Tempe Police Release Video of Uber Accident • Hacker News Board

Re: Tempe Police Release Video of Uber Accident

#191

Post by LukaAl » Thu, Mar 22, 2018, 12:22 AM UTC

freehunter wrote: post #42

jonknee wrote: post #30

Earlier quoted context omitted.

> there is no such thing as a too reflective vest for bicyclists And to not jaywalk a bike across a street in the dark!

Since I've seen this be a very controversial statement here in the past, I'd just like to point out to everyone that American drivers on four lane roads are not expecting pedestrians to be in the middle of the road and not at a crosswalk. That is not a normal thing to happen in America, and no one expects it to happen. It is a very rare situation. --edit: okay guys... no where did anyone argue that you shouldn't be w…

Wow! You are saying no one in the US expects, or at least consider that could be an obstacle on the street in front of you? It doesn't mean necessarily a pedestrian. An animal? Cargo lost by a truck in front of you?

It is the responsibility of the driver to be alert and try to minimize the impact. Just because a cyclist shouldn't cross a four-lane road in the middle of the night, doesn't mean you shouldn't look for obstacles.

Re: Tempe Police Release Video of Uber Accident

#192

Post by hcknwscommenter » Thu, Mar 22, 2018, 12:22 AM UTC

jonknee wrote: post #127

netsharc wrote: post #114

Earlier quoted context omitted.

Which makes me wonder if the Uber autocar is just relying on camera vision to drive itself... If it is, and it's been lying to authorities (I don't know what they said to the authorities about their cars' capabilities), that could be big.

Far more likely is that this is a standard dash cam that the police know how to pop the SD card out of and access the video...

The point is not germane. What is germane is that a car that supposedly uses LIDAR and infrared, and presumably was approved by the regulators on the basis of such, should have had no problem seeing the pedestrian as LIDAR and infrared are unaffected by night and at least shown some indication of braking but did not. This suggests that the car does not in fact utilize any of those fancy (non-vis) detection methods. Alternatively, these fancy detection methods were fooled by the bicycle and thus misclassified as an error or something.

Re: Tempe Police Release Video of Uber Accident

#193

Post by danso » Thu, Mar 22, 2018, 12:22 AM UTC

jedberg wrote: post #40

While I'm surprised the sensors on the car didn't catch the cyclist, I pretty sure if I had been driving and paying full attention I would have hit the bike too. Personally, given this video, I'd still be totally comfortable with a self driving car on the road with no safety driver (who clearly didn't make a difference here anyway, not that they could have).

I agree in that I consider the safety driver to be irrelevant in this situation, because of the reaction time needed to take control, nevermind hit the brakes, nevermind after a complacent state of mind.

But as others have pointed out, it's hard to tell what the driver could have seen just from the dashcam video, which for starters, has a much less dynamic range than human vision. Because in the accident scene photos/videos, the ambient street light is much more visible [0]. I'm guessing driving down that Arizona road at night can't possibly be as pitch dark as the camera shows it, given how that area seems to be used for entertainment venues/festivals.

[0] http://sanfrancisco.cbslocal.com/2018/03/19/uber-self-drivin...

Re: Tempe Police Release Video of Uber Accident

#194

Post by Asooka » Thu, Mar 22, 2018, 12:22 AM UTC

ChuckMcM wrote: post #24

This time it actually has the video. And it is a poignant reminder that there is no such thing as a too reflective vest for bicyclists.

Even if you're not a cyclist, it's a good idea to wear something with reflective stripes when you're out on dark roads. This is actually the law in several European countries.

Re: Tempe Police Release Video of Uber Accident

#195

Post by pdonis » Thu, Mar 22, 2018, 12:23 AM UTC

tedd4u wrote: post #126

A few people have commented along the lines of "I probably couldn't have stopped in time even if I had been driving." I think it's important as a driver to hit the brakes as hard as you can even if you don't think you'll be able to avoid collision, because the energy is reduced greatly even with a small amount of deceleration (F=ma). Probability of death in car/pedestrian collision is reduced from over 80% to under 1…

> I think it's important as a driver to hit the brakes as hard as you can even if you don't think you'll be able to avoid collision

It's good to hit the brakes, yes, but the best way to avoid almost all collisions is to steer, not just to brake. Braking combined with steering sharply to the left (since the pedestrian was moving to the right) would have been the best response to this situation.

(Note that, before anti-lock brakes became common, "hit the brakes as hard as you can" was not good advice, precisely because it would lock the wheels and prevent you from steering. In a car with anti-lock brakes, that's not an issue.)

