Tempe Police Release Video of Uber Accident • Hacker News Board

Re: Tempe Police Release Video of Uber Accident

#571

Post by jozzas » Thu, Mar 22, 2018, 2:49 AM UTC

linkregister wrote: post #558

In Uber’s own tech blog, it describes its use of LIDAR to generate a wire frame of each object and track it. The lighting should be irrelevant for the vehicle, while it is essential for the driver. From the below article: “Last fall, Uber officials showing off their vehicles on Mill Avenue said their radar and lidar were able to detect objects, including jaywalkers, as far as 100 yards away and avoid collisions.” Som…

I suspect she was picked up by the sensors and categorised as an object. These are absolute fundamentals, and I can't fathom the car working at all if it can't do this. If it was being driven with a limited set of sensors or a malfunctioning sensor, hoo boy the NTSB will have a few things to say.

I reckon it's more likely that this is was due to a software issue where she was flagged as "not a collision concern" or similar due to being in the other lane on a 2-lane road, or due to her path not being predicted accurately due to the odd returns received by a bicycle in profile at that range (it may have thrown off their regular pedestrian detection). In any case a terrible shame.

Re: Tempe Police Release Video of Uber Accident

#572

Post by in_cahoots » Thu, Mar 22, 2018, 2:49 AM UTC

u801e wrote: post #491

Based on the video, it took about 1.5 seconds from when the pedestrian came into view till the vehicle collided with them. At 38 mph, a car is traveling about 56 feet/second. That means that the car traveled about 84 feet in that timeframe. VOL (Visual Aim on Left) headlamps are supposed to be aimed 2.1 inches below headlamp level at a distance of 25 feet [1]. If the headlamps on the vehicle were about 2 feet off the…

We don’t know when the pedestrian car into view for the human eye. It’s possible that the actual scenario was much brighter in real life than on camera.

Re: Tempe Police Release Video of Uber Accident

#573

Post by oska » Thu, Mar 22, 2018, 2:49 AM UTC

Treblemaker wrote: post #32

The 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'…

> If you're choosing to cross a road in the dark, you, too, have a responsibility to understand the drivers' limited visiblity and act accordingly.

You have a self-responsibility, which most people act on (self-preservation). But as a driver you have a responsibility to pedestrians who can be killed by the machine that you are driving. That is the greater responsibility and why active care for pedestrians should be one of the highest regards of a driver (or autonomous driving system).

Re: Tempe Police Release Video of Uber Accident

#574

Post by cozzyd » Thu, Mar 22, 2018, 2:50 AM UTC

drawkbox wrote: post #542

atonse wrote: post #25

How 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…

Arizona and Tempe especially have lots of darker roads. The LIDAR/computer vision tuning at night doesn't seem right. Maybe it was adjusting for the changing street light brightness but yes this is one situation and the vehicle had just come from the Mill bridge that has festive lights that are strung along the bridge [1]. This is a situation though where the LIDAR should have clearly been better than it was. Maybe i…

What the heck are those trails doing in the median if it's not a legal crosswalk?

Re: Tempe Police Release Video of Uber Accident

#575

Post by bigtunacan » Thu, Mar 22, 2018, 2:50 AM UTC

Clearly there are still issues with the car. The sensors should have picked up on this. I would be surprised if the automated emergency braking in my Honda wouldn’t have picked this up based on past experience of a close call at night where it saved my ass.

I want to focus a bit though on the human failure here. Uber has been piloting this program with human backup drivers just for situations like this until there is enough confidence in the maturity and safety of these cars.

Whether this driver could have prevented the collision if their eyes were on the road, we will never know. What is clear is the driver was looking down, most likely on their cellphone.

For that there is negligence on both the part of the driver and on the part of Uber. Uber is recording the interior of these cars, so there should have been some type of review process in place to capture this type of behavior from the drivers so it could be addressed.

Re: Tempe Police Release Video of Uber Accident

#576

Post by jozzas » Thu, Mar 22, 2018, 2:51 AM UTC

smogcutter wrote: post #242

darkerside wrote: post #79

Earlier quoted context omitted.

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…

I dunno, if it decided to ignore the pedestrian because she was in the other lane that's extremely troubling. The pedestrian was moving laterally across the road. If the car has detected that, it should infer that she might become an obstacle very shortly. Driving is all about predicting the future. Think of every time you've been able to tell that someone is going to change lanes even though their blinker is off, or…

When combined with her walking a bicycle, I suspect this is an edge case they never tested. It may have tripped up their pedestrian detection or path prediction (or both) if they got inconsistent hits on the bike.

Re: Tempe Police Release Video of Uber Accident

#577

Post by ScottBurson » Thu, Mar 22, 2018, 2:52 AM UTC

WiseWeasel wrote: post #539

ScottBurson wrote: post #380

Earlier quoted context omitted.

> I hope that the driver isn't the sole focus of punishment As sad as this accident was, I don't see how the driver or Uber is legally at fault. Pedestrians crossing outside of a crosswalk are supposed to yield to traffic. The victim in this video is stunningly oblivious to the fact that she's on a roadway. Obviously we want self-driving tech to be good enough to avoid such an accident, particularly since a human pro…

> Pedestrians crossing outside of a crosswalk are supposed to yield to traffic. Pedestrians always have the right-of-way. They might get a ticket for jaywalking in some cities, but it's never permissible to hit someone with your car, regardless of whether they're in a crosswalk. Vehicular manslaughter is no minor infraction, and Uber can expect a civil suit from the victim's family even if there are no criminal charg…

I'm sure you're right about the civil suit.

Re: Tempe Police Release Video of Uber Accident

#578

Post by jcims » Thu, Mar 22, 2018, 2:52 AM UTC

Animats wrote: post #496

atonse wrote: post #25

How 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…

How did LIDAR and IR (?) not catch that? That seems like a pretty serious problem. Something is badly wrong there. That should have been detected by LIDAR, radar, and vision. Yes, they need a wide dynamic range camera for night driving, but such things exist.[1][2] They're available as low-end dashcams; it's not expensive military night vision technology. Radar should pick up a bicycle at that range. The old Eaton VO…

I wonder if the driver grabbing the wheel disabled the response from the car.

Re: Tempe Police Release Video of Uber Accident

#579

Post by musage » Thu, Mar 22, 2018, 2:54 AM UTC

david-gpu wrote: post #363

bobbles wrote: post #43

Earlier quoted context omitted.

You're getting downvoted but the guy was obviously reading his phone the entire time...

We don't know if he was looking at a display showing the road from the perspective of lidar, etc.

We also don't know if the driver was looking at snuff videos. However, we can make an educated guess that it likely wasn't either.

Re: Tempe Police Release Video of Uber Accident

#580

Post by mtgx » Thu, Mar 22, 2018, 2:54 AM UTC

jerf wrote: post #197

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 whi…

This is why I've always criticized Tesla over its defense of Autopilot accidents, even when most people here were defending them. Then Waymo and Ford came out and confirmed what I and several others here were saying - that humans can't react to "take over the wheel" if something shows up in front of the car.

It's either Level 5, or nothing.