exactly what the parent said.

[AC-x]

If that machine is incapable of detecting and avoiding unexpected obstacles and that unexpected obstacles is less squishy than a human it could easily be the occupant of said machine that dies...

Re:Uber hatred turned political a long time ago

[K. S. Kyosuke]

Yes, because nobody had criticized Uber before. Projecting much?

Re:Pressured to proceed despite poor test results.

[complete loony]

Sounds like just about every failed IT project. Rush to market, ignore test failures, probably a thermocline of truth.

From what we've been hearing, somebody in the chain of command between the inattentive driver and the CEO, deliberately created this situation and should be charged with manslaughter.

Re:What?

[Junta]

Privatized space companies seem to have been working out ok, at least better than I thought it would.

This may end up being a bust for now at least, but this specific article suggests that Uber isn't good at this compared to others, and this deficiency may have cost a human life and thrown a huge roadblock for the entire industry. Other companies may have well been doing well enough, though it of course may be the case things are far worse than they imagine. Certainly I feel the reporting is more enthusiastic than the state of the technology (reporting seems to suggest go anywhere is right around the corner, as yet the trials are running in very favorable conditions only).

Re:Uber hatred turned political a long time ago

[elrous0]

When it became "a taxi, but paying drivers less and trying to get out of the same regulations for no other reason than somehow being 'cooler' than taxi companies", a lot of deserved criticism came about.

If that's the case, then why isn't everyone piling on Lyft as well?

Re: Pressured to proceed despite poor test results

[Anonymous Coward]

Well then, stop using them. The only way for the market to punish this behavior is to stop using the service. But as we see time and time again, convenience, compliance, and human stupidity will allow them to succeed. We really are out worst enemies.

Re:Uber hatred turned political a long time ago

[JonnyCalcutta]

Nice story, very imaginative, Did you make it up yourself?

Re:Uber hatred turned political a long time ago

[stabiesoft]

Did you not even bother to read TFS? In the summary it quite clearly states the other self drive companies were achieving 5600 miles between interventions while uber could not meet it's 13 mile goal. Sounds to me like uber's system is just plain not ready. I'd even question 5600 miles. Once they get to 1 million I'd say they are there.

I think you are the one making this political.

Re:Uber hatred turned political a long time ago

[Anonymous Coward]

The hilarious thing is your comment received 15 replies yet is downvoted to troll.

Welcome to Slashdot.

Re:What?

[phantomfive]

I seem to remember a few years back that some major auto-maker said they would have self-driving cars on the streets by 2018. Now a bunch of them are making predictions for 2020, and Ford plans on removing the steering wheel by 2024.

This++

[Anonymous Coward]

The video the police released, did it come directly from the car or did it come from Uber engineers extracting it from the car?

Because it's seriously shit, like its been put through a bokeh or vignette filter to darken the outside.

Uber will probably walk away from this blaming the driver, but you can't have the driver as a safety for the car, because the driver does not know what decisions the car has made till the effects are known.

Re:This++

[Anonymous Coward]

It probably did come from the car directly. But, that's not to say that there's any guarantee that the undoctored footage is an accurate representation of what it looked like to the humans on scene. If they didn't adjust the camera for the night, it would make everything look artificially darker than it really was. Cameras are also easily fooled by small amounts of bright light because they're usually set up to preserve the highlights as best as possible, even if that means losing some detail to shadows.

Unfortunately, in this case, that would render the video extremely generous for Uber and not really good enough.

The linked video is probably more accurate given the number of street lights, but the only way to really know what it looks like would be to actually go there for yourself.

Re:This++

[burtosis]

No camera installed on a driverless car should be a shitty throwback to 1990s spy cams. Its 2018 now and the performance you are talking about is barely acceptable in 40 dollar phones.

Re:Pressured to proceed despite poor test results.

[sunking2]

This isn't an IT project though, it's the real world. Failed IT projects seldom costs lives, just money. This is what the app IT world needs to understand when moving into the real world of engineering. Rushing to market has real consequences.

Re: Pressured to proceed despite poor test results

[hey!]

Yes, indeed. Stop using them. However your alternatives may in the long term become limited, because Uber is one of those companies built around an ideology of market disruption.

