Stories
Slash Boxes
Comments

SoylentNews is people

posted by Fnord666 on Monday March 19 2018, @10:31PM   Printer-friendly
from the bound-to-happen dept.
 
This discussion has been archived. No new comments can be posted.
Display Options Threshold/Breakthrough Mark All as Read Mark All as Unread
The Fine Print: The following comments are owned by whoever posted them. We are not responsible for them in any way.
  • (Score: 5, Insightful) by Knowledge Troll on Monday March 19 2018, @10:37PM (25 children)

    by Knowledge Troll (5948) on Monday March 19 2018, @10:37PM (#655162) Homepage Journal

    I'm positive that the company that is known for pushing the limits, ignoring regulations, and in general being total shits and douche bags will cooperate fully with giving the investigators all the footage from their black boxes that make them look innocent. Uber can be trusted, yessir.

    Nothing can go wrong with Uber doing this at all.

    Starting Score:    1  point
    Moderation   +4  
       Insightful=2, Informative=1, Underrated=1, Total=4
    Extra 'Insightful' Modifier   0  
    Karma-Bonus Modifier   +1  

    Total Score:   5  
  • (Score: 2) by Fluffeh on Monday March 19 2018, @10:46PM

    by Fluffeh (954) Subscriber Badge on Monday March 19 2018, @10:46PM (#655167) Journal

    The problem is that with the reputation that Uber has, the more it holds back, the more it looks like Uber is just being Uber here - and that's not going to work well in the court system for them. The downside is that they will just keep throwing more litigation at it until everyone gets weary of reading about it.

    I really think that this is one of those cases where the state/federal government should step in and sue on behalf of the family/estate of the deceased - just to make it an even fight in the courts rather than Joe Three-Partner-Practise up against the might of hundreds of Uber Lawyers.

  • (Score: 5, Interesting) by Eristone on Tuesday March 20 2018, @12:01AM (23 children)

    by Eristone (4775) on Tuesday March 20 2018, @12:01AM (#655201)

    There is a preliminary report that the car was not at fault. https://www.sfgate.com/business/article/Exclusive-Tempe-police-chief-says-early-probe-12765481.php?t=11286f4b07 [sfgate.com]

    • (Score: 5, Interesting) by vux984 on Tuesday March 20 2018, @12:46AM (12 children)

      by vux984 (5045) on Tuesday March 20 2018, @12:46AM (#655212)

      That's quite interesting. It's normally *really* hard to prove you aren't at fault when you hit a pedestrian. The claim that they came out of nowhere is usually not terribly convincing. Even in the event the pedestrian comes out from behind a parked truck or something, the police and insurance will usually counter that one was driving too fast for the conditions if there is insufficient time to react to someone stepping into traffic.

      The fact that the vehicle was technically speeding, even if only by a few mph; is also often cited as an aggravating factor to the accident, as well as increasing the severity of the injury; and impacts your liability -- not only was the driver going too fast for the conditions, but was going too fast period.

      • (Score: 2) by MostCynical on Tuesday March 20 2018, @02:09AM

        by MostCynical (2589) on Tuesday March 20 2018, @02:09AM (#655225) Journal

        "too fast for the conditions" includes "too fast to avoid a pedestrian in full suicide mode"
        Driver has a one-to-two tonne vehicle; pedestrian has skin.

        --
        "I guess once you start doubting, there's no end to it." -Batou, Ghost in the Shell: Stand Alone Complex
      • (Score: 3, Informative) by pTamok on Tuesday March 20 2018, @08:02AM (4 children)

        by pTamok (3042) on Tuesday March 20 2018, @08:02AM (#655297)

        According to the reports I've seen, it wasn't 'technically' speeding - it was actually speeding - that is, its speed was greater than the applicable speed limit. Many people take the view that the posted speed limit is advisory, and being a few miles per hour (or kilometres per hour) over the limit is O.K., so long as you don't exceed some imaginary leeway. Such leeway is assumed by drivers in the UK as 10% + 2 mph*, so 35 in a 30 zone is OK, but legally, anything over the posted number is breaking the law, even if it is difficult to prove (radar sensor accuracy and stability since the last calibration is arguable in court).
        I think it is reasonable to require an autonomous vehicle to observe the limits - if you can only measure velocity to a certain accuracy, make sure your error margins don't exceed the posted speed. Car speedometers already do this: they deliberately read over the actual speed, as manufacturers are held to high standards on this point - speedometers are allowed to show higher than the actual speed, but not lower.

