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

971–980 of 1001 posts

Re: Tempe Police Release Video of Uber Accident

#971

Post by singingboyo »

yaps8 wrote: post #882
JohannesH wrote: post #825

In my mind this accident is on Uber no matter how you interpret the video. Scenario 1: Lets say the pedestrian was visible to the naked eye and sensors. The model and safety operator still didn't "see" her and act in time. Who to blame? Uber. Scenario 2: The lighting and environment was in a condition where neither the model nor the safety operator could see the road more than 10 feet in front of the car, yet neither…

This. Many people comment that the lighting is poor and that a human might understand what's happening too late. This is debatable and misses the point: if visibility is bad, you (and it applies with full force to automated drivers) should reduce your speed accordingly, maybe with the exception of freeways where you are not expected to encounter pedestrians.

> maybe with the exception of freeways where you are not expected to encounter pedestrians.

Freeways require you to reduce speed in those cases as well, possibly even more so. This is pretty well demonstrated by the pileups that occur every winter, and similar (if rarer) incidents caused by dense fog or similar. Pileups at freeway speeds are much worse than the 2, maybe 3 car crashes that occur on city streets.

Yes, there's no pedestrians, but one person hitting a pothole or large puddle and spinning out is all it takes to get a pileup if you're not careful.

Re: Tempe Police Release Video of Uber Accident

#972

Post by aecs99 »

jiri wrote: post #811
aecs99 wrote: post #700

I currently work full-time in the self-driving vehicle industry. I am part of a team that builds perception algorithms for autonomous navigation. I have been working exclusively with LiDAR systems for over 1.5 years. Like a lot of folks here, my first question was: "How did the LiDAR not spot this?". I have been extremely interested in this and kept observing images and videos from Uber to understand what could be th…

Is it possible for car to do some calibration of some sort to decide what is current "sensor visibility"? Like a human would do in a fog. Is this a common practice to use this information to reduce or alter speed of the car?

Great question. At least in our algorithms we do this - to adjust the driving speed based on the conditions (e.g., visibility or perception capabilities).

At the end of the day, you can drive only as fast as your perception capabilities. A good example of that is how fast humans can perceive when influenced by drugs/alcohol/medications vs. when uninfluenced.

What is baffling is the fact that the car was driving at 38 mph in a 35 mph zone. This should not happen regardless of how well/poor your sensing/perception capabilities are.

Re: Tempe Police Release Video of Uber Accident

#973

Post by phyller »

tzs wrote: post #688
phyller wrote: post #611

This was not what I was expecting. If this had been an accident with a normal vehicle, I think the driver would have been found to be at least partially at fault, and guilty of involuntary manslaughter. But I'm not a lawyer. I thought the victim jumped out into the street, but that clearly isn't the case. The pedestrian always has the right of way, a human would have seen her and slowed and swerved, and no one should…

> The pedestrian always has the right of way That's a very common misconception, but in fact in the majority of states in the US vehicles have the right of way on roads over pedestrians outside of marked crosswalks and crosswalks at an intersection [1]. I think this misconception arises because many people overestimate what privileges and protections having the right of way confers. Some people seem to think having i…

Yea good points. I didn't mean "right of way" as a legal term, but as a common phrase which means you are responsible as a vehicle driver to not hit people, even if they are doing stupid, illegal things. If you've had a bad day at work you don't get to speed up and swerve into little Johnny as he runs across the street after his ball, then tell his parents "he shouldn't have been in the street" and drive away.

Re: Tempe Police Release Video of Uber Accident

#974

Post by Obi_Juan_Kenobi »

rkangel wrote: post #856
noobermin wrote: post #707

Earlier quoted context omitted.

Alright, so given your observations, which I don't doubt, here's a question I have: why have a pilot on public roads? If uber's software wasn't robust, why "test in production" when production could kill people?

Disclaimer: I am playing Devils Advocate and I don't necessarily subscribe to the following argument, but: Surely it's a question of balancing against the long term benefit from widely adopted autonomous driving? If self driving cars in their current state are at least close to as safe as human drivers, then you could argue that a short term small increase in casualty rate to help development rate is a reasonable cos…

It's an n of 1, but we're nowhere close to 'human driver' levels of safe.

Humans get 1 death per 100 million miles.

Waymo/Uber/Cruise have If you consider just Uber, they have somewhere between 2 and 3 million miles, suggesting a 40x more deadly rate. I think it's fair to consider them separately as my intuition is that the other systems are much better, but this may be terribly misguided.

This is a huge deal.

I honestly never thought we'd see such an abject failure of such systems on such an easy task. I knew there would be edge cases and growing pains, but 'pedestrian crossing the empty road ahead' should be the very first thing these systems are capable of identifying. The bare minimum.

This crash is going to result in regulation, and that's going to slow development, but it's still going to be justified.

