4 ms·

I still see facts as: 1) Nothing was reflective 2) She was walking across the road between street lights in a random space. 3) If a person hit the woman it w

I still see facts as: 1) Nothing was reflective 2) She was walking across the road between street lights in a random space. 3) If a person hit the woman it would not have been homicide or manslaughter even if we didn't know the driver was on their phone. I would say the pedestrian was legally more than 50% at fault in this horrible accident. http://www.alllaw.com/articles/nolo/auto-accident/driver-at-fault-pedestrian-car.html http://www.alllaw.com/articles/nolo/auto-accident/driver-at-...

Filligree 9y ago
The test for a self-driving car isn't "does it drive as well as a human," it's "does it drive as well as a self-driving car which normally beats a human should." This might sound like it makes no difference, but the car had LIDAR. That the scene was badly lit, whether true or not, is irrelevant; the car didn't use ambient light to see. If it could hit her under these circumstances, then it could just as easily hit her at noon.
- zodPod 9y ago
  This is what gets me! I really wish they would talk about why the LIDAR didn't notice her. Was it off? Was it malfunctioning? Was there some other issue there? I've been puzzled about how it didn't see her!
  - FreeFull 9y ago
    The theory I've seen the most is that the LIDAR probably did detect her just fine, but Uber's software failed to act on the data.
  - thaumaturgy 9y ago
    According to Reuters, Uber had changed from a network of 7 LIDAR sensors on their vehicles to just one: https://www.reuters.com/article/us-uber-selfdriving-sensors-insight/ubers-use-of-fewer-safety-sensors-prompts-questions-after-arizona-crash-idUSKBN1H337Q https://www.reuters.com/article/us-uber-selfdriving-sensors-...
- drjasonharrison 9y ago
  According to discussions I have read. The LIDAR was disconnected as part of a test of "visual only" navigation. Why didn't they keep the LIDAR on for safety?
  - baldfat 9y ago
    If they did that I would say sue them out of existence. That killed a woman and harmed the industry.
- SpikeDad 9y ago
  Which brings up the real question - do we want self-driving cars that ONLY drive as well as humans. I for one would be terrified at that criteria.
panarky 9y ago
Human drivers and other autonomous cars could have avoided this accident. Uber's systems failed spectacularly, and should not have been on public roads. Now we learn that Uber and the governor secretly collaborated to enable this, to neutralize any effective oversight, and to keep the public in the dark. Blaming the victim doesn't illuminate any of these issues.
dragonwriter 9y ago
> I would say the pedestrian was legally more than 50% at fault in this horrible accident. Which is fine for Uber-as-operator if they were operating in a state which uses a contributory negligence or modified comparative negligence rule, but a number of jurisdictions—Arizona, for instance—use a pure comparative negligence rule where, e.g., an injured party that is 99% responsible can recover 1% of damages from a party that is 1% responsible. Moreover, it's no help for Uber-as-manufacturer if strict product liability principles are applied. Now, it's not clear to me that they would apply to Uber as manufacturer as long as the product isn't sold to a third party, but even if they don't apply under current law, Uber using operating a service directly with the product rather than selling it to other operators as an end-run around liability for an unreasonably dangerous product invites a change to the law which would extend such liability to manufacturer-operators.
andrewla 9y ago
I would like to see the more detailed telemetry that the car gathered at the time. For a human driving the car, the dash cam together with personal testimony would probably put the issue to rest; although the fact that the pedestrian was in the roadway before they entered the lane looks very suspicious because a human likely would have seen them because of headlights on an open road even if not illuminated by streetlights. Nonetheless, the dash cam shows what it shows, and human sense and memory is fragile. Not so for a self-driving car -- the memory is not fragile, and should be preserved. If it is not preserved and there is no remaining copy of the car's telemetry at the time, that in itself is a detail that can point towards gross negligence on the part of Uber. If it is preserved, the question becomes, would a "reasonable person/software" have acted based on what was shown there, and that goes towards whether there is negligence or recklessness. But we (or at least, investigators) need to see that telemetry to make that decision.
- hcknwscommenter 9y ago
  >the dash cam shows what it shows The dash cam is either proof that: i) that dash cams are far inferior to human eyes in terms of dynamic range (certainly true); and/or ii) the video was purposely darkened or compressed to mislead the public into believing a human driver would have hit the pedestrian (see, https://arstechnica.com/cars/2018/03/police-chief-said-uber-victim-came-from-the-shadows-dont-believe-it/ https://arstechnica.com/cars/2018/03/police-chief-said-uber-... ). I think it was both i) and ii).
  - cr0sh 9y ago