Why set the bar so high

[monkeyxpress]

Self driving cars are mostly hype. They're primarily self driving on very good, very clean, very well mapped roads only. Take them out of perfect conditions, and they fail miserably.

But even a car that could drive under such conditions would be extremely useful. Take, for example, public buses. They just drive around the exact same route everyday, and the route in many places is upgraded with special lanes and signalling infrastructure to make their job easier. There is no reason why you couldn't start with replacing such bus routes in cities with moderate weather conditions. Over time a combination of roading infrastructure improvements (special lanes, intersection redesigns, beacons etc) and the tech getting better could easily expand out to cover the majority of vehicle uses in a city. Again, we do this for bicycles and buses, so why is it impossible to imagine it would make sense to do some road works to cater towards driver less cars?

Another example is motorway driving. Motorways are already an extremely controlled and regular environment. It would be great to have a driver less truck that can go door to door, but there is no reason why we can't start with depots built off the side of motorways where local human drivers pick up and drop off longhauled trailers. As for weather conditions guides in the road way and other navigation infrastructure could be added if these problems cannot be dealt with using lidar and cameras.

Yes, I agree that a car you can just dump into an unknown urban environment is a long long way off. But I don't understand the fascination with meeting this goal before driver less vehicles can be useful to us.

Comment removed

[account_deleted]

Comment removed based on user account deletion

Re:Uber hatred turned political a long time ago

[Applehu Akbar]

So Uber lags Waymo in self-drive? Microsoft has been getting away with being the laggard in operating systems for years, yet it still makes billions.

Testing on live monkeys.

[Anonymous Coward]

I appreciate that its optimal to test in a live environment but in most cases where clear danger is involved its done with consent.

So i'm wondering when we gave consent for expert system driven cars to be tested on us directly.

Re:Uber hatred turned political a long time ago

[goose-incarnated]

Did you not even bother to read TFS? In the summary it quite clearly states the other self drive companies were achieving 5600 miles between interventions

No, it didn't state that. It states that the other companies claimed to achieve 5600m/intervention. Prior to this crash Uber also made outrageous claims about the progress and state of their SDC capability.

Waymo claims 5600m between interventions. Maybe it's true, but until they are forced to release some data (say, via a fatality) I see no reason for believing a claim made by a company spokesperson. Hell, I shudder when I hear the claims my employer makes about our products. Same with every previous employer.

Re: Pressured to proceed despite poor test results

[Anonymous Coward]

Or, just for novelty's sake, the government could actually enforce the law against a corporation. They've been operating illegally for years and they haven't been taken to task for it. In the early days, the drivers didn't even have insurance for commercial driving.

Re:What?

[goose-incarnated]

I seem to remember a few years back that some major auto-maker said they would have self-driving cars on the streets by 2018. Now a bunch of them are making predictions for 2020, and Ford plans on removing the steering wheel by 2024.

In 2012 they were all saying "five years time", so we are now in the sixth year of the five year prediction.

You think?

[slashmydots]

The smartest AI we have right now, not counting IBM's watson, is about as intelligent as a brain-damaged cockroach and that actually is a real scientific assessment from real scientists. I think Time posted that article actually. So they're supposed to know a truck from a building in full context. Really? Does anyone expect a computer to be on par with a human brain on this one? That's ridiculous. We're decades away from a working self-driving car.

Re:Uber hatred turned political a long time ago

[Anonymous Coward]

Perhaps because they are much quieter about their business whereas Uber was louder than the loudest fool blaring about how "disruptive" their great service was and that local laws/regulations didn't apply to them?

I've seen that video

[rsilvergun]

thanks to Arstechnica, but none of the main stream press covered it. I hate using the phrase, but it's hard not to notice them all towing Uber's line. But that's kinda the thing with the MSM. They always seem to tow the corporate line. Thanks to that most folks who don't read /. or ars are going to go to bed thinking this crash was unavoidable.

Re: I've seen that video

[Lanthanide]

Yeah, you should hate using that phase, because the actual phrase is "toe the line".

Re:I've seen that video

[HiThere]

To be fair, it seems as if that is also the police line, so that's probably where they got there information. If a large number of groups say the same thing, the first thing to suspect is not that they were conspiring, but that they used the same source.

Re: I've seen that video

[Known Nutter]

Give GP a brake. It's a doggy dog world out there.