        Doing 38 in a 35 zone doesn't seem like much, but it adds 17.9% to the kinetic energy of the vehicle, and reduces the reaction time by 7.9% - that's enough to turn an accident that 'merely' injures a pedestrian (or just misses them) into a fatal accident. Someone else on this discussion has posted the pedestrian fatality rates by speed of collision at 20/30/40 mph. It's one reason (of several) why 20mph limits in residential areas is becoming more popular.

        *Some semi-official bodies have publicised this as a recommendation, basically saying that most (not all) police forces in the UK tend to apply that rule, but will apply the limits exactly if it is deemed justified i.e. you have been driving like a pillock and/or had an accident.

        • (Score: 0) by Anonymous Coward on Tuesday March 20 2018, @11:16AM (1 child)

          by Anonymous Coward on Tuesday March 20 2018, @11:16AM (#655324)

          > reasonable to require an autonomous vehicle to observe the limits

          Disagree -- it should be a judgement call and this could be quite difficult to turn into a useful algorithm--

            + Sometimes (as mentioned elsewhere in this discussion), the speed limits are stupidly high. Not all roads near schools have low limits during school hours. Personally, if I'm on a dark street with cars parked on both sides, I may be going slower than the limit. Same for suburban streets at dawn and dusk where there are deer (this is when deer seem to be most likely to move around and cross the road)...deer move a lot faster than pedestrians and pop out of wooded areas very quickly.

            + Sometimes it makes sense to observe the limits, in particular where the limits have been sensibly posted in towns/cities. Much of the time the limits are set by traffic engineers that have a clue (but not everywhere).

            + Other times it makes much more sense (and I believe is generally accepted to be safer) to move with the flow of traffic. One car obeying the limit makes trouble...on a freeway/motorway where everyone else is speeding.

          I believe that "driving at imprudent speed" is a judgement call cops can make just about any time they like. For example, failure to slow down through a temporary construction zone.

          Based on the limited data available at this time, I think Uber has a lot of work left to do.

          • (Score: 2) by vux984 on Tuesday March 20 2018, @08:49PM

            by vux984 (5045) on Tuesday March 20 2018, @08:49PM (#655611)

            "+ Sometimes (as mentioned elsewhere in this discussion), the speed limits are stupidly high."

            One isn't required to drive "stupidly fast" though; and in fact you are legally obligated to reduce your speed where safety requires it. The speed limit is the *maximum* it is legal to go; it's not a requirement.

            In theory if the majority of vehicles end up autonomous and they are programmed to observe the limits then they will effectively dictate the flow.

            Until then though, your note that it is safer to go with the flow is not wrong. But it's also a catch-22; speeding to go with the flow may be safer, but it is still speeding - its still illegal and if there is an accident, the damage will have been increased by the speeding factor. There is no real winning move there.

            There are so many such cases when driving, where you are damned if you do, and damned if you don't. Where the legal system is at odds with safety. Where the liability conflicts with safety. Where you can avoid an incident that won't be your fault, but in doing so increases the chances of an incident that will be your fault.

        • (Score: 0) by Anonymous Coward on Tuesday March 20 2018, @12:36PM (1 child)

          by Anonymous Coward on Tuesday March 20 2018, @12:36PM (#655340)

          I drive a car with more than 100 HP. Which is nothing by American standards, btw... That enough that if I look away from the speedo for a couple of seconds, the speed will easily climb from just below the speed limit to high enough to get a ticket (2 km/h over the limit).

          Yet, for some strange reason, they expect me to spend more time looking at traffic and pedestrians, than I do looking at the speedo.

          • (Score: 2) by vux984 on Tuesday March 20 2018, @05:53PM

            by vux984 (5045) on Tuesday March 20 2018, @05:53PM (#655521)

            That's just it though. Speedometers read a touch low. So if you are intentionally keeping your speed at 60km/h as shown by the speedo, and then climb to 62km/h by accident you are still ok. The manufacturers deliberately calibrate the speedos to overestimate your speed.

            The standard in the UK for examples is that a speedo must *never* show less than the actual speed, and must never show more than 110% of actual speed + 6.25mph. So at 100mph, its legal for your speedo to show anything from 100mph to 116.25mph. And from my (limited) experience most will probably show around 105-110. So if you are driving, and intentionally keeping it at 100mph as read by the speedo, you are *really* going mid-90s, and if you drift up to 102 now and again that's fine.

