Idiot Leaves Driver's Seat In Self-Driving Infiniti, On the Highway 406
cartechboy writes Self-driving cars are coming, that's nothing new. People are somewhat nervous about this technology, and that's also not news. But it appears self-driving cars are already here, and one idiot was dumb enough to climb out of the driver's seat while his car cruised down the highway. The car in question is a new Infiniti Q50, which has Active Lane Control and adaptive cruise control. Both of which essentially turn the Q50 into an autonomous vehicle while at highway speeds. While impressive, taking yourself out of a position where you can quickly and safely regain control of the car if needed is simply dumb. After watching the video, it's abundantly clear why people should be nervous about autonomous vehicles. It's not the cars and tech we need to worry about, it's idiots like this guy.
Submission with a spelling error, say it isn't so. (Score:0, Insightful)
"self-driving cars are already hear,"
Uh, do people not know to read back their posts before they submit them, or did they let their smartphone autocorrect?
Self-driving cars will never succeed. WE DON'T HAVE VERY MANY AUTOMATIC TRAINS IN THE WORLD, the reason is because they have to contend with stupid drivers and jaywalkers. What makes anyone think automatic cars will work with human drivers still on the road?
Huh? (Score:5, Insightful)
After watching the video, it's abundantly clear why people should be nervous about autonomous vehicles.
No, it's clear why we should be worried about almost-but-not-really autonomous vehicles, in the real deal this would be fine. If we could get this guy as far away from a steering wheel as possible, it'd be perfect.
It's not autonomous (Score:5, Insightful)
It's partially autonomous. And that's why it is so dangerous.
After watching the video, it's abundantly clear why people should be nervous about autonomous vehicles. It's not the cars and tech we need to worry about, it's idiots like this guy.
Once we actually have autonomous vehicles --- this won't be an issue as a human operator won't even be required for safe operation; only to provide instructions about where to go.
Re:Huh? (Score:5, Insightful)
Absolutely. Anything that *almost* removes the need for you to be behind the wheel is an accident waiting to happen. Even if you remain in your seat, what are the odds that you'll remain alert and aware of the surrounding traffic after the 100th commute where it proved completely unnecessary to do so?
Either give me a car that will let me take a nap while it drives, or leave me in control. I've got better things to do than babysit a computer
No, this is proof we need autonomous vehicles (Score:3, Insightful)
No, this is proof we need autonomous vehicles. Consider the current alternative that these "idiots" are driving.
Re:Submission with a spelling error, say it isn't (Score:5, Insightful)
I would guess that the number of stupid jaywalkers will diminish quickly
Re:Huh? (Score:4, Insightful)
Absolutely. Anything that *almost* removes the need for you to be behind the wheel is an accident waiting to happen. Even if you remain in your seat, what are the odds that you'll remain alert and aware of the surrounding traffic after the 100th commute where it proved completely unnecessary to do so?
This is why autonomous cars are a long while away, sure we'll be 99% of the way there by 2018, it's that last 1% that's the bitch.
We cant remove the human from the link until we are 100% certain that the computer can make decisions better than a very good driver.
A few weeks ago a friend of mine was in a rear end collision with one of those new Ford Kuga SUV's which advertise automatic emergency brakes... They dont work, even if they did engage they wouldn't have stopped in time because their based on a dry stopping distance, not a wet one (it's winter here in Oz). We cant even get simple systems to work reliably.
Right. This is the "deadly valley" (Score:5, Insightful)
No, it's clear why we should be worried about almost-but-not-really autonomous vehicles, in the real deal this would be fine.
That's right. Automatic lane keeping plus radar-based cruise control is right in the middle of the "deadly valley" - good enough to allow hands-off driving 98% of the time, not good enough to handle trouble. This is why that Cruise startup building a budget self-driving system [slashdot.org] worries me. Thos guys are from "social" apps. They're thinking they can ship something that sort of works, and that's good enough. It isn't.
Auto manufacturers are held to a much higher standard than the computer industry is used to. GM is being sued because their ignition switches could turn off if people hung too much crap on their keychain. (Something unlikely to be caught in testing, because, at the test track, each key hangs on a separate key tag.) "Speeding, cellphone texting, intoxication... irrelevant. We are not looking at the driver, or the circumstances of the driver's negligence. We are looking at the automobile, and only the automobile." - terms of the GM settlement. [gmauthority.com]
The minimum safe level of performance for a self-driving car is that the vehicle must be able to bring itself to a safe stop, preferably at the side of the road, in any emerging bad situation. Even after any single-point failure.