Re: I've seen that video

[vtcodger]

Naw, probably the police just assessed the situation as they would for a human driver. Nighttime. Pedestrian in a place a pedestrian shouldn't have been. An unimpaired human driver might well have been unable to avoid an accident and very likely wouldn't have been held responsible. BUT. The damn car seemingly SHOULD have done better than it did. Disturbingly it never seemed to try to avoid the accident even though it might have been too late to do so by the time the pedestrian was recognized. THAT seems very bad.

Conclusion: Probably not really ready for prime time.

Re: I've seen that video

[Brockmire]

I see what you did, there.

Re: I've seen that video

[HumanEmulator]

What are you talking about? Arsâ(TM) article is based on an article in The NY Times that uncovered the relevant documentation that Uberâ(TM)s cars are performing badly.

Re: I've seen that video

[sonamchauhan]

The Uber was doing 3mph over the limit (that's 17% more kinetic energy according to someone). Yet the police chief was quick to excuse it, and also the safety driver who seemed to be ignoring her job. The chief who made the statememt also served in California. It'd be good to vet her links to Uber, just to ensure undue influence didn't take place.

Re: I've seen that video

[mentil]

Better odds that Uber's legal department sent them a nastygram saying to shut their trap until the investigation is complete.

Re: I've seen that video

[Dog-Cow]

I see what you did, their.

FTFY

Re: Pressured to proceed despite poor test results

[LifesABeach]

This may be an ignoreable point, but someone is dead. Maybe some of our attention should be focused on why did the car not Brake, or Veer?

That Adam ruins thing has a bit on it

[rsilvergun]

it was derived from a racial slur to make it easier to hand the roads over to cars.

I don't undertand

[OYAHHH]

Why a company would want to get into the self driving car angle of the rideshare market.

We see the reports of the incredibly low profits the drivers currently make, and UBER wants to buy a bunch cars as well? The one thing that drives down profits?

Re: I don't undertand

[Type44Q]

Fear; fear of their market disappearing. Surely it's obvious...?

Is this the triggering event of the AD winter?

[DrTJ]

Uber lost $4.5 bn last year, in a revenue stream of $7.5bn. They were evaluated to be next-to-last in a large survey when it comes to AD technology and strategy.
They are the ones with the largest stakes as their current business model is less than solid, and really need to bring AD to the streets ASAP in order to survive.

This accident shows how far they are from that - this accident was not a difficult case from any perspective, sensory (the video shows dashcam footage, without any kind of HDR functionality so the perceived "blackness" is misleading), world model, trajectory estimation, situation analysis or vehicle dynamics. This is a really a 101 case.

If the investors get the picture, they could be less fond of funding this cash burner, which will make Uber default lightening fast. Ubers' demise will probably affect the other AD inverstors' willingess to put up cash, slowing down the entire AD industry... and the "AD winter" may persist until the technology matures.

Even Waymo's 5600miles-between-interventions status is orders of magnitude off from any reasonable target. Do the math; US citizens drive 2.5 trillion miles annually, how many driver-is-supposed-to-intervene incidents would that imply per year? (about 450 million). Multiply that with accident probability for a driver with the alertness of someone who just saw that last 5000 miles pass OK). Even if the probability of a fatal accident would as low as 1 in 1000 for such events, that would translate into about half a million road fatalities. It's slightly larger than the current 30 000 deaths.

At any rate, I doubt that Uber can keep up appearances for the investors long enough for AD to save them.

Re: Is this the triggering event of the AD winter?

[Type44Q]

Uber lost $4.5 bn last year

That's misleading as hell. Uber's is profitable with shitloads of cash flow; they chose to spend that money, which they did. Are their investors stupid? Perhaps but that's another debate.

Re: Is this the triggering event of the AD winter?

[DrTJ]

I don't think that is particularly misleading. It just tells us that is ridiculously expensive to develop AD, and that Uber needs to do that for quite some time yet before they can start to make money from it. Will the investors sign up en-masse for a long period? I doubt it.

Re: Is this the triggering event of the AD winter?

[raftpeople]

Yes, they spent that money subsidizing the cost of the rides (especially the drivers). We call that a "cost" to their operations which reduces profit.