            Plus most speed traps themselves give a small bit of grace; to account for their own potential for error, except they are calibrated 'the other way'. So your speedo is always overestimating your speed (to ensure it never reads lower than you really are going), and police radar is underestimating it to ensure they never give you a ticket when you are within the limit, and the upshot is that if your speedo says 2km/h over the limit, you should have nothing to worry about. Unless your speedo is broken.

      • (Score: 2) by Knowledge Troll on Tuesday March 20 2018, @08:08AM (1 child)

        by Knowledge Troll (5948) on Tuesday March 20 2018, @08:08AM (#655298) Homepage Journal

        I'm untrusting of the reporting on the topic of autonomous vehicles because there is a really strong narrative being pushed that the robot cars are safer than human drivers. This narrative is being pushed very hard and with out justification because the technology is simply too new. Objective criticism of the robot car actions is almost impossible - for some reason people don't fault the self driving car for driving like crap.

        Maybe it's because most people can't drive very well to begin with, I don't know, but criticism of these machines is extremely difficult even when they behave poorly.

        • (Score: 2) by Wootery on Thursday March 22 2018, @10:52AM

          by Wootery (2341) on Thursday March 22 2018, @10:52AM (#656548)

          Agree that there doesn't seem to be any serious analysis of this stuff - or if there is, no-one's talking about it.

          How safe are the autonomous cars really? It doesn't do to just look at fatalities per mile - that ignores not only non-fatal incidents, but also the fact that city driving (which Uber presumably disproportionately favours) isn't the equivalent of high-speed driving.

          If a decent study on this question exists, please point me to it!

      • (Score: 2) by Freeman on Tuesday March 20 2018, @05:31PM (3 children)

        by Freeman (732) on Tuesday March 20 2018, @05:31PM (#655511) Journal

        The pedestrian was pushing a vehicle and apparently they weren't paying enough attention when they were crossing the road. Accidents happen, it sucks, but you can't stop a big heavy machine on a dime. My 2 year old knows they shouldn't cross the street without Mommy or Daddy and that you're supposed to look both ways for cars. While going 38 in a 35 is speeding, I wouldn't call it malicious.

        Uber's autonomous car being the first to kill anyone isn't the kind of publicity they need and definitely sends up all kinds of red flags. Since they have a history of doing unethical and possibly illegal things. I just don't think this is one of those things. There definitely should be a thorough investigation, though

        --
        Joshua 1:9 "Be strong and of a good courage; be not afraid, neither be thou dismayed: for the Lord thy God is with thee"
        • (Score: 3, Insightful) by vux984 on Tuesday March 20 2018, @11:14PM (2 children)

          by vux984 (5045) on Tuesday March 20 2018, @11:14PM (#655703)

          Take a look at the photo of the site:

          https://jalopnik.com/video-shows-pedestrian-in-fatal-uber-crash-stepped-in-f-1823922228 [jalopnik.com]

          This *seems* to show a fairly open area. A fairly wide median. The bike had white plastic bags on it that would have caught the headlights. The person was walking the bike so not moving terribly fast. I understand that these kinds of accidents happen, but it's kind of hard to believe that an attentive driver would have had zero warning, zero chance to react. And would have failed to at least hit the brakes before striking the pedestrian. Even if the accident couldn't have been avoided, I'm still surprised the brakes weren't even applied; especially by a computer with better reaction times than me.

          I also find the comments that she wasn't at the crosswalk to be ... misleading. Would the vehicle really have stopped if she'd been at the crosswalk? Is uber running a different algorithm for crosswalks? If she'd been approaching the crosswalk and just walked out into the crosswalk without looking, would the vehicle have stopped in time then? Crosswalks at best have better lighting, and you aren't allowed to park right adjacent to them to help sight lines, but im not seeing a sightline issue here, and id thought autonomous vehicles could cope with lower light levels better than people so the "shadows" should have been less a problem.

          I'm also seeing some followup that the vehicle didn't have "time to stop", which I don't really find suspicious. What i find worrying is that everything i can see suggests the vehicle really should have had time to at least hit the brakes though.

          • (Score: 1) by pTamok on Wednesday March 21 2018, @08:37AM (1 child)

            by pTamok (3042) on Wednesday March 21 2018, @08:37AM (#655991)

            There is a really good Twitter thread about the infrastructure design at the accident site.

            https://twitter.com/EricPaulDennis/status/975891554538852352/photo/1 [twitter.com]

            There are pedestrian footways build on the 'median', which are exactly where a good engineer would put them, but 'closed off' with signs, but no fences.

