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

Tesla issues strongest statement yet blaming driver for deadly crash

abc7news.com

Post reply on HN
Print view View original

661–670 of 850 posts

Re: Tesla issues strongest statement yet blaming driver for deadly crash

#661

Post by ranman »

Animats wrote: post #552

Tesla's system demands more from drivers than manual driving. The driver has to detect automation failures after they occur, but before a crash. This requires faster reaction time than merely avoiding obstacles. Here's a good example - a Tesla on autopilot crashing into a temporary road barrier which required a lane change.[1] This is a view from the dashcam of the vehicle behind the Tesla. At 00:21, things look norm…

Do you own a Tesla? I do. I’ve driven one for 3 years now. It definitely makes road trips easier and reduces fatigue. I believe it requires the same amount of attention but less physical exertion. Your polemic description of “trail of blood and death” is really based on anecdotal descriptions instead of reality. Car accidents happen and will continue to happen for a long time. The goal is to reduce the number of accidents.

Re: Tesla issues strongest statement yet blaming driver for deadly crash

#662

Post by toomanybeersies »

axaxs wrote: post #572

This is so frustrating to me. 50 years ago we had the ability to fly to the moon. A played out trope, sure. But today we can't even do a basic 'dont run into shit' on a perfectly clear day. Physics in this way is and has been a solved problem for decades. The fact that people are dying from hitting stationary objects should be criminal. At lightest, they should bar these companies from even participating.

There aren't a lot of things to run into in space.

Going to the moon and self driving cars are two completely different problem spaces.

Re: Tesla issues strongest statement yet blaming driver for deadly crash

#663

Post by jijojv »

ranman wrote: post #661
Animats wrote: post #552

Tesla's system demands more from drivers than manual driving. The driver has to detect automation failures after they occur, but before a crash. This requires faster reaction time than merely avoiding obstacles. Here's a good example - a Tesla on autopilot crashing into a temporary road barrier which required a lane change.[1] This is a view from the dashcam of the vehicle behind the Tesla. At 00:21, things look norm…

Do you own a Tesla? I do. I’ve driven one for 3 years now. It definitely makes road trips easier and reduces fatigue. I believe it requires the same amount of attention but less physical exertion. Your polemic description of “trail of blood and death” is really based on anecdotal descriptions instead of reality. Car accidents happen and will continue to happen for a long time. The goal is to reduce the number of acci…

I do since Dec'16. Any common car with adaptive cruise control should make road trips easier and reduce fatigue without any false sense of security. You own the older gen AP1 which should be way more stable and apparently doesn't have this particular "barrier" bug.

Re: Tesla issues strongest statement yet blaming driver for deadly crash

#664

Post by blackrock »

If the guy didn't use the Tesla Autopilot, then he would still be alive today.

I don't know how else to clearly say this.

When you are travelling at 65 MPH, then there is no ability to safely react, should the car decide to go berserk.

You probably have less than 0.25 seconds to react. And even if you had your hands on the wheel, there is just no way for you to safely process the situation, and react to safely take control of the vehicle.

And if you do react, then you might end up over-reacting, which would cause a different kind of accident. And which, Tesla would blame on you, for human error for over-reacting.

This is not a good situation. The only way to survive, is to not play.

I have over hundreds of thousands of miles of driving experience, and I as a human, would have never gotten into this situation.

Re: Tesla issues strongest statement yet blaming driver for deadly crash

#665

Post by CaptainZapp »

azernik wrote: post #644
axaxs wrote: post #572

This is so frustrating to me. 50 years ago we had the ability to fly to the moon. A played out trope, sure. But today we can't even do a basic 'dont run into shit' on a perfectly clear day. Physics in this way is and has been a solved problem for decades. The fact that people are dying from hitting stationary objects should be criminal. At lightest, they should bar these companies from even participating.

This isn't "physics in this way" - it's a computation and machine learning problem that still has not been solved with purely visual sensors.

Fine, but then ot's beneath totally irresponsible to let such technology loose on the public.