They are not profitable. They will never be profitable at current ride rates unless they eliminate their biggest expense which is drivers.

Re: Uber hatred turned political a long time ago

[Type44Q]

The pink mustache scares 'em away (it's pretty crusty; might even be a old tampon string tucked in there)...

Re: Pressured to proceed despite poor test results

[hey!]

Someone is dead because of a faulty development process, which in turn is the result of a toxic business climate.

I suspect this happens more often than we know; it's just seldom that you can connect the dots so readily.

Re:Dubious

[Whorhay]

Given that Waymo has cars that go 5600 miles on average between human interventions, and Uber was only averaging 13, I'd say your being alarmist. If I had a Waymo car I could in theory get away with intervening less than twice a year on average. What all of this has shown is that Uber isn't competent or responsible enough to be working in this field.

Excessive Hate On Uber - do the math

[FeelGood314]

Human drivers will kill about 100 people in the USA today, world wide it will be close to 2000. Driving also consumes the valuable resource of 4.5 million workers in the USA alone.

Uber might be an easy company for some people to hate (and there are even some people with financial incentive to hate them) but other than the slashdot hate for crypto currencies this is excessive. They also might have the worst self driving car on the road but someone has to be the worst*. We need self driving vehicles and there are going to be accidents. If lax standards means killing 10000 people so that self driving cars that have 50% less fatalities are on the road one month sooner then we have a net saving of 50 000 people. So everyone calling Uber murderers, for them to stop testing and for more stringent controls on self driving cars, statistically you are significantly worse than them. While your at it why don't you donate to Green Peace so they can protect big coal and rail delivery of crude oil?Poor math and critical thinking are what will doom the human race.

*What do you call the person who graduates last in medschool? --------- Doctor.

Re:Excessive Hate On Uber - do the math

[raftpeople]

You're assuming that the only way to learn and improve the systems is to allow situations like the one that just happened. But there are ways to reduce risk. For example, keep the auto-braking system from Volvo engaged, if it is triggered and the AV didn't respond, then you now have a learning moment that didn't require a death.

Re:Excessive Hate On Uber - do the math

[fluffernutter]

We need self driving vehicles only IF they ever make a dent in the fatality rate. So far I see no evidence at best and I'm not prepared to give them leniency on killing people now based on a mere promise and I don't think anyone should. All evidence right now is that automation isn't coming anywhere close to human safety, which really is quite safe if you take it in perspective of miles driven by humans every day. Being overweight is probably more dangerous and people volunteer for that.

Re:Excessive Hate On Uber - do the math

[corydoras]

It's not good enough to improve the numbers overall by being better at ideal driving circumstances. If they can't avoid accidents like these (it didn't even detect that there was a pedestrial there from all available evidence), they shouldn't be on the road. The standard needs to be, at a minimum, as good an the best humans in ALL CIRCUMSTANCES.

Re: Pressured to proceed despite poor test results

[fluffernutter]

I'm surprised you didn't get down voted for recommending that.

Re: Uber hatred turned political a long time ago

[Type44Q]

When Uber started as a 'ride sharing app', ostensibly helping people coordinate carpooling where they were going to be going anyway, it was fine.

You're not all that familiar with the early history of Uber, are you.

Re:Dubious

[fluffernutter]

As I said above, 80% of driving is easy, 20% of driving is hard. Not difficult for Waymo to stick to the situations they know they are good at while still being a ticking time bomb on 20% of edge cases. We need full disclosure from these companies, including proof that they can in fact see and recognize ALL objects at ALL times of day and in ALL weather that may be in danger from the car, or be a danger to the car and that they react appropriately.

Re:Pressured to proceed despite poor test results.

[plopez]

But marketing tools us it was all rainbows and unicorns as far as the eye could see.

Re: Pressured to proceed despite poor test results

[Potor]

I have never will never used Uber. I also have never owned a car. Car-sharing, public transportation, and bikes have been enough for me. As for disruption, I am willing to bet that Uber will disappear before public transportation does. Disruption cannot be an absolute model.

Re: Pressured to proceed despite poor test results

[plopez]

It happens all the time which is why I am finally running away screaming fro software.

Stop the victim blaming

[plopez]

Cars are not allowed to do that. Period. People blind, kids fail to pay attention, roads are icy, people have strokes and heart attacks crossing the street etc. SDV *must* accommodate.