    I'd like to see the actual data it gathered as well. Plus, I'd like to see the software system, ML models, etc - to run the data against. The dash cam video is very suspicious. What I want to see is the actual computer vision camera footage; it's never been made clear whether the footage that has been publically made available is from a dashcam or CV camera; I suspect the former. On the notion of the LIDAR - the device spits out 64 beams; it appeared in the video the woman who was hit was wearing a fuzzy and dark color jacket (was she? I don't know for certain). If this is so - it is possible that any reflections for the LIDAR returns only came from the bottom half and the bicycle at best. The top half would have been either absorbed, or possibly filtered as "noise" prior to the classification algorithm. Which is why the ML algo's matter - how did the classifier classify the data? Did it think what was moving across the road was something other than a human? Without the data and the algorithm, there's no good way to know. But now I hear it's possible the LIDAR was turned off to run a "vision only" driving system. Was it? Who knows; we may never get the truth of what happened here. Form a personal standpoint, it saddens and perplexes me - and makes me upset to a degree - that this happened here in my city. I recently completed the Udacity Self-Driving Car Engineer nanodegree, and I tend to wonder if this incident will somehow potentially color any future employment opportunities in that field, should I pursue them... The whole political angle of this thing - sigh.
jadedhacker 9y ago
These facts are only relevant if you disregard LIDAR and you believe that the camera showed what the human eye would see. Videos on youtube (https://www.youtube.com/watch?v=CRW0q8i3u6E https://www.youtube.com/watch?v=CRW0q8i3u6E and also https://www.youtube.com/watch?v=1XOVxSCG8u0&feature=youtu.be https://www.youtube.com/watch?v=1XOVxSCG8u0&feature=youtu.be) and ABC news footage (http://abc7.com/video-shows-self-driving-ubers-fatal-collision/3243922/ http://abc7.com/video-shows-self-driving-ubers-fatal-collisi...) at the scene of the crash paint a much brighter picture. The street lights in the footage appear to be distinct while the background is dark, indicating the gain of the sensor would appear to be reduced. Try taking a picture of streetlights at night, if the street is visible, they will appear as stars. Additionally, the safety driver was not paying attention. EDIT: In my opinion, the more you think about this crash the worse Uber looks. The LIDAR should have worked, the lighting was likely brighter than the video shows, the weather was good, no obvious distractor objects, the road was straight, no parties were impaired. The pedestrian had a big bulky bike and crossed at right angles to the car. This is nearly the most basic test of self-driving cars and it failed. The only way Uber would look worse is if it had happened in day time. On top of that, Uber disabled the Volvo auto-braking safety system. (https://www.sfgate.com/business/article/Uber-Disabled-Volvo-SUV-s-Standard-Safety-System-12782878.php https://www.sfgate.com/business/article/Uber-Disabled-Volvo-...)
vkou 9y ago
> If a person hit the woman it would not have been homicide or manslaughter even if we didn't know the driver was on their phone. That's because the bar for vehicular manslaughter is incredibly high. There is a huge range of shitty road behaviour that is not illegal, but you should never do, for the sake of yourself and other people. Driving faster then you can see is one of those behaviors. Hitting someone because you had legal right of way is another. Right of way is good and all, but your first job as a driver is to avoid crashes. Right of way is just a set of guidelines for avoiding crashes... And a way to assign legal blame when one happens. When I sea kayak, I have right of way against large boats in some situations. I'd be a complete, and dead idiot if I were to exercise my legal right to stay on course.
phyzome 9y ago
If you even believe the video is representative of what a human would see (which is totally false, as recent local videos of that stretch have shown), then the car was driving beyond its headlights. That's unsafe driving. Deer are not reflective. Random objects that have fallen off a truck are not reflective. People are often not reflective, and there is no law requiring them to be. And this woman was wheeling a bright pink bike, which shows up quite well. (It may also have had standard spoke reflectors, which might not show up in this video. hard to say.) And finally, there weren't any crosswalks in that area for her to use. Yes, she misjudged the speed of the car, but her choice to cross the road at all is pretty solid.
- koreyb 9y ago
  Well there is a crosswalk at the light..intersection of Curry & Mill. What's interesting is the X paved walkway in the median. You can see it here: https://www.google.com/maps/place/Tempe,+AZ/@33.4364759,-111.9429782,241m/data=!3m1!1e3!4m5!3m4!1s0x872b0898a6aa80e3:0xa2dd4aad392cb41f!8m2!3d33.4255104!4d-111.9400054 https://www.google.com/maps/place/Tempe,+AZ/@33.4364759,-111...
- romwell 9y ago
  >If you even believe the video is representative of what a human would see (which is totally false, as recent local videos of that stretch have shown), then the car was driving beyond its headlights. That's unsafe driving. This point needs to be understood by everyone discussing the issue.