Few computer based consumer products meet that standard, but a some do. The Segway is a good example. There's enough redundancy in a Segway to keep single failures from face-planting the user. Five rate gyros instead of three, two batteries, two processors, and a safety shutdown mode that brings the vehicle to a stop and sounds alarms to tell you to get off before it fails.
Re:Huh? (Score:5, Insightful)
Nonsense. The computer only needs to be markedly better than an *average* driver to be a huge safety win. It doesn't even need to *always* be better than the average driver - if it can reliably avoid 90% of the most common accidents, then even it it fails spectacularly in the last 10% of edge cases, and even if humans would have avoided 100% of those, the autonomous systems will still have reduced the number of accidents by a factor of 9.
That said, I wouldn't trust the current auto manufacturers to do the job properly, they mostly can't even install a media system without also potentially letting anyone with a bluetooth rifle take complete control of the critical electronics. But there are folks doing some really impressive driving systems - I've even seen one that can drive professional road-racing courses flawlessly, with near-professional lap times. And yes, it can even do so in the rain.
Re:Huh? (Score:5, Insightful)
In the olden days, the horse knew the way home.
Re:Huh? (Score:4, Insightful)
Nope,
Firstly people will never accept that. It needs to be far superior in order to be accepted by the general public.
Secondly, the average driver in the US is considered a very poor driver by Australian standards... and the Germans/English are better than us.
Thirdly, idiots tend to make different mistakes, computers will always make the same ones. For this reason, autonomous systems need to be far superior to the average driver as 1 in 1000 events on the road will catch maybe 1% of average drivers (because they fail in different ways) but if it gets 1 autonomous car, it will get others as well (because they all fail in the same way). For this reason, it needs to be able to avoid more risks than the average driver. Ironically, the unpredictability of bad drivers helps protect them and make them better than autonomous systems even thought their unpredictability is the biggest thing making them a bad driver.
It's getting that last 1% right that will take the most time.
That said, I wouldn't trust the current auto manufacturers to do the job properly, they mostly can't even install a media system without also potentially letting anyone with a bluetooth rifle take complete control of the critical electronics.
Back to my example about the Ford SUV.
Ford programmed it for dry conditions, it does not perform as well in the wet (sometimes it does not even work at all, as my friends Prelude found out the hard way). That system doesn't take into account the differing environmental conditions, it doesn't know if the road is smooth or coarse, wet or dry, tarmac or gravel. All of these factors make massive differences in braking distance.
Re:Submission with a spelling error, say it isn't (Score:4, Insightful)
In the end, the need to get home from the bar w/o a D.U.I. will drive the technology. ,in a self driven chariot.
Perhaps this was just a drunk, in question. We'll call him a philosopher scientist, in a flash of dead brilliance, he reasons the cops can't give him a D.U.I. if he isn't driving. Others round the world, resonating sympathetically, will do the same. THEN, the self driving GOOGLE map using Smart Cars will take your drunk ass through a bad neighborhood before inevitably breaking down. Oh, of course there will be nice stories, like pregnant lady gets to the E.R. to spew quints
Strange stories, guy tells car to go to Best Buy, car takes him to Best Buy in a distant city. Auto Erotic stories, Exhibitionist couples lacking the task of driving, screw in the back seat, Daredevils on the hood.
I don't think safety issues will stop anyone, anymore than obvious ones with explosives, alcohol and indiscriminate unprotected sex. It doesn't matter if they don't work , at first, it makes good news copy. These are humans we are talking about , right?
Re:Huh? (Score:4, Insightful)
People tend to think of average as mean, so it's entirely possible (and likely I think) that most drivers are better than the mean.
That's because there are a bunch of decent drivers, and a few who are really atrocious and shouldn't be driving. I'm guessing the distribution looks something like:
2% Excellent: not only do they not cause accidents, they avoid them and make everybody safer with their driving habits
88% Good: not always 100% defensive so susceptible to things like sudden braking, but generally good drivers if nothing unexpected happens
10% Bad: weaving through traffic, speeding in residential areas, turning or backing up without looking, etc
So 90% of the population is "above average" (mean).
Re:Right. This is the "deadly valley" (Score:5, Insightful)
They did test it, identified a problem early, and then the engineer responsible suppressed the results to avoid having the mistake next to his name.
There were already engineering standards regarding heavy keychains, and related testing procedures. That is why they fired 15 people, instead of calling it a learning opportunity. The only opportunity for learning is learning not to be dishonest. That is also why there is a criminal probe.
The only reason the public found out and the recall happened was because of a wrongful death lawsuit where the lawyer learned about it through discovery. This is one of those cases where the lawyers were actually more honest (or more risk-averse) than the engineers!