Re: Tesla issues strongest statement yet blaming driver for deadly crash

#666

Post by sokoloff »

dclowd9901 wrote: post #653
tjomk wrote: post #581

Earlier quoted context omitted.

Or you don't use beta features where the ultimate mistake can cost you your life.

People don’t get this when you call the feature “Autopilot”.

It is comparable to an airplane autopilot, but not to the public misconception of an autopilot. An airplane autopilot is another complex control, not another pilot.

If I command my autopilot to fly me into terrain or into a thunderstorm, it will do so. If I command it to climb at a vertical speed that the engine power can’t support, it will fly the plane into an aerodynamic stall. If I’m flying on autopilot, I still have to pay close attention and still log pilot time as “sole manipulator of the controls”. If I crash on autopilot, it’s still overwhelmingly [human] “pilot error”, not “autopilot error”.

People hear autopilot and incorrectly assume it’s, “Hey George, fly me to LaGuardia and wake me upon landing.”

Re: Tesla issues strongest statement yet blaming driver for deadly crash

#667

Post by petre »

javagram wrote: post #592
lsc wrote: post #586

Earlier quoted context omitted.

>But today we can't even do a basic 'dont run into shit' on a perfectly clear day. so,uh, the emergency braking systems on other cars...so called 'AEB' systems, or what Honda calls Collision Mitigation Braking System... would that have caught this? I would expect this sort of thing "big object in road" to trigger the AEB system... even if it didn't stop me, I'd expect it to trigger before the collision and slow me do…

Honda's documentation for the Civic doesn't indicate it would stop in this case. "Collision Mitigation Braking System™ (CMBS™) Can assist you when there is a possibility of your vehicle colliding with a vehicle or a pedestrian detected in front of yours." I think like Tesla, it would run right into it - the systems have been tuned to reduce false positives and so only detect and stop on objects matching vehicles or p…

The way traditional automakers are doing it is the proper way to do it - cautiously and responsibly. Incrementally implementing driver assistive technology instead of shipping an autopilot which doesn't deliver and crashes the car into obstacles. Traditional automakers are used to screw ups that lead to recalls and lawsuits so they're quite cautious.

The Hyundai Ionic assistive tech slows down the car to a stop if you take your hands off the wheel.

Re: Tesla issues strongest statement yet blaming driver for deadly crash

#669

Post by LoSboccacc »

Matheus28 wrote: post #526
LoSboccacc wrote: post #421

Earlier quoted context omitted.

the car should stop itself to the emergency lane then when user disengagement is detected.

That's a very naive comment. What if there's no emergency lane? Should it just crash towards whatever is on the right side of the car? :) Tesla's autopilot already slows down then stops moving if the user doesn't touch the steering wheel for a period time, but it gives plenty of warning to the user before that happens, as it should be.

So let me see, what are the alternative here.

Inconveniencing the driver and making it a learning experience vs crashing into concrete and killing him

Well. Maybe you are the naive one. Machinery shouldn’t give warnings. We already learned those lesson in industrial automation settings.

The average mentality of “just user error bro” seen on hackers new is what scares me the most. That is not how you build secure machines, far from it, and at this point my only hope not to be killed from an autonomous car in the next 30 years is for the regulations hammer coming crashing down full force on the wanton engineers that are treating loss of life as user error

Re: Tesla issues strongest statement yet blaming driver for deadly crash

#670

Post by cm2187 »

omarforgotpwd wrote: post #593
maxxxxx wrote: post #554

Earlier quoted context omitted.

" Tesla's system demands more from drivers than manual driving. The driver has to detect automation failures after they occur, but before a crash. This requires faster reaction time than merely avoiding obstacles." I have been thinking that too. At what point do you decide that the autopilot is making a mistake and take over? That's an almost impossible task to perform within the available time.

Try it. it’s not impossible. I do it all the time. Have you ever noticed how passengers are still paying attention to what’s going on and sometimes even turn their head to see where the car is turning? it’s like that, except you’re in the drivers seat. it’s not more work than driving.

It’s not a matter of work but of reaction time available.
Post reply on HN