Filligree 9y ago

The test for a self-driving car isn't "does it drive as well as a human," it's "does it drive as well as a self-driving car which normally beats a human should." This might sound like it makes no difference, but the car had LIDAR. That the scene was badly lit, whether true or not, is irrelevant; the car didn't use ambient light to see. If it could hit her under these circumstances, then it could just as easily hit her at noon.

zodPod 9y ago
This is what gets me! I really wish they would talk about why the LIDAR didn't notice her. Was it off? Was it malfunctioning? Was there some other issue there? I've been puzzled about how it didn't see her!
- FreeFull 9y ago
  The theory I've seen the most is that the LIDAR probably did detect her just fine, but Uber's software failed to act on the data.
- thaumaturgy 9y ago
  According to Reuters, Uber had changed from a network of 7 LIDAR sensors on their vehicles to just one: https://www.reuters.com/article/us-uber-selfdriving-sensors-insight/ubers-use-of-fewer-safety-sensors-prompts-questions-after-arizona-crash-idUSKBN1H337Q https://www.reuters.com/article/us-uber-selfdriving-sensors-...
drjasonharrison 9y ago
According to discussions I have read. The LIDAR was disconnected as part of a test of "visual only" navigation. Why didn't they keep the LIDAR on for safety?
- baldfat 9y ago
  If they did that I would say sue them out of existence. That killed a woman and harmed the industry.
SpikeDad 9y ago
Which brings up the real question - do we want self-driving cars that ONLY drive as well as humans. I for one would be terrified at that criteria.

panarky 9y ago

Human drivers and other autonomous cars could have avoided this accident. Uber's systems failed spectacularly, and should not have been on public roads. Now we learn that Uber and the governor secretly collaborated to enable this, to neutralize any effective oversight, and to keep the public in the dark. Blaming the victim doesn't illuminate any of these issues.

dragonwriter 9y ago

> I would say the pedestrian was legally more than 50% at fault in this horrible accident. Which is fine for Uber-as-operator if they were operating in a state which uses a contributory negligence or modified comparative negligence rule, but a number of jurisdictions—Arizona, for instance—use a pure comparative negligence rule where, e.g., an injured party that is 99% responsible can recover 1% of damages from a party that is 1% responsible. Moreover, it's no help for Uber-as-manufacturer if strict product liability principles are applied. Now, it's not clear to me that they would apply to Uber as manufacturer as long as the product isn't sold to a third party, but even if they don't apply under current law, Uber using operating a service directly with the product rather than selling it to other operators as an end-run around liability for an unreasonably dangerous product invites a change to the law which would extend such liability to manufacturer-operators.

andrewla 9y ago

I would like to see the more detailed telemetry that the car gathered at the time. For a human driving the car, the dash cam together with personal testimony would probably put the issue to rest; although the fact that the pedestrian was in the roadway before they entered the lane looks very suspicious because a human likely would have seen them because of headlights on an open road even if not illuminated by streetlights. Nonetheless, the dash cam shows what it shows, and human sense and memory is fragile. Not so for a self-driving car -- the memory is not fragile, and should be preserved. If it is not preserved and there is no remaining copy of the car's telemetry at the time, that in itself is a detail that can point towards gross negligence on the part of Uber. If it is preserved, the question becomes, would a "reasonable person/software" have acted based on what was shown there, and that goes towards whether there is negligence or recklessness. But we (or at least, investigators) need to see that telemetry to make that decision.

