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

Cyclist hit by driverless Waymo car in San Francisco, police say

sfgate.com

Post reply on HN
Print view View original

661–670 of 750 posts

Re: Cyclist hit by driverless Waymo car in San Francisco, police say

#661

Post by shadowgovt »

gnicholas wrote: post #648
shadowgovt wrote: post #380

Earlier quoted context omitted.

Broadly speaking (and this is a very broad brush, details may differ), autonomous vehicles will resolve collision avoidance by doing something legal. Swerving is not legal and it's not legal because you might hit something you can't see. On the other hand, an autonomous vehicle maintains perpetual 360° situational awareness so it already knows whether it's safe to swerve.

> Swerving is not legal and it's not legal because you might hit something you can't see. If I'm driving down the road and someone pushes a kid in front of my car, I'm very likely to swerve, even if I'm not 100% sure if there is someone next to me. If I hit a car next to me and cause damage/injure people, there is a good chance that a jury would find in my favor because a "reasonable person" (the relevant legal stand…

AVs aren't doing near enough calculus to make a decision like that; they have obstacles and they avoid them. Their avoidance can allow for lane departure (that 360 situational awareness is handy), but they aren't going to solve a trolley problem in realtime if boxed in.

The thing is... Neither are people. A kid darts out on front of you and you won't have time to make a conscious decision. The maneuver you make will be too fast for rational thought, and your brain will explain to itself why it did that later.

... Incidentally, this is one of the things that scares me about human drivers. Put a human on the stand to testify and most will say they never saw that kid, no matter what physical process happened to their eyes and neurons. I think most even actually believe it.

In any case, I'm glad that in this incident they'll be able to do a full NHTSA workup, pull logs, and reprogram the machine to avoid an accident like this in the future. Can't do that to a human driver.

Re: Cyclist hit by driverless Waymo car in San Francisco, police say

#662

Post by Symbiote »

ghaff wrote: post #603
School-Cotton wrote: post #598

Earlier quoted context omitted.

At least in New York you could make this work if your little town is on a metro north line. The last trains do leave after 1 AM I believe.

NYC is not exactly typical although I'm sure there are other one-off examples per the comments, especially if you relax the constraints to 11pm or so. But I'm not sure it's terribly common most places to take a late train to a small town and get home from there without driving from the train station. I think my last commuter rail is around 10:30pm but I'd have little confidence that I could get a taxi/Uber from the t…

It's common surrounding larger cities in many countries. The trains from about 23.00 to about 01.00 in Copenhagen are busy with people going home to the suburbs and beyond.

Picking some random village I've never been to, from Copenhagen to Uggeløse (population around 2000) takes 65 minutes, the last connection to the bus requires leaving Copenhagen at 23:16.

Picking some other village that's on a railway line (Veksø, population 1800), the trains run all night every 30 minutes on Friday and Saturday nights. On other days the last train is at 00:30-ish.

The original claim was from a city to a town, which is easier than either of these examples.

Re: Cyclist hit by driverless Waymo car in San Francisco, police say

#663

Post by CobrastanJorji »

_dain_ wrote: post #492
flakes wrote: post #350

Earlier quoted context omitted.

I don't agree with most laws that impose absolutes. The courts exist to deal with grey areas. A large number of pedestrians struck by cars each year (at least in North America), are themselves intoxicated, and not fit to walk along roadways. There are many cases where the driver should not reasonably be at fault for a pedestrians erratic behavior. There's also the "hit-to-kill" phenomenon in China, where it can be fi…

>A large number of pedestrians struck by cars each year (at least in North America), are themselves intoxicated, and not fit to walk along roadways. There are many cases where the driver should not reasonably be at fault for a pedestrians erratic behavior. Excuse me? How can someone be unfit to just walk somewhere just because they're drunk? You don't need a license to walk!

You don't need a license to go downhill skiing, but you still shouldn't do it while intoxicated.

But yes, if it is unsafe to walk home while drunk, that's an infrastructure and safety problem more than it is a failing of the drunk person.

Re: Cyclist hit by driverless Waymo car in San Francisco, police say

#664

Post by Symbiote »

cloverich wrote: post #623
fmobus wrote: post #608

Earlier quoted context omitted.

It is more efficient though. But if you say you want to avoid at all costs, and assuming there are more people like you out there, maybe we should just start _charging that cost_.

What do you mean by efficient? When I lived in Austin, we lived in a neighborhood where the (only) train picked up, and it dropped me off 5 blocks from my office downtown. It was an extraordinarily pleasant period of my life. But even then, in ideal conditions, it still took longer than driving, and I had to time my departures to avoid waiting 15-30 minutes at the station. I'm definitely in favor of re-designing our…

They might mean energy efficiency or time efficiency (reading a book etc on the train). Or maybe construction energy efficiency, which in Austin is presumably a sunk cost.

