Live data from Hacker News

Tesla hacker reveals what driver-facing camera is looking for

electrek.co

311–320 of 455 posts

Re: Tesla hacker reveals what driver-facing camera is looking for

#311

Earlier quoted context omitted.

> I don't want autopilot so I can actually sleep at the wheel, I want it in case I do. You should never get behind the wheel if there's a chance you'll fall asleep! You shouldn't be driving. Period. This is incredibly dangerous and puts innocent people at risk. No technology should compensate for this. We don't design technology to help drivers under the influence feel better about their driving. Please don't get beh…

> You shouldn't be driving. Period. what a stunningly pretentious and ignorant thing to say. reasons to be tired and drive: rotating shiftwork, working multiple jobs, caring for a newborn, caring for the sick. that covers millions of people, off the top of my head, without doing any research. and those are only the most black-and-white examples. more ambiguous cases include: driving from physical activity like OP men…

> what a stunningly pretentious and ignorant thing to say.

You might want to brush up on the HN guidelines. Talking like that to people is never OK here, nowhere near it.

Re: Tesla hacker reveals what driver-facing camera is looking for

#312

Earlier quoted context omitted.

The fact that this comment is a plethora of words trying to win an argument how falling asleep at the wheel shouldn't be considered unsafe is incredible.

That's an ungenerous reading of the comment. They would agree that falling asleep at the wheel is unsafe. We've designed a system, however, that is set up such that it's bound to happen sometimes. Not everyone can call in to work and tell their boss "sorry, I slept badly last night, I'm not coming in" or order a $60 Uber to get somewhere on a whim. Moralizing at people is just virtue signaling.

I agree that moralizing doesn't help, and that we should fully take into account "realities". But that also means you can't just assume autopilot will help you - it could very well be that having autopilot will end up making people to fall asleep at the wheel more often, and therefore increase the chance of accidents or severity of the accidents. We know for a fact that it does involve some trade off (see those people misusing autopilot and getting into accidents), and we simply don't know autopilot provides better safety overall.

There has been enough indirect evidence to suggest advanced driver assist without a significant driver monitoring is dangerous (Google's report of their employees doing other things, the fatal accident of Uber's self-driving vehicle, various fatal and non-fatal accidents of Tesla autopilot due to misuse), but there's been no systematic study and we don't have enough public data to suggest it's one way or another.

Re: Tesla hacker reveals what driver-facing camera is looking for

#313

Earlier quoted context omitted.

The thing is, to be susceptible to that kind of effect, a normal adult -- meaning one without an underlying medical condition -- would already have to be very short of good quality sleep and/or doing foolish things like driving for several hours at night without taking a break. It is not normal to be susceptible to just dozing off at the wheel without realising like that. If you are in that position, you should be aw…

These are pretty absolute terms you're talking in with little corresponding evidence. Monotonous activities frequently make people drowsy. This is not some medical condition, it is a common occurrence. Driving is often an acutely monotonous activity. I know a huge number of people that get drowsy while reading books. Do they all suffer from the same "condition"?

We weren't talking about merely being drowsy. We were talking about having a pattern of falling asleep at the wheel, and knowingly continuing to drive while unfit for that reason. It's right there in the top comment that started the thread.

Re: Tesla hacker reveals what driver-facing camera is looking for

#314

Earlier quoted context omitted.

Some people don't have that privilege. Ever known someone that works two jobs trying to make ends meet? You don't have a choice except to drive home from work.

Ever known someone that works two jobs trying to make ends meet? You don't have a choice except to drive home from work. And will it be easier to make those ends meet when you are dead, or paralysed, or locked up because you left someone else dead or paralysed? The arguments that a few people are making here, as if not driving while so tired you might fall asleep would cause some devastating loss to society and heral…

>And will it be easier to make those ends meet when you are dead, or paralysed, or locked up because you left someone else dead or paralysed?

Oftentimes for people forced to work 2, 3, 4 or more jobs just to survive, these are also acceptable "solutions" to the problem. The risk analysis comes down to [low chance of dying now driving impaired] versus [medium chance of dying later, homeless/starving after losing job(s)].

I once arrived at work with blood all over my face, neck, and hands from biting my lips to stay awake.

Autopilot is so far, far, far out of reach for most people working multiple jobs that it's really a non-issue not worth debating for a long time, but the technology behind it would be a literal lifesaver if it were more widely available to the unfortunately large population of people that are forced to drive regardless of mental state in order to survive and/or provide for their families and/or other circumstances "out of their control".

Re: Tesla hacker reveals what driver-facing camera is looking for

#315

Earlier quoted context omitted.

This is a matter of degree. The start of this thread wasn't about someone who was driving home at the end of a long day at work and feeling a little tired. It was about someone who, by their own admission, had a history of dozing off at the wheel, and wanted to knowingly continue driving while unfit to do so. If we're being brutally objective here, road accidents are extremely expensive in purely economic terms. Obvi…

The start of the thread was actually: > You should never get behind the wheel if there's a chance you'll fall asleep! I would say anyone that feels drowsy has a change of falling asleep. So 50% of the population should regularly not get behind the wheel, by GC's standard.

No, the start of the thread was:

One of the reasons I would love to have autopilot is because I have a history of dozing off at the wheel. Never so much where I've had an accident, but bad enough where I've scared myself badly.

This was followed by arguing for the safety feature in case it happened again.

Re: Tesla hacker reveals what driver-facing camera is looking for

#316

Earlier quoted context omitted.

