This is definitely not a case of someone jumping out in the road, as was previously reported. Sure, jaywalking at night is a bad idea, but this is exactly the type of thing an automated driving system should catch and avoid
Tempe Police Release Video of Uber Accident
51–60 of 1001 posts
Re: Tempe Police Release Video of Uber Accident
#52How did LIDAR and IR not catch that? That seems like a pretty serious problem. It's clear from the video that a human driver actually would've had more trouble since the pedestrian showed up in the field of view right before the collision, yet that's in the visible spectrum. When I argue for automated driving (as a casual observer), I tell people about exactly this sort of stuff (a computer can look in 20 places at t…
Yeah releasing the sensory data beyond human visible spectrum would be way more informative about if a better designed AV would have dealt with this better.
I'm glad it was not me driving down that road that night, I don't think I could have prevented it.
Re: Tempe Police Release Video of Uber Accident
#53There was almost a second where the woman was clearly visible and yet the car did not attempt to emergency break. I would not expect the vehicle to safely stop, but it could have definitely slowed down and reduce the energy of the collision.
This is not criticism against self-driving tech, I would not expect alert human driver to avoid collision either due to limited reaction time. With technology however, we should be able to do better than humans, particularly when it comes to reaction times. Clearly, there is still a some work to do.
Re: Tempe Police Release Video of Uber Accident
#54Wow, I had thought it impossible for the Uber AV to be in the right lane, and for the victim to be crossing left-to-right, for the police to claim that the victim couldn't be seen until it was too late. But I didn't expect the road to be so dark, given what we saw in the accident photos (which might have been over-exposed?) and Google Maps, which showed a lot of street/sidewalk lighting. In the moment that we can ful…
Digital image quality is generally far inferior to naked eye visibility, but even in the image, the pedestrian can be seen for at least a full second.
The driver clearly was unfortunately not paying attention. She clearly reacts within a few hundred milliseconds of looking up, which is to say immediate reaction time in terms of mental processing. That means that we cannot use her reaction time to gauge how early she could have seen the pedestrian to compensate for poorer optics of digital cameras.
As I understand LIDAR, it should work even at night (as it generates its own light and measures response time). This is a pedestrian, walking across a road, with nothing possible to occlude her. There is no reason it should not have identified an obstacle in the road (she's been in the road for several seconds, after all, having crossed at least 3 lanes of traffic by that point). Even if the visual camera had problems identifying the object, LIDAR should have flagged it.
Re: Tempe Police Release Video of Uber Accident
#55"There is no such thing as an accident. Only negligence." This was no accident. The jaywalker was negligent for crossing the street as she did. Uber is negligent for not having autonomous vehicle standards that clearly exceed what humans can do (I don't think a human could have avoided her based on the video). Not sure about the driver. Depends on his training/job etc...
The 'safety driver' wasn't even looking out the windscreen. If your job is to act as a backup for an untrusted system, you are responsible.
Re: Tempe Police Release Video of Uber Accident
#56Okay, isn't this the kind of a situation where the machine was supposed to excel?
The dynamic range of the video is very low so it looks like the victim comes out of complete darkness but weren't all these sensors supposed to see the obstacle even in complete darkness?
BTW, please consider the low dynamic range of the video when commenting on the human's ability to avoid that accident. After driving in the dark for a while your eyes will adapt and you'll be able to see much more details in the shadows than a regular video camera can record.
Re: Tempe Police Release Video of Uber Accident
#57How did LIDAR and IR not catch that? That seems like a pretty serious problem. It's clear from the video that a human driver actually would've had more trouble since the pedestrian showed up in the field of view right before the collision, yet that's in the visible spectrum. When I argue for automated driving (as a casual observer), I tell people about exactly this sort of stuff (a computer can look in 20 places at t…
You mean like the Mercedes system from 2007 - https://www.youtube.com/watch?v=l6DMidknVCo ? I agree that IR should have caught the bicyclist earlier, and that it didn't suggests that these cars might also kill quite a few deer with a similar low light profile. (or this one from Cadillac : https://www.youtube.com/watch?v=SPIpNG6mLCY)
Re: Tempe Police Release Video of Uber Accident
#58Wow, I had thought it impossible for the Uber AV to be in the right lane, and for the victim to be crossing left-to-right, for the police to claim that the victim couldn't be seen until it was too late. But I didn't expect the road to be so dark, given what we saw in the accident photos (which might have been over-exposed?) and Google Maps, which showed a lot of street/sidewalk lighting. In the moment that we can ful…
The dynamic range of small sensor digital cameras is so poor that it is impossible to judge whether or not she would have been visible to the naked eye based on the video.
That's a great point. I did think, looking at the video, that the street was unnaturally dark. The camera seems to be exposed for night lights and the headlights too.
Re: Tempe Police Release Video of Uber Accident
#59How did LIDAR and IR not catch that? That seems like a pretty serious problem. It's clear from the video that a human driver actually would've had more trouble since the pedestrian showed up in the field of view right before the collision, yet that's in the visible spectrum. When I argue for automated driving (as a casual observer), I tell people about exactly this sort of stuff (a computer can look in 20 places at t…
It isn't at all clear from the video. Video cameras of this quality vastly underperform human vision in low-light conditions.
Like parent said, it's low-light in the visible spectrum, but I'd totally expect these vehicles to scatter tons of light in non-visible spectrums, making these conditions well-lit in those spectrums. Like a bat using echolocation.
Re: Tempe Police Release Video of Uber Accident
#60The first question that comes to my mind watching this, is, why didn't the _pedestrian_ see the _car_? It looks as though she didn't even look up until the car was only a few feet away from here. I'm not just trying to "blame the victim". The car "should" have responded sooner. But see-and-avoid works both ways. If you're choosing to cross a road in the dark, you, too, have a responsibility to understand the drivers'…
A theory that occurred to me last night is that if she was homeless, she may also have had an untreated eye condition such as myopia.