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

NHTSA’s full investigation into Tesla’s Autopilot shows 40% crash rate reduction

techcrunch.com

Post reply on HN
Print view View original

221–230 of 327 posts

Re: NHTSA’s full investigation into Tesla’s Autopilot shows 40% crash rate reduction

#221

Post by maxander »

mjevans wrote: post #212
uncoder0 wrote: post #211

After looking at the report it looks like Tesla ran into the same issue we did in the 2007 DARPA Urban Challenge. The trailer was higher than the front facing sensors. We and most other teams had all assumed 'Ground Based Obstacles' meant that any obstacles on the test track would make contact with the ground in the lane of travel. DARPA decided to put a railroad bar across the street and expected cars to back up and…

Predictions like that only work out if the lesson is sufficiently broadcast. Clearly, since this is still newsworthy here, that is not the case. (However this back channel is helping.)

If there isn't some compilation or review article along the lines of "all serious failure modes encountered in autonomous vehicles since 2005-ish," there should be.

Re: NHTSA’s full investigation into Tesla’s Autopilot shows 40% crash rate reduction

#222

Post by secabeen »

trevyn wrote: post #94
mikeash wrote: post #80

Earlier quoted context omitted.

It doesn't require instant attention, no. But it's hard envision a scenario where the car needs attention at some point in the future, not particularly soon, but can't drive itself. Maybe I'm drawing the line at the wrong point, though. If a car can drive itself but may occasionally need to pull over and ask the driver to tell it which way to go, at the driver's leisure, is that level 3 or level 4? I thought 4, but I…

Level 4: "Such a design anticipates that the driver will provide destination or navigation input, but is not expected to be available for control at any time during the trip. This includes both occupied and unoccupied vehicles." (emphasis added) In my mind, if a level 4 vehicle is unable to reach a given valid destination without assistance, that would be considered a failure. In reality, macro-scale rerouting around…

>In reality, macro-scale rerouting around obstacles that the self-driving system does not know how to overcome is not a particular challenge, GPS systems do this easily. The definition doesn't say it has to take the most efficient route. ;)

Oh, that's going to go over really well. Say you're drunk, and want to go home in your self-driving car. However, there's a blind left turn that your car knows it can't safely navigate, so it happily routes you a 15-minute drive through back roads for a trip that should take 5. People are going to love that.

(I've heard that UPS does route planning with a system that discourages left turns, even sometimes doing three rights to avoid some dangerous lefts.)

Re: NHTSA’s full investigation into Tesla’s Autopilot shows 40% crash rate reduction

#223

Post by jayjay71 »

semi-extrinsic wrote: post #180
celticninja wrote: post #146

Earlier quoted context omitted.

What events do you consider rare? Car crashes?

Presumably they meant airbag deployment crashes in Teslas. We can estimate the absolute number from Tesla's press release in December stating that 1.3 billion miles have been driven on Autopilot, while the NHTSA report states the accidents they count occur 0.8 times per million miles. So in that case it's data from about 1000 crashes in the Autopilot case, and more in the non-Autopilot case. Which tells us that the s…

So they're comparing miles driven manually before a certain date to miles driven manually and automatically after a certain date? It certainly implies that Autopilot is safer, but why not separate it properly (Autopilot vs manual) and give us far more accurate information?

It just seems...odd.

Re: NHTSA’s full investigation into Tesla’s Autopilot shows 40% crash rate reduction

#224

Post by secabeen »

huangc10 wrote: post #28

Can anyone who is in the industry comment on how Autopilot performs in poor weather (ie. flash floods, thunderstorms, snowstorms etc...) All I can find from the article about weather was in section 3.1: > The manual includes several additional warnings related to system limitations, use near pedestrians and cyclists, and use on winding roads with sharp curves or with slippery surfaces or poor weather conditions. The…

It depends. The Tesla system is reliant on good contrast between the road stripes and the road surface. With grey skies and rain, that contrast can go down to a level where the lines cannot be reliably discerned, and AutoSteer is not available.

Re: NHTSA’s full investigation into Tesla’s Autopilot shows 40% crash rate reduction

#225

Post by greglindahl »

bluGill wrote: post #210
celticninja wrote: post #146

Earlier quoted context omitted.

What events do you consider rare? Car crashes?

Actually you can argue that car crashes are rare. 1.2 deaths per 100 million miles is rare. (I'm not able to find statistics for accidents without deaths, but we can extrapolate) A lot of people drive a lot of miles which is why there are a lot of crashes. So you can argue either side of this depending on how you want to twist the numbers.

Auto crashes are the leading cause of death in some age ranges. You have to decide what you want to normalize to... other kinds of death is a good choice.