hcknwscommenter 9y ago
>the dash cam shows what it shows The dash cam is either proof that: i) that dash cams are far inferior to human eyes in terms of dynamic range (certainly true); and/or ii) the video was purposely darkened or compressed to mislead the public into believing a human driver would have hit the pedestrian (see, https://arstechnica.com/cars/2018/03/police-chief-said-uber-victim-came-from-the-shadows-dont-believe-it/ https://arstechnica.com/cars/2018/03/police-chief-said-uber-... ). I think it was both i) and ii).
- cr0sh 9y ago
  I'd like to see the actual data it gathered as well. Plus, I'd like to see the software system, ML models, etc - to run the data against. The dash cam video is very suspicious. What I want to see is the actual computer vision camera footage; it's never been made clear whether the footage that has been publically made available is from a dashcam or CV camera; I suspect the former. On the notion of the LIDAR - the device spits out 64 beams; it appeared in the video the woman who was hit was wearing a fuzzy and dark color jacket (was she? I don't know for certain). If this is so - it is possible that any reflections for the LIDAR returns only came from the bottom half and the bicycle at best. The top half would have been either absorbed, or possibly filtered as "noise" prior to the classification algorithm. Which is why the ML algo's matter - how did the classifier classify the data? Did it think what was moving across the road was something other than a human? Without the data and the algorithm, there's no good way to know. But now I hear it's possible the LIDAR was turned off to run a "vision only" driving system. Was it? Who knows; we may never get the truth of what happened here. Form a personal standpoint, it saddens and perplexes me - and makes me upset to a degree - that this happened here in my city. I recently completed the Udacity Self-Driving Car Engineer nanodegree, and I tend to wonder if this incident will somehow potentially color any future employment opportunities in that field, should I pursue them... The whole political angle of this thing - sigh.

jadedhacker 9y ago

These facts are only relevant if you disregard LIDAR and you believe that the camera showed what the human eye would see. Videos on youtube (https://www.youtube.com/watch?v=CRW0q8i3u6E https://www.youtube.com/watch?v=CRW0q8i3u6E and also https://www.youtube.com/watch?v=1XOVxSCG8u0&feature=youtu.be https://www.youtube.com/watch?v=1XOVxSCG8u0&feature=youtu.be) and ABC news footage (http://abc7.com/video-shows-self-driving-ubers-fatal-collision/3243922/ http://abc7.com/video-shows-self-driving-ubers-fatal-collisi...) at the scene of the crash paint a much brighter picture. The street lights in the footage appear to be distinct while the background is dark, indicating the gain of the sensor would appear to be reduced. Try taking a picture of streetlights at night, if the street is visible, they will appear as stars. Additionally, the safety driver was not paying attention. EDIT: In my opinion, the more you think about this crash the worse Uber looks. The LIDAR should have worked, the lighting was likely brighter than the video shows, the weather was good, no obvious distractor objects, the road was straight, no parties were impaired. The pedestrian had a big bulky bike and crossed at right angles to the car. This is nearly the most basic test of self-driving cars and it failed. The only way Uber would look worse is if it had happened in day time. On top of that, Uber disabled the Volvo auto-braking safety system. (https://www.sfgate.com/business/article/Uber-Disabled-Volvo-SUV-s-Standard-Safety-System-12782878.php https://www.sfgate.com/business/article/Uber-Disabled-Volvo-...)

vkou 9y ago

> If a person hit the woman it would not have been homicide or manslaughter even if we didn't know the driver was on their phone. That's because the bar for vehicular manslaughter is incredibly high. There is a huge range of shitty road behaviour that is not illegal, but you should never do, for the sake of yourself and other people. Driving faster then you can see is one of those behaviors. Hitting someone because you had legal right of way is another. Right of way is good and all, but your first job as a driver is to avoid crashes. Right of way is just a set of guidelines for avoiding crashes... And a way to assign legal blame when one happens. When I sea kayak, I have right of way against large boats in some situations. I'd be a complete, and dead idiot if I were to exercise my legal right to stay on course.

phyzome 9y ago

If you even believe the video is representative of what a human would see (which is totally false, as recent local videos of that stretch have shown), then the car was driving beyond its headlights. That's unsafe driving. Deer are not reflective. Random objects that have fallen off a truck are not reflective. People are often not reflective, and there is no law requiring them to be. And this woman was wheeling a bright pink bike, which shows up quite well. (It may also have had standard spoke reflectors, which might not show up in this video. hard to say.) And finally, there weren't any crosswalks in that area for her to use. Yes, she misjudged the speed of the car, but her choice to cross the road at all is pretty solid.

koreyb 9y ago
Well there is a crosswalk at the light..intersection of Curry & Mill. What's interesting is the X paved walkway in the median. You can see it here: https://www.google.com/maps/place/Tempe,+AZ/@33.4364759,-111.9429782,241m/data=!3m1!1e3!4m5!3m4!1s0x872b0898a6aa80e3:0xa2dd4aad392cb41f!8m2!3d33.4255104!4d-111.9400054 https://www.google.com/maps/place/Tempe,+AZ/@33.4364759,-111...
romwell 9y ago
>If you even believe the video is representative of what a human would see (which is totally false, as recent local videos of that stretch have shown), then the car was driving beyond its headlights. That's unsafe driving. This point needs to be understood by everyone discussing the issue.