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

Studies find automatic braking can cut crashes over 40%

apnews.com

Post reply on HN
Print view View original

131–140 of 177 posts

Re: Studies find automatic braking can cut crashes over 40%

#131

Post by branon »

pwinnski wrote: post #126
branon wrote: post #10

Oof, my experience with with auto-brake in a 2021 Subaru Crosstrek (anecdotal, sample size of one, all that good stuff) is much the opposite. It's too sensitive and tends to brake shockingly hard during fairly mundane maneuvers. When backing up near a wall/obstacle, the auto-brake seems to kick in when anything is closer than three or four feet away from the bumper. I use my mirrors and have an accurate impression of…

"Over-sensitive" or not, what's the actual harm in driving in a way that avoids triggering it? When your driving safety system gives you feedback while driving, you have two choices: - Learn from the system and modify your driving to be safer for yourself and those around you, or - Ignore the system, describe it as "too sensitive," and disable it. Before the driving safety system feedback, you might have been an unsa…

This is fair. I do (and have) adjusted my behavior to avoid setting off the auto-brake. Mainly I did this because disabling it entirely seemed like a bad idea, despite my misgivings.

Sometimes I still trigger it when reversing towards obstacles though. It still prompts an eyeroll and a muttered curse. Not sure if that will ever go away, somebody else mentioned that it tends to treat bushes as solid objects. Not that I'm reversing over top of the bushes, but reversing _towards_ them should not be as big of a problem as the auto-brake would have the driver believe.

Re: Studies find automatic braking can cut crashes over 40%

#132

Post by branon »

pwinnski wrote: post #129
branon wrote: post #26

Earlier quoted context omitted.

Fair enough, could definitely happen. I figure I _do_ still leave enough space even when closing though. I've never wrecked a car yet, so I tend to doubt the machine when it challenges a what I'm pretty sure is a well-refined heuristic.

Everybody who ever wrecks a car has "never wrecked a car yet" right up to that point.

Totally fair. If (when!) the day comes, hopefully it will have been for some other reason instead of a maneuver I incorrectly thought was safe. I trust myself enough to think this will be the case.

Re: Studies find automatic braking can cut crashes over 40%

#133

Post by r00fus »

giantg2 wrote: post #61
branon wrote: post #10

Oof, my experience with with auto-brake in a 2021 Subaru Crosstrek (anecdotal, sample size of one, all that good stuff) is much the opposite. It's too sensitive and tends to brake shockingly hard during fairly mundane maneuvers. When backing up near a wall/obstacle, the auto-brake seems to kick in when anything is closer than three or four feet away from the bumper. I use my mirrors and have an accurate impression of…

"I'm only one car-length away and still going 40 MPH. Another half-second and they'll be gone," These numbers must be wrong, and are terribly unsafe. Just off the top of my head, that's like travelling 60fps for .5 seconds is 30 feet, which exceeds 1 car length, even for my truck.

Anyone discussing braking margin/safety at speed in car-lengths ... is measuring the wrong thing.

Your reactions don't correlate in distance (car-lengths), they're measured in time response (seconds).

Always measure your safety buffer in the appropriate metric: time.

Re: Studies find automatic braking can cut crashes over 40%

#134

Post by pwinnski »

branon wrote: post #101
giantg2 wrote: post #92

Earlier quoted context omitted.

Please provide a source for that definition. I've never seen that used anywhere.

I don't have a source, it's just what makes sense to me - which at the end of the day, is how we all drive. As far as definitions for "car length" go, I feel that "average length of an actual car" is just as open to interpretation as "minimum safe length at which to follow a car" - really, these are both non-definitions. Today I've learned NEVER to reference "car length" online because it's a shitty and imprecise uni…

Subaru is wrong for thinking you closing on a barely-moving car at speed is dangerous, and the entire internet is wrong for thinking "car length" refers to the length of a car. Two doesn't quite make a pattern, but it suggests one. You being obviously-wrong on the second point doesn't mean you're definitely wrong on the first, but I'd be more inclined to trust the automotive company with access to crash statistics than your "never wrecked a car yet" instincts, personally.

Then again, I've known two people who both proudly proclaimed that they'd never even gotten a ticket before whose driving was so terrifyingly unsafe I wouldn't get into a car with them a second time, so maybe that's my bias.

Re: Studies find automatic braking can cut crashes over 40%

#135

Post by frumper »

yonaguska wrote: post #105
berberous wrote: post #49

Earlier quoted context omitted.

Have you ever looked up stopping distances? https://www.automotive-fleet.com/driver-care/239402/driver-c... I don’t know how accurate the above is, but it suggest that the stopping distance at 40mph is 5-9 car lengths, so I have to agree with the above poster that if you are routinely leaving one car length and your car system is freaking out, you are probably driving unsafely. Keep in my that statistically this coul…

It's effectively impossible to maintain safe stopping distances in many places. Speaking from experience in Chicago, it's just a concept that doesn't exist for many people. And if you are driving slower in order to maintain a safe stopping distance, you're creating an extra hazard on the road by maintaining a larger speed differential from the rest of traffic. Not to mention, you increase your own risk of accidents b…

In my experience they will aggressively cut you off anyway, so you might as well give yourself some breathing room.

Re: Studies find automatic braking can cut crashes over 40%

#136

Post by scruple »

linuxftw wrote: post #117
yonaguska wrote: post #105

Earlier quoted context omitted.