Re: NHTSA’s full investigation into Tesla’s Autopilot shows 40% crash rate reduction

#226

Post by oaktowner »

duskwuff wrote: post #137
JBReefer wrote: post #83

Earlier quoted context omitted.

The left lane is only for passing, and if you are passing >80% of the cars on the road, you are a reckless driver.

> if you are passing >80% of the cars on the road, you are a reckless driver. Or you are driving at the speed limit on a 2x2 highway used as a trucking corridor. (For example: most of I-5.)

Right. In California trucks have a lower speed limit than cars, which results in what feel like pretty dangerous situations.

The right lane is pretty much for trucks going 55.

The left lane is a combination of: * Trucks going 60-65 to pass a truck in the right lane * 'Typical' cars who are going 75ish * Fast cars who are going 80-90

So there is a huge disparity of speed in that lane -- and that disparity seems like it must be a huge factor in many highway accidents.

It always feels really dangerous. I wonder if it is safer than just having the trucks have the same speed limit as the cars.

Re: NHTSA’s full investigation into Tesla’s Autopilot shows 40% crash rate reduction

#227

Post by yohui »

greglindahl wrote: post #220
yohui wrote: post #113

Earlier quoted context omitted.

For those unaware, "AEB" stands for "Autonomous Emergency Braking". See: https://en.wikipedia.org/wiki/Collision_avoidance_system Did Tesla cars have AEB prior to Autopilot installation? If not, then this suggests the 40% reduction in crashes may simply be due to the installation of an AEB system. What effect Autopilot's other features may have would remain uncertain.

Yes, Tesla cars had AEB since March 2015. Autopilot (with autosteer) came out in October 2015. Presumably Tesla's AEB gets improved a little with every firmware release.

Thanks. But the NHTSA report looks at data starting with "MY 2014" (i.e. calendar year 2013) cars.

If AEB did lead to a 40% reduction in crash rates for Tesla cars, as it did for other car models, I suspect that moving the dividing line by five months later wouldn't change the figures much: you would still have a lot more crashes prior to AEB and fewer after.

Re: NHTSA’s full investigation into Tesla’s Autopilot shows 40% crash rate reduction

#228

Post by mikeash »

bluGill wrote: post #209
mikeash wrote: post #196

Earlier quoted context omitted.

If the car can detect something in the adjacent lane, then it'll give you a warning if you try to change lanes into it manually too.

My car has that - there are too many false positives for it to be useful. I most commonly hear get the warning when there are two left turn lanes and I'm in the right one (because I need to turn right at the next road). Sure there is a car to my left, if it doesn't turn left I'll hit it, but the other car would be at fault in that case. I will admit that there have been a couple times it saw the car I was about to me…

Mine has never had a false positive. I've yet to accidentally merge into anybody with it, so I don't know how it behaves in the situation it's supposed to handle.

Re: NHTSA’s full investigation into Tesla’s Autopilot shows 40% crash rate reduction

#229

Post by NamTaf »

1209120931 wrote: post #126
DonHopkins wrote: post #119

Earlier quoted context omitted.

What kind of "CARTCHA" challenges and games might be fun to play, to prove to your car that you were paying attention to the road and traffic, yet not distract you from being ready to take over driving? Computer vision and speech recognition could be useful for proving the driver's awake, as well as driving the car itself. https://en.wikipedia.org/wiki/Car_numberplate_game

The challenge of driving the car manually, without any hyped autopilot.

This. There's nothing else because any other task requires diverting mental focus away from the act of constantly assessing the changing environment around your car. That by definition leads to slower reaction times.

Re: NHTSA’s full investigation into Tesla’s Autopilot shows 40% crash rate reduction

#230

Post by mikeash »

secabeen wrote: post #222
trevyn wrote: post #94

Earlier quoted context omitted.

Level 4: "Such a design anticipates that the driver will provide destination or navigation input, but is not expected to be available for control at any time during the trip. This includes both occupied and unoccupied vehicles." (emphasis added) In my mind, if a level 4 vehicle is unable to reach a given valid destination without assistance, that would be considered a failure. In reality, macro-scale rerouting around…

>In reality, macro-scale rerouting around obstacles that the self-driving system does not know how to overcome is not a particular challenge, GPS systems do this easily. The definition doesn't say it has to take the most efficient route. ;) Oh, that's going to go over really well. Say you're drunk, and want to go home in your self-driving car. However, there's a blind left turn that your car knows it can't safely nav…

I've heard the same thing about UPS, but the way I heard it, it's only because left turns can be so much slower. A route that leans more heavily on right turns will usually be faster overall.
Post reply on HN