SDV will never take humans out of the look as all software is written by humans

Re: Stop the victim blaming

[LinuxLuver]

Sometimes the victim is to blame. It's that simple.

Re: Pressured to proceed despite poor test results

[LifesABeach]

It might be interesting to know who is in charge of this cluster f.

Re:Uber hatred turned political a long time ago

[Kjella]

Waymo claims 5600m between interventions. Maybe it's true, but until they are forced to release some data (say, via a fatality) I see no reason for believing a claim made by a company spokesperson.

It's public because California's regulations require it to be public. 352454 miles driven, 63 disengagements = once every 5600 miles. Read the report (pdf) yourself.

Re: Pressured to proceed despite poor test result

[Type44Q]

In the early days, the drivers didn't even have insurance for commercial driving.

Shooting down shills is too easy: In the early days, Uber contracted with fully-insured and regulated livery/limo companies.

Don't let the door hit your dumb, lying ass on the way out.

Re: Pressured to proceed despite poor test results

[ShanghaiBill]

Stop using them. However your alternatives may in the long term become limited

How so? Lyft provides a near identical service, and tends to behave more ethically. So why use Uber?

It wasn't always this way in the USA

[bjdevil66]

Adam Ruins Everything gives a great/funny history lesson on how jaywalking became a term for an act that wasn't even a crime until the automobile industry saw their new business model threatened around 100 years ago.

In the past, calling someone a "jay" a bad slur, so calling someone a jaywalker back then was more like, "Dumbass-Walker". Now that slur is a legal term. Imagine having to stand before a judge and plea guilty or not guilty to "dumbf**kery".

Re: Uber hatred turned political a long time ago

[reanjr]

No one wants livery laws. They're a protection racket. The fact that Uber is breaking the law is a social benefit.

.... It was the worst of times.

[TheRealHocusLocus]

Imagine the plight of these poor human beings, "test drivers", who have logged countless miles with their hands hovering near the wheel of an autonomous car. Which takes far more effort than driving. Imagine the stress of being placed in a position of complete helplessness by your very job description. And every time you move the wheel or tap a pedal, a time-stamped log entry is created and you know that your own intuitive sense that something was not quite right, will be analyzed by other persons whose cumulative judgement could cost you your job... should they decide that your correction into the lane or speed adjustment has deprived them the opportunity of analyzing what the AI might have done, seconds hence.

Even worse, imagine what happens to people who completely 'let go' to reduce this stress. They fall within the spectrum of useful idiot to patsy -- set up for a fall -- because the AI they're using may pass a driving test under perfect conditions but it's no licensed driver. If they were in the passenger's seat and a human driver had done something there is always the "what were you thinking" interaction, the human driver catching themselves and (most often) apologizing, thanking the passenger. They are alone in a bobsled, not permitted to steer. Their deep pocket corporation may have promised them the Moon, but it is not capable of absolving them legally from a driver's responsibility.

That is only a taste of what is to come.

I consider the push for self driving cars to share the road with humans to be a malignant global cultural aberration of stupidity. There are no ultimate winners. Economically it is a job-killer in plain sight and I am glad to see Uber drivers rebelling against their duplicitous corporation. AI developers and their in-pocket insurance companies will demand that human drivers be taken off the road NOT from AI's superiority, but to cut corners and push their product onto the market before it is as safe (overall) as once promised, even as safe as a human driver. The corner-shaving tech elite -- and their unwitting stooges -- will become a bourgeoisie menace to everyone, forcing 'lower class' people to walk more than ever before in any developed country, for miles. It is a Darwinian regression. Their children will be helplessly dependent on black box technology, the kind you cannot take charge of and master. Orwellian tracking will flourish because the tech-bastards will decide that every object near public roads must have a transponder, to make their job easier. You will see an era when 'jaywalking' becomes anyone who is out and about without a transponder. Whether they are crossing a road or not. This is a DUMB SORRY-ASS FUTURE .

Re: .... It was the worst of times.

[Brockmire]

I should have stopped reading after the first paragraph. Your logic is bizarre. The driver should be a member of the engineering team with deep understanding of the vehicle. You think test jet fighter pilots don't feel the operation and feedback how it operates with the software guys heavily relying on the pilot's experience and skills? Don't cook the fucking tests, collect fucking data so you improve your tech.