That's not how real life works. It's dangerous for everyone to drive and in the future manual driving will be as legal as riding horses on highways is today. People do potentially deadly risk reward calculus all the time. And of all potentially deadly things that people regularly do to crack down on, this one would have a pretty huge negative ROI for society. This issue impacts a non-trivial percentage of the populat…

The fact that this comment is a plethora of words trying to win an argument how falling asleep at the wheel shouldn't be considered unsafe is incredible.

Nobody said it was safe. They merely pointed out that some people have to choose between safety and feeding their families.

Re: Tesla hacker reveals what driver-facing camera is looking for

#317

Earlier quoted context omitted.

You've never had a bad night's sleep? Never had to make a longish unexpected trip because of a family emergency? Never had to work 2 jobs to make ends meet? I've had to do all of those things. At some times in my life, I've had the first and last for an extended period. To my knowledge, I have never as a result driven in a condition that made me unsafe behind the wheel, though. Yours is another comment that talks abo…

If you have done those things, then you've accepted some increased risk of falling asleep, just like most of us have at one time or another. We're human and can't say for certain that falling asleep is impossible (or some other medical event won't happen). We live with that risk, and some others have to live with more risk because of their circumstances. I'm more reacting to all the posts that no one should ever driv…

If you have done those things, then you've accepted some increased risk of falling asleep, just like most of us have at one time or another.

Comments like this are side-tracking the discussion. The original premise wasn't that sometimes everyone is a bit more alert or tired than others. It was someone who knew very well that they had a history of falling asleep at the wheel and wanted to continue driving apparently knowing it was likely to happen again.

We're human and can't say for certain that falling asleep is impossible (or some other medical event won't happen).

Of course not. But normal people with a healthy lifestyle don't just randomly fall asleep at the wheel in the absence of a serious (but perhaps undiagnosed) medical condition. The way a lot of the comments in this thread are defending driving while tired, you'd think millions of people could all just doze off without any warning or awareness that they were at unusually high risk.

You know they let some people drive who have extremely bad eyesight--bad enough that that they can't read road signs?

In my country, such poor vision would (if not properly corrected) immediately disqualify you from holding a driving licence. This is a policy that I entirely agree with, so we apparently come from very different perspectives here.

Re: Tesla hacker reveals what driver-facing camera is looking for

#318

Earlier quoted context omitted.

The problem is, that like drunk driving, the person making the call doesn't always have the experience to know when they are impaired, or how close they are to falling asleep. I've driven the family from Northern Califorina to Sourthern California quite a few times, many of them overnight. It took one instance of actually falling asleep at the wheel for a few seconds to realize "oh, that's the difference between tire…

The problem is, that like drunk driving, the person making the call doesn't always have the experience to know when they are impaired, or how close they are to falling asleep. Then they can follow the same rule as for overtaking: if in doubt, don't. It's not complicated. There is no way you are tired enough to be actually falling asleep without realising you're far too tired to be driving safely. It's not as if it's…

> There is no way you are tired enough to be actually falling asleep without realising you're far too tired to be driving safely.

Except I was, and I did. And now I can recognize the signs that weren't obvious before, so I don't anymore. You're saying there's no way, and I'm giving you a concrete example of an instance.

> It's not as if it's a close thing where one minute you're alert and driving responsibly and the next minute you're out.

There's a huge range between "alert" and "asleep". I'm tired after a long day, yet I can drive responsibly. Even a the point where I could lay down and easily go to sleep if I chose to, I'm fine to drive for quite a while after that (preferred sleep time is not necessarily the same as responsible driving time, there are many factors that go into it).

> Anyone who can't make that determination reliably is demonstrably unfit to be in charge of a vehicle and should hand in their licence.

So, have you been in that situation? Or just always stopped before you got to that point? Or not even driven over a long enough period to be in that situation? You haven't even presented any anecdotal evidence, you've just made assertions that what I said I personally experienced can't be how I described it (and implied I was demonstrably unfit to drive at the point I said this happened to me).

Re: Tesla hacker reveals what driver-facing camera is looking for

#319

Earlier quoted context omitted.

These are pretty absolute terms you're talking in with little corresponding evidence. Monotonous activities frequently make people drowsy. This is not some medical condition, it is a common occurrence. Driving is often an acutely monotonous activity. I know a huge number of people that get drowsy while reading books. Do they all suffer from the same "condition"?

We weren't talking about merely being drowsy. We were talking about having a pattern of falling asleep at the wheel, and knowingly continuing to drive while unfit for that reason. It's right there in the top comment that started the thread.

Who said there was a pattern and people continuing to drive after it happened once?

Re: Tesla hacker reveals what driver-facing camera is looking for

#320

Earlier quoted context omitted.

> You shouldn't be driving. Period. what a stunningly pretentious and ignorant thing to say. reasons to be tired and drive: rotating shiftwork, working multiple jobs, caring for a newborn, caring for the sick. that covers millions of people, off the top of my head, without doing any research. and those are only the most black-and-white examples. more ambiguous cases include: driving from physical activity like OP men…

The 'you shouldn't be driving' commenters are people who are are either living in a privileged high-income bubble and don't understand what real life is like for most of the population, or are logical purists thinking about this issue in isolation without considering what real life is like for most of the population. Neither of those demographics has considered that their 'you shouldn't be driving, period' policy wou…

Exactly how have you established what the commenters have or haven't considered? To me, its silly to make this about class or economic status. Also, this is not just a made up "policy" . Its the law (well, in most countries). If you want to relax the laws so that drowsy/reckless driving is accepted, you are welcome to present an actual argument.
Post reply on HN