            Well worth reading.

            • (Score: 3, Insightful) by vux984 on Wednesday March 21 2018, @10:29PM

              by vux984 (5045) on Wednesday March 21 2018, @10:29PM (#656369)

              Wow, yeah... I mean, it's still clear the pedestrian shouldn't be there. But talk about setting things up to fail. And I find it ever less convincing that a human driver would ever found completely faultless for striking a pedestrian walking there, without so much as attempting to brake. It's just wide open. The pedestrian was walking. Its not like they were hiding behind a truck and jumped out in front of traffic to try and commit suicide.

              The pedestrian has responsibility too; they aren't innocent but the driver/vehicle should have had some time to react.

              But I just don't see how uber could get a complete walk on this; even if they weren't charged for the accident (which I'd understand; lots of accidents end without charges -- accidents happen and I'm not at the point of saying uber was criminally malicious or negligent); but you'd think in terms of insurance they'd have at least some liability.

    • (Score: 1, Informative) by Anonymous Coward on Tuesday March 20 2018, @02:35AM (6 children)

      by Anonymous Coward on Tuesday March 20 2018, @02:35AM (#655235)

      "Traveling at 38 mph in a 35 mph zone on Sunday night, the Uber self-driving car made no attempt to brake

      • (Score: 2, Interesting) by Anonymous Coward on Tuesday March 20 2018, @06:28AM

        by Anonymous Coward on Tuesday March 20 2018, @06:28AM (#655281)

        "Traveling at 38 mph in a 35 mph zone on Sunday night, the Uber self-driving car made no attempt to brake

        Which means it's at fault. Period. A stupid human may not be liable, but a computer that is suppose to be reactive to situations like that, should have hit the brakes.

      • (Score: 4, Interesting) by Knowledge Troll on Tuesday March 20 2018, @08:13AM (4 children)

        by Knowledge Troll (5948) on Tuesday March 20 2018, @08:13AM (#655299) Homepage Journal

        If that machine can't hold a constant velocity of 34.9 mph in a 35 zone it isn't fit for the purpose of being on the road. That control loop is one of the very first things they would be doing with the control system.

        • (Score: 2) by pendorbound on Tuesday March 20 2018, @03:26PM (3 children)

          by pendorbound (2688) on Tuesday March 20 2018, @03:26PM (#655415) Homepage

          That's unrealistic given the way gasoline automobiles work. It doesn't take much external force to alter the car's velocity a small amount. The nature of the drive train means over & under corrections are inevitable. A slight rise or dip in the terrain or a gust of wind would be plenty to change the car's velocity by a few MPH. The physical system that the computer is controlling (engine & brakes) isn't accurate enough to avoid inevitable overshoot when the computer applies control. You can see that at work with a simple cruise control on human-driven cars.

          The only way you'd get a control system to keep the car exactly at a given speed would be to have it constantly alternate gas & brake to speed up and slow down. That's ruinously bad for mileage and the mechanical lifespan of the car, and would make for a really bumpy ride, but that's the least of the problems. A car constantly braking & accelerating would appear erratic to other drivers, probably resulting in them speeding around it to "get away from" the erratic driver.

          Until/unless 100% of cars become fully autonomous, self-driving cars need to behave in a fashion that's similar to what human drivers expect of other human drivers. Drifting a few MPH over or under the posted limit is normal, and thus it's actually the safest way for an autonomous car to behave. The reactions of human drivers around it are the most likely source of accidents.

          • (Score: 0) by Anonymous Coward on Tuesday March 20 2018, @08:03PM

            by Anonymous Coward on Tuesday March 20 2018, @08:03PM (#655600)

            That's unrealistic given the way gasoline automobiles work.

            I don't agree at all with this statement.

            It doesn't take much external force to alter the car's velocity a small amount.

            And if the control loop on those cars is under 1khz I'd be surprised - that means with in 1/500th of a second corrections can be applied to any external disturbance. Running at 1khz is trivial for the velocity control loop if they wanted to.

            The nature of the drive train means over & under corrections are inevitable.

            True but that is actually the nature of any control system. They always go over and under their set point and perform constant adjustments when that happens. They effectively always have error that is always being corrected with an average that is the correct value. This is a standard problem solved all the time.

            A slight rise or dip in the terrain or a gust of wind would be plenty to change the car's velocity by a few MPH.