Re: Tempe Police Release Video of Uber Accident

#196

Post by ThrustVectoring » Thu, Mar 22, 2018, 12:23 AM UTC

jcranmer wrote: post #124

lhorie wrote: post #10

When I cross-reference the lane lines at 0:03 with google maps[1], I get an estimate of the car being about 60ft-70ft away from the victim when I was first able to spot her shoe in the video. Some googling[2] says human reaction at 40mph would've taken 59ft, with an additional 80ft of actual stopping time. This seems in line with the police report that "it would have been difficult to avoid this collision in any kind…

There are two caveats to interpreting the video: 1. Human eyes have a better dynamic range than cameras do. The camera is thus displaying a worse contrast than a human eye can see. 2. Motion is preattentive in the visual system. (When you think about it, panicking about subtle motion in the corner of your eye tends to be evolutionarily advantageous). So the time we have to identify the pedestrian using the video is a…

On a collision course, there's zero relative motion of the object you're about to collide with. It's one of the quirks of the human vision system that makes it particularly ill-suited to operating motor vehicles.

Re: Tempe Police Release Video of Uber Accident

#197

Post by jerf » Thu, Mar 22, 2018, 12:23 AM UTC

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…

"The driver specially selected and trained for exactly this job, does not look at the road!"

Humans gonna human. Our brains are trained by millions of years to not pay attention to things not directly important. It's hard enough sometimes to keep our attention on the road when we're doing the driving and the driving is too easy, like being alone on a highway at night. Expecting humans to passively observe driving while making no decisions and taking no actions... well... could you watch eight straight hours of video footage from a car, tomorrow? I know I couldn't. I couldn't even pre-HN/Reddit/instant entertainment constantly in my pocket, let alone now.

The idea that sitting a human in an otherwise-automated car will do anything after the first few hours is absurd. It's a feel-good measure and nothing more. I wouldn't have expected any other result.

Re: Tempe Police Release Video of Uber Accident

#198

Post by bcheung » Thu, Mar 22, 2018, 12:23 AM UTC

Looks like somewhere between 1.0 and 1.2 seconds between when I can see the pedestrian in the video and when the collision occurred.

I don't think a normal human, even if they were paying attention to the road, could have braked or swerved in time to avoid hitting them.

Not sure if visibility is better for a real human than that video but from what I remember about the eye test at the DMV it is extremely lax. They had a vision test sign with a really big font and it was really close. I was appalled that they let people drive who can't see the vision test sign on the opposite side of the room -- even that was ridiculously easy to read.

Also, even if they reacted within 0.1 seconds (unlikely) I doubt the car would have been able to slow down in that timeframe.

I'm surprised IR and LIDAR didn't pick it up though. Would be useful to have the other data released. Video is only a small part of what the car sees and excluding this information is a huge disservice.

Not sure if they already do, but would be nice, no, imperative, for car companies to collect "black box" data of collisions and shared this with all other interested parties. This is a dataset that would be extremely beneficial for all self-driving AI engineers.

Re: Tempe Police Release Video of Uber Accident

#199

Post by jcranmer » Thu, Mar 22, 2018, 12:23 AM UTC

pdonis wrote: post #136

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…

> The gap between street lights (and hence the person) was in the field of view of the camera the entire time But the gap between street lights is going to be very hard to see into. > I'm confident my eyes are good enough that I would have been able to see this person at night in these lighting conditions. I think you're overconfident. Human low light vision is very good if there is low light everywhere. But it is no…

> But the gap between street lights is going to be very hard to see into.

Looking, right now, at a parking lot between two lights from a well-lit room. I can make out most of the outline of the black car in the middle of the "darkness" without any trouble. This isn't even the low light vision kicking in (which I agree isn't going to kick in if you're driving). Human vision should be able to make out the pedestrian earlier than the video footage.

Re: Tempe Police Release Video of Uber Accident

#200

Post by eanzenberg » Thu, Mar 22, 2018, 12:23 AM UTC

pdonis wrote: post #180

eanzenberg wrote: post #164

Although it seems from the dashcam footage that the scene is dark and the bike/pedestrian comes out of nowhere, humans are able to handle extreme low-light situations and I don't doubt that this pedestrian would have been visible to an attentive driver. If autonomous cars cannot handle low-light conditions given the current state of hardware, either from visible, ir, or lidar, then they need to run in the daytime onl…

> humans are able to handle extreme low-light situations But this wasn't an extreme low light situation. It was a situation with street lights alternating with extreme low light regions. Under those conditions the human eye will not be dark adapted and your low light vision won't be very good.

Ive driven at night and the video is not representative of what I see. With your car's headlights and the road lights, there are no blind spots on the road. If there is not enough street illumination and my headlights cannot reach far enough ahead I turn the high beams on and travel cautiously.