It's effectively impossible to maintain safe stopping distances in many places. Speaking from experience in Chicago, it's just a concept that doesn't exist for many people. And if you are driving slower in order to maintain a safe stopping distance, you're creating an extra hazard on the road by maintaining a larger speed differential from the rest of traffic. Not to mention, you increase your own risk of accidents b…

Little known fact: You can travel the exact speed of the car in front of you while also being any distance N behind them. If I'm traveling greater than 50MPH, I leave at least 3 car lengths between me and them, this is to allow for reaction time to stop at the same rate they are stopping.

In California that just means that in the span of 15 seconds 3 other cars are going to move in front of you and the car that was ahead of you.

Re: Studies find automatic braking can cut crashes over 40%

#137

Post by foobazgt »

branon wrote: post #101
giantg2 wrote: post #92

Earlier quoted context omitted.

Please provide a source for that definition. I've never seen that used anywhere.

I don't have a source, it's just what makes sense to me - which at the end of the day, is how we all drive. As far as definitions for "car length" go, I feel that "average length of an actual car" is just as open to interpretation as "minimum safe length at which to follow a car" - really, these are both non-definitions. Today I've learned NEVER to reference "car length" online because it's a shitty and imprecise uni…

Safe follow distance is usually measured in time, not distance. It's a function of human reaction time plus your vehicle's ability to brake, which is mostly a function of its velocity (also greatly affected by tires and weight but there's far less variance there in passenger cars compared to personal variance in velocity).

Most DOTs claim around 1.5s for reaction time, and then up to another 1.5s for braking at typical highway cruising speeds. So, 2-3s follow distance in total. Adjust down for low velocity and up for poor driving conditions.

Re: Studies find automatic braking can cut crashes over 40%

#138

Post by notacoward »

brookst wrote: post #64
dazc wrote: post #51

Earlier quoted context omitted.

> 'Sometimes, though, I will allow the distance between myself and the next car to close, for example, if they're turning and will be gone in a second...' I often wonder what advantage in life people think they are gaining by this behaviour. Care to share?

Reducing the odds of getting rear-ended by absorbing the slack of a conservative following distance rather than braking to maintain the same following distance from 35 to 20 mph.

You can't control what the person behind you does, but you can control what you do and rushing people through turns is unsafe. If the person in front of you signaled (or did anything else to indicate) their turn in sufficient time, then you should be able to do likewise for the person behind you with a light tap on the brakes. The odds of being rear-ended probably won't change by any significant amount, and IMX people who cite those odds weren't really thinking about them at the time. It's just an excuse for prioritizing speed over safety.

Re: Studies find automatic braking can cut crashes over 40%

#139

Post by garyrichardson »

I have a 2018 VW Atlas and its automated braking system has done the right thing 99.9% of the time. It's definitely saved me from 5-6 accidents -- being cut off a number of times, a racoon jumping in front of my car, and that one time my attention waned.

My Tesla, on the other hand, get's it right about 10% of the time. I wanted to type a higher percentage, but phantom breaking happens so often. The number of times I've seen the red X on the atlas in the entire time owning it is about the number of times the M3 loses its mind on a monthly basis.

The other interesting thing: I could probably write the algorithm for the adaptive cruise control on the Atlas. It's so predictable that I know exactly what it's going to do in reaction to traffic around me. If a car ahead of me slows down, I slow down at a constant rate. If the car speeds up, I know exactly how much it's going to speed up. Same with people changing lanes ahead of me, etc.

The M3 OTOH feels like a buggy machine learning model. I have no idea why it's at the current speed it's at. If I set it to 120KM/H and there's no car in front of me the chances of it being 120KM/H is about 30%. If a car pulls away from and my car should speed up there is no rhyme or reason to when and how quickly it speeds up.

All that being said, I'd still take the Tesla automated braking over nothing at this point. Reading the other posts, it sounds like VW is doing something right.

Re: Studies find automatic braking can cut crashes over 40%

#140

Post by giantg2 »

VBprogrammer wrote: post #77
berberous wrote: post #49

Earlier quoted context omitted.

Have you ever looked up stopping distances? https://www.automotive-fleet.com/driver-care/239402/driver-c... I don’t know how accurate the above is, but it suggest that the stopping distance at 40mph is 5-9 car lengths, so I have to agree with the above poster that if you are routinely leaving one car length and your car system is freaking out, you are probably driving unsafely. Keep in my that statistically this coul…

> And then another driver t-boned her because he didn’t have enough distance to stop, also cleary partially at fault IMO. Not partially at fault; completely and totally at fault (except if we're talking about someone deliberately trying to cause a crash in which case I'll settle for partially). The reality is there could be a hundred reasons the car stopped, the driver being an idiot is only one of them. Mechanical f…

It sounds like the example was not of a car they were following, but one that pulled out infront of them and stopped in their lane. Just as you are referencing possible excuses for that driver, we could come up with possible excuses for the driver that pulled out (like pulling out suddenly without enough distance to react even if the other car was going under the speed limit).

We also need to understand that being at fault isn't 1-to-1 with who was being safe. Typically, if you pull out and get t-boned by someone who has the right of way and is generally following the law, the person being t-boned is at fault for failing to yield. Where a vehicle is impacted plays a huge role in investigators determining who is at fault. Now if you run into the back of the person in front of you, it's almost always your fault, unless the person in front was driving carelessly or recklessly (brake check, for example) but would still involve partial fault in many cases.

So really, we can't say either way without knowing a lot more.

Post reply on HN