            The sensors are accurate enough that the deceleration can be measured as soon as the car starts traveling up hill at a fraction of a degree. With in a tiny fraction of a second the throttle will be adjusted and another reading from the sensors is taken.

            The physical system that the computer is controlling (engine & brakes) isn't accurate enough to avoid inevitable overshoot when the computer applies control.

            Control systems always overshoot, this is not realistic to complain about. Perhaps you could say the performance is too high and the car will overshoot too far and not maintain the speed requirements but that's a problem the control loop tuning solves.

            The only way you'd get a control system to keep the car exactly at a given speed would be to have it constantly alternate gas & brake to speed up and slow down.

            Is that how you drive? Jesus christ - you are always on the throttle or brakes? Do you even have a license or drive a car?

            A car constantly braking & accelerating would appear erratic to other drivers, probably resulting in them speeding around it to "get away from" the erratic driver.

            Dude the brakes and the gas pedal are not binary.

            Until/unless 100% of cars become fully autonomous, self-driving cars need to behave in a fashion that's similar to what human drivers expect of other human drivers. Drifting a few MPH over or under the posted limit is normal, and thus it's actually the safest way for an autonomous car to behave. The reactions of human drivers around it are the most likely source of accidents.

            Well moving above the speed limit because it's safer in traffic isn't really a thing to worry about in the middle of the night, now is it?

          • (Score: 0) by Anonymous Coward on Wednesday March 21 2018, @10:04PM

            by Anonymous Coward on Wednesday March 21 2018, @10:04PM (#656360)

            While I don't do it very often, I can keep within 1 mph of any speed limit on a reasonably flat straight road. Oftentimes even on curves and differing grades if I have the patience and attention. A robot has none of these limitations and having worked on cars from the 90s through late 2000s, I can tell you almost all of them have both the sensors and the ECU/ABS speedometer accuracy to hold speed at tenths of a mph within the vehicles operating range, and that is on MANUAL cars. Automatic cars with ABS, TCS, throttle and steering by wire should have no problems under conditions outside of 5mph or 5 percent grade in holding their speed steady at or below the posted speed limit. Furthermore having had a discussion about this just the other day: Most of the modern net-enabled GPS navigation systems have sign-accurate speed limit markers including with their navigation service, to notify you to slow down/speed up on a stretch of road. Given that, there really is no excuse for an autonomous car to not be driving almost exactly the speed limit. The entire point of autonomous vehicles is to provide better safety, reliability, and reproducability than human drivers can. If they aren't doing that then they fail at autonomous cars.

          • (Score: 0) by Anonymous Coward on Thursday March 22 2018, @12:54AM

            by Anonymous Coward on Thursday March 22 2018, @12:54AM (#656407)

            It's unrealistic to expect the car to be able to maintain 35.00 MPH, but we're talking 38 in a 35 MPH zone here. That's almost 10% off. The simple cruise control on my relatively uninteresting car would have no problem with that, and it doesn't even have control of the brakes, only the throttle. Most human drivers could do a pretty reasonable job at it too, which makes me wonder what it must be like following you around if you can't maintain your speed within 35 +/- 3 MPH.

    • (Score: 2) by rigrig on Tuesday March 20 2018, @08:15AM (2 children)

      by rigrig (5129) <soylentnews@tubul.net> on Tuesday March 20 2018, @08:15AM (#655301) Homepage

      it’s very clear it would have been difficult to avoid this collision in any kind of mode (autonomous or human-driven) based on how she came from the shadows right into the roadway,

      Not the same as not being at fault, especially as the car was speeding (a bit), and aren't autonomous vehicles supposed to have a bunch of sensors that don't care about lighting conditions?

      --
      No one remembers the singer.
      • (Score: 0) by Anonymous Coward on Tuesday March 20 2018, @12:42PM (1 child)

        by Anonymous Coward on Tuesday March 20 2018, @12:42PM (#655341)

        based on how she came from the shadows right into the roadway

        Wait, wasn't a big part of the argument for why autonomous vehicles will be safer that they are using radar and lasers to locate obstacles, and so won't be affected by shadows...

        • (Score: 2) by pendorbound on Tuesday March 20 2018, @03:31PM

          by pendorbound (2688) on Tuesday March 20 2018, @03:31PM (#655420) Homepage

          Better isn't the same thing as perfect. If you're only willing to accept a completely perfect and 100% safe car, you should probably never leave your home.