Re:Uber hatred turned political a long time ago

[aaarrrgggh]

As a publicly traded company, Waymo lying would be a bigger offense than a private company. Given the number of years they have been doing this, I would say 5,600 miles might be reasonable. The real issue though is how quickly do they discover new limitations and improve that rate. If each of the interventions they experience now is a one-in-a-million, unique event, then getting to 10,000 miles might be a challenge. I'm not even sure how you address that, as it is about once every couple months for a safety driver; how can that be reliable as a countermeasure.

Re:Uber hatred turned political a long time ago

[careysub]

I'd even question 5600 miles. Once they get to 1 million I'd say they are there.

Depends on what an intervention is. There is huge difference between "coming to a safe stop because the vehicle is confused" and "slamming into an obstacle without stopping". The former would be an annoyance, and might be unacceptable in a consumer product, but be quite acceptable for commercial trucking where they can factor in the cost of handling such incidents.

The Google and Uber metrics are useful for evaluating the relative maturity of the systems though.

Re:What?

[djinn6]

If you learned anything from software management, you'd know estimates are notoriously inaccurate. I personally take any estimate I receive and quadruple it before passing it further up.

That said, the 2020 predictions are pulled out of someone's ass. It's only there to keep investors happy, since most of them cannot comprehend timelines longer than 8 quarters (and some even less).

Re:Uber hatred turned political a long time ago

[Uberbah]

There is huge difference between "coming to a safe stop because the vehicle is confused"

Because if it didn't it would hit something or drive at an unsafe speed - thus being a recordable intervention?

Re:Uber hatred turned political a long time ago

[Uberbah]

If that's the case, then why isn't everyone piling on Lyft as well?

Because Lyft is Moe's Tavern next to Uber Mr. Burn's nuclear power plant. Any more questions?

Re: Uber hatred turned political a long time ago

[Uberbah]

No one wants livery laws. They're a protection racket. The fact that Uber is breaking the law is a social benefit.

Paying untrained drivers with little insurance a non-living wage is the social benefit? Or is it the assraping monopoly prices they'll start charging if they manage to drive taxi companies out of business - taxi companies that have to make a profit and can't afford to lose billions of dollars a year. Or maybe its if Uber manages to get an AI that doesn't kill everyone at which point they drop all their human riders while continuing to assrape you on prices.

Re:Uber hatred turned political a long time ago

[complete loony]

It's also likely that those 63 interventions were because the car was *too* cautious. For example, around construction zones.

I remember hearing one anecdote. Workmen were moving around their vehicle, inside the border of traffic cones. The car was predicting that they might step out in front, so it just didn't move.

Re: Pressured to proceed despite poor test results

[Jeremi]

It happens all the time which is why I am finally running away screaming from software.

For goodness sake, don't scream, that will only make it easier for it to hunt you down.

Re: Uber hatred turned political a long time ago

[Brockmire]

I support the change from car analogies to Simpsons analogies.

Re: Pressured to proceed despite poor test resu

[Type44Q]

And you're still wrong.

Re: Pressured to proceed despite poor test result

[LinuxLuver]

While the car could have handled the situation better, there was a human driver present who also failed to deal with a pedestrian crossing the multi-lane road in the dark. The pedestrian was the primary error element and the vehicle - and driver - weren't up to compensating for the pedestrian's poor judgement. Why would you walk in front of a lit-up morning vehicle at night? Why?

Re:Pressured to proceed despite poor test results.

[lucien86]

Real AI is hard, very hard, and when it gets things wrong it can kill people. The other problem with it is that simply throwing money or huge teams at it isn't enough, there is a serious shortage of competent experts - and that's even for ordinary weak AI. For strong AI (what cars ideally need) the number of even half competent experts is next to zero. We know this because there is not a single working true strong AI machine in the world. Yet. :)

Re:Dubious

[david_thornley]

If 20% of situations are edge cases, then, considering the number of "situations" a Waymo car has, it does very well indeed. Personally, I don't know under what conditions Waymo cars are tested; where do you get your information?

There will never be 100% confidence that cars can detect all possible obstacles. Lots better than humans is the best we're going to be able to do.