Note a 15-30 minute wait is what I have to endure at night (21.00 to 01.00) if I want to travel to/from the suburbs of Copenhagen on a weekday, similar population to Austin. On Friday and Saturday nights each line runs every 20 minutes all night.

Re: Cyclist hit by driverless Waymo car in San Francisco, police say

#665

Post by lex-lightning »

petesergeant wrote: post #215
lex-lightning wrote: post #167

Earlier quoted context omitted.

That’s such a low bar though. Zero is achievable. Part of it is to slow cars down and take more cars off the road rather than changing who drives. Invest in public transit, redesign problematic infrastructure, and create mixed use urban areas so people don’t feel the need to travel through a city. A city should be a safe place. More deaths in a slower zone than on freeways is preposterous. I understand the reasons wh…

> A city should be a safe place Is that a fundamental attribute of a city?

I’ll grant that this was a statement of a value I personally hold, an ethical axiom which I can’t convince you of. Fair.

Would you care to present a your own views? Or to dispute the reasoning that follows from my stated values?

If not, then we aren’t going to agree. I’m very much in favor of streets not being designed like the video game Frogger. Listen, I like snowboarding; adrenaline is a blast. But I don’t know why getting groceries has to be an adrenaline rush.

I’m making a joke. I’m coming into this debate with established opinions. But even if you’re not going to win me over, the floor is yours to persuade future readers that speed and carrying capacity are more important than safety. Or to argue that I’m presenting a false dichotomy.

What I don’t find acceptable is accepting the status quo just because it’s the status quo.

Re: Cyclist hit by driverless Waymo car in San Francisco, police say

#666

Post by lex-lightning »

illirik wrote: post #368
petesergeant wrote: post #215

Earlier quoted context omitted.

> A city should be a safe place Is that a fundamental attribute of a city?

Of course; that's why our ancestors first built walls around granaries.

I’ll go deeper still. It’s why we’re social animals. Trust is the foundation of secure relationships and secure relationships keep us safe. Cities wouldn’t exist if it were every man for themself.

Yeah, for sure I believe that cities, walls, friendships, food chains, economies, armies, language… this all rests on meeting the basic human needs of shelter (security from elements and hostile actors), food (security from starvation), community (security of a support network).

Re: Cyclist hit by driverless Waymo car in San Francisco, police say

#667

Post by mint2 »

ra7 wrote: post #166
fprog wrote: post #144

Earlier quoted context omitted.

Public trust is a pretty good reason.

There’s enough to establish public trust with their voluntary release of safety data and regulators closely overseeing their operations. Setting a precedent for releasing footage means fault is automatically assumed when they don’t voluntarily release it. It’s a problem if they’re not even releasing it to the regulators, but that’s not happening here.

Sounds like the same argument bad cops claim after questionable police shootings

Public trust requires building it first. As one of the first instances of a waymo crash, yes the public needs to see it. If after reading the footage in the first 99 crashes and in each time waymo’s assessment was valid, that’s when Waymo has public trust and can credibly not release every single video but only do it on a case by case basis.

Re: Cyclist hit by driverless Waymo car in San Francisco, police say

#668

Post by tbrownaw »

alemanek wrote: post #545
aubanel wrote: post #475

Earlier quoted context omitted.

I think the end goal of self driving cars is simply to let people do something else than drive. Safety is only a condition for this to happen.

There are other benefits of self driving cars. For example average utilization of a self driving car would be at least an order of magnitude greater than non-self driving one. Think about how much time a car spends parked and waiting to be used. With greater utilization we should have a lower absolute number of vehicles that are replaced more frequently. This would lead to less polluting and better maintained vehicle…

What fraction of the day is rush hour?

How many cars are needed for that?

How many cars are needed for typical evening or night traffic?

Re: Cyclist hit by driverless Waymo car in San Francisco, police say

#669

Post by mint2 »

PeterisP wrote: post #644
ra7 wrote: post #590

Earlier quoted context omitted.

Unless you’re willing to watch millions of hours video of every Waymo vehicle driving and judge them in aggregate, watching a one-off incident will tell you nothing about their safety. At that point, you might as well trust their aggregate statistics reported to the regulators because they capture everything. If you’re not an expert, it’s best to let the regulators and insurance adjusters do their job.

Your argument is much less effective at making me trust Waymo than the accident video could be.

Esp if one considers it’s the same argument cops make when trying to suppress incriminating police shooting footage.

Re: Cyclist hit by driverless Waymo car in San Francisco, police say

#670

Post by mint2 »

belltaco wrote: post #134
hx8 wrote: post #57

Earlier quoted context omitted.

If the footage is not released to the public I'm going to assume it makes Waymo look bad.

They did say that the Waymo only braked after the cyclist was 'fully visible'. They won't be releasing the footage on their own.

Yeah that was an interesting weasel word. It raised a lot of questions.
Post reply on HN