Hacker News Board Hacker News, browsed like a classic message board
Live data from Hacker News

Tempe Police Release Video of Uber Accident

twitter.com

Post reply on HN
Print view View original

231–240 of 1001 posts

Re: Tempe Police Release Video of Uber Accident

#231

Post by brokenmachine »

lolsal wrote: post #8
foobarbazetc wrote: post #5

Earlier quoted context omitted.

That’s only because there’s no light on her until the end. The car hardware should have seen her. If it doesn’t work in the dark it shouldn’t be on the road imho.

> That’s only because there’s no light on her until the end. Yes, quiet true. > The car hardware should have seen her. If it doesn’t work in the dark it shouldn’t be on the road imho. I tend to agree. I think unfortunately this video indicates that the car did not perform worse than a human. It's still valuable to have cars that perform better than humans, even if they aren't perfect.

> I tend to agree. I think unfortunately this video indicates that the car did not perform worse than a human. It's still valuable to have cars that perform better than humans, even if they aren't perfect.

I disagree. I think this video shows that the car performs much worse than I would expect a human to.

I have my own dashcam videos at night that look very similar to this, and I definitely think I would not have hit a large object such as a pedestrian walking a bicycle on a straight road such as this in a 35mph zone.

Re: Tempe Police Release Video of Uber Accident

#232

Post by perennate »

aianus wrote: post #155
brokenmachine wrote: post #119

Earlier quoted context omitted.

What's your point? She wasn't driving a car. Imagine if the uber hardware was installed into a petrol tanker. It just just as pointless a discussion. If self driving cars can't be trusted to not smash into any obstacle that appears such as a deer/roadwork/child/runaway shopping cart, then they shouldn't be permitted on the road.

The point is that humans who have poorer sensors and do even stupider things are allowed to drive. So the bar for self-driving is a lot lower than people around here make it sound.

I actually believe that the kind of situation shown in the video (pedestrian walking across road at night in the path of a car, where car has to brake hard to avoid a collision) happens much more frequently than most people think. And most of the time the pedestrian comes out fine because most drivers are looking ahead most of the time, and someone walking across the road is a big abnormal event that is easy for a human to notice and react to.

I don't have any data to support this, other than a vague memory of doing something similar many years ago. Although if I remember correctly, that was a more major road that was probably better lit. But I still find it hard to believe that human vision would have been as lousy as the camera's.

Re: Tempe Police Release Video of Uber Accident

#233

Post by lsaferite »

darkerside wrote: post #79
HankB99 wrote: post #27

This seems like a simple case. I would have expected the driver assist in my 2017 Subaru to have reacted to something in the road. I'm surprised that the much more sophisticated self driving system did not.

I wonder if this was affected by the fact that the pedestrian was in another lane up until the very last second. Perhaps the car detected the pedestrian but failed to consider it an obstacle since it wasn't in its direct path. It could be unexpectedly difficult to account for pedestrian crossing speed if it caused automated cars to stop when a car in the next lane happened to "wobble" towards the automated car's lane…

Even if you imagine it thought it was another vehicle stopped in the other lane of traffic, the SDV should have slowed. There's no way you should assume a stopped vehicle isn't hiding another danger and blowing through while speeding is what stupid humans do and what SDVs shouldn't.

Re: Tempe Police Release Video of Uber Accident

#234

Post by gok »

hcknwscommenter wrote: post #203
gok wrote: post #188

Earlier quoted context omitted.

In pitch darkness, IR cameras can only see more than a visual light camera if you somehow had IR headlights that were more powerful than the visual headlights (they probably don't; that would blind other AVs just like high beams blind other drivers). It doesn't grant you the ability to see in the dark with infinite range. Lidar can sense shapes with more range, but at the cost of dramatically worse resolution and lat…

Human beings radiate IR do they not? Silly question, of course they do. Perhaps not above an Arizona level of thermal background?

Thermal imaging and near-infrared imaging are not the same. Waymo/Uber/etc are not using thermography for their autonomous vehicles.

Re: Tempe Police Release Video of Uber Accident

#235

Post by ars »

ajross wrote: post #92
TrainedMonkey wrote: post #53

There 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…

Do we know the vehicle didn't brake? There's no audio to tell us anything, and a car takes "almost a second" to rock forward on its suspension when reacting to brake input too. FWIW, I think I might see the nose bobbing a bit just before the collision, but honestly can't tell. I certainly don't think you can say "did not attempt to brake" based on the evidence at hand, basically.

> Do we know the vehicle didn't brake?

Because the nose of the car doesn't go down. Especially in an emergency braking situation the nose of the car will go significantly down, and you would see that in the video.

Re: Tempe Police Release Video of Uber Accident

#236

Post by heartbreak »

hcknwscommenter wrote: post #137
jackpirate wrote: post #78

Earlier quoted context omitted.

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. That's not at all clear to me. I don't know too much about cameras, but it looks to me like the camera is making the scene appear much darker than it actually is. In the video, you can see many street lights projecting down…

Exactly. The dynamic range of the human eye is vastly better than a visible spectrum camera.

> The dynamic range of the human eye is vastly better than a visible spectrum camera.

Certainly better than any camera mounted on a dashboard.

It's honestly a bit surreal how the pedestrian appears out of the splotch of pure darkness in the frame. That's low dynamic range and resolution (or high compression) at work, not how light behaves in reality.

Re: Tempe Police Release Video of Uber Accident

#237

Post by hcknwscommenter »

stefan_ wrote: post #31

If there is something that isn't going to help the public perception of autonomous cars at all, it's releasing a compressed to shit capture of another video showing a single camera angle from dozens. I would say it's a deliberate attempt to manipulate if I didn't also strongly believe ignorance on part of the police department has lead them to believe that autonomous cars could even exit a parking lot without data fr…

I think releasing much higher quality and higher dynamic range and multi-spectrum video is going to be a LOT more damming of driverless car tech. Just look at all the commentators here saying they would have not been able to stop in time based on this terrible camera footage (ignoring the vastly better capabilities of their own human eyes).

Re: Tempe Police Release Video of Uber Accident

#238

Post by TillE »

stefan_ wrote: post #163
frgtpsswrdlame wrote: post #133

Earlier quoted context omitted.

Uber is more than welcome to release their video(s) themselves if this one represents an inaccurate view.

Given there are no obstructions in the road, LIDAR (or even the decent cameras from the sensor array on top) would obviously show her on a clear collision path many seconds before impact. So I guess this is the ultimate litmus test if the culture at Uber has changed. They let this video stand and I don't personally think they should be allowed to drive another metre autonomously on public roads.

Any engineer on that project would have to be horrified that their system didn't prevent this accident. It failed, completely.

I would be shocked if we see Uber resume testing any time soon, if at all.

Re: Tempe Police Release Video of Uber Accident

#240

Post by philip1209 »

ars wrote: post #74

Direct link: https://twitter.com/TempePolice/status/976585098542833664 The interior view tells you everything you need to know about why anything except perfection in self driving cars will lead to more accidents. The driver specially selected and trained for exactly this job, does not look at the road! If someone specially trained for that won't do it, the general public certainly will not. Also, in the video you ca…

What is displayed on the center console? Is it possible that the driver was looking at a live video feed? I believe that some standard cars show their night vision feed there.
Post reply on HN