Re: Tempe Police Release Video of Uber Accident

#975

Post by rootusrootus »

smileysteve wrote: post #647
rootusrootus wrote: post #533

Regardless of how suddenly she appears in the camera, it is pretty logical that she was probably visible to the human driver if he had been watching the road. That's not nearly fast enough on a stretch of road like that to be overdriving the visibility provided by the modern headlights of a brand new Volvo. I bet she was visible to the driver in plenty of time to have stopped. I'm an optimist, but still I think we ar…

I agree; and particularly like that you mention this is a brand new Volvo -- I regularly drive much faster on roads that are likely to have deer -- and this driver is clearly not paying attention. I think the driver should be considered for manslaughter charges (to which, as they were an employee, Uber is civilly liable for)

I am not totally onboard with the idea of manslaughter charges, but in principle I'm with you. I am nervous about how the liability question gets worked out. When a human behind the wheel is directly liable for their actions we at least have some incentive for good behavior. When we detach that liability and assign it generically to a company instead, then it feels like our options for coercive behavior shaping has been significantly reduced. We can't put the corporation in jail, and history shows that we rarely punish corporations significantly, even large fines are frequently pretty small when you look at the budget of the corporation involved.

Re: Tempe Police Release Video of Uber Accident

#976

Post by aecs99 »

whiskyant wrote: post #796
aecs99 wrote: post #700

I currently work full-time in the self-driving vehicle industry. I am part of a team that builds perception algorithms for autonomous navigation. I have been working exclusively with LiDAR systems for over 1.5 years. Like a lot of folks here, my first question was: "How did the LiDAR not spot this?". I have been extremely interested in this and kept observing images and videos from Uber to understand what could be th…

Maybe the question isn't why the LIDAR didn't spot it. I feel it's more likely it did spot it, but couldn't make the correct decision.

You summed up all my speculations in one sentence

Re: Tempe Police Release Video of Uber Accident

#978

Post by Balgair »

abiox wrote: post #606
Balgair wrote: post #424

Earlier quoted context omitted.

> Wow, that self driving car is total crap if it can't pick up that obstacle. Not to nit-pick too much but 'that obstacle' is a person. A person that died because engineering hubris. I think we should be careful in our language in this situation, as it may get away from us. She was not 'an obstacle' or a 'thing to avoid' or 'hobo', she was a living person.

i'm not sure i follow what you're trying to communicate. what do you mean by "get away from us"? also, framing things as "obstacle" vs "living person" seems like a false dichotomy.

I mean that Uber, being a very shady company, has a large incentive to 'sterilize' the language. The person becomes an 'object', the person driving the car becomes a 'rouge employee', the cruddy software and cost cutting becomes an 'undiscovered malfunction', the burden of safety on the AV manufacturer becomes a 'shared breakdown'. Responsibility and humanity start to get stripped away via the word choices used.

Re: Tempe Police Release Video of Uber Accident

#979

Post by Balgair »

brokenmachine wrote: post #608
Balgair wrote: post #424

Earlier quoted context omitted.

> Wow, that self driving car is total crap if it can't pick up that obstacle. Not to nit-pick too much but 'that obstacle' is a person. A person that died because engineering hubris. I think we should be careful in our language in this situation, as it may get away from us. She was not 'an obstacle' or a 'thing to avoid' or 'hobo', she was a living person.

I 100% agree, however arguing about semantics is not productive. I deliberately chose the word "obstacle" to hammer home the point that the car did not even need to identify that the object in front of it was a person. If that obstacle had happened to be a fallen tree, it would be the car's passengers that would be dead instead. Literally all it needed to do was stop, or at the very least slow down. In this case, I b…

I agree, the car is not even close to prime-time

Re: Tempe Police Release Video of Uber Accident

#980

Post by badthingfactory »

kalleboo wrote: post #907
baking wrote: post #584

This was the original story: "After the Uber collision, the car continued traveling at 38 miles per hour, according to the Tempe police chief" https://www.bloomberg.com/news/articles/2018-03-21/for-self-... In other words, the car _never_ detected the pedistrian and never slowed down on its own. This has nothing to do with _when_ it saw her. It clearly didn't, yet both camera and LIDAR should have been able to.

This doesn't even make sense, we can see in the video that the "driver" does look up and react. Why didn't they hit the brakes?

At least one study been done to see how responsive a passive "driver" of an autonomous vehicle can be. The study concluded a person that is required to intervene with an autonomous vehicle is no better than a drunk person.

It makes sense. If the car is able to get you from point A to point B 99.9% of the time without an issue, are you really going to be paying attention that one time when a person walks out in front of you and the car fails to respond appropriately? If the vehicle enables people to not pay attention, they aren't going to be attentive.

I'd be willing to bet almost every driver that gets to ride around in these test vehicles thinks the job is awesome because they get paid without having to really do anything.

Post reply on HN