Federal regulators says car makers “cannot wait for perfect” on automation

“Pre-market approval steps” could clarify responsibility in the future.


Read the whole story

 

[Infinity4011]

I think it's worth noting that cars didn't start out as safe as they are today. My first and second vehicles did not have airbags, or ABS. My current vehicle doesn't have traction control or cruise control. Every iteration of autonomous vehicle that gets put on the road will be better than the previous, but no car design is perfect. And the most imperfect part of any car is the driver.

 

[Statistical]

Steveha7[/url]":w754emue]Is there really a demand for a driverless car? Driving is the fun part and is why I got a license. This may help handicapped people and if kids can ride in them without an adult. Other than that, I don't see much demand.

Nothing fun about a commute or I would say any fun wears thin after the first 1000 times. I would gladly pay $20K extra for a self driving car today even if it was limited to only self driving on the highway (user took over at on/off ramps).

 

[MisterGrumps]

Is there really a demand for a driverless car? Driving is the fun part and is why I got a license. This may help handicapped people and if kids can ride in them without an adult. Other than that, I don't see much demand.

I'm guessing you have a sub 45 minute, no rush hour bumper to bumper traffic commute?

 

[infernallexicon]

Steveha7[/url]":11cufg8z]Is there really a demand for a driverless car? Driving is the fun part and is why I got a license. This may help handicapped people and if kids can ride in them without an adult. Other than that, I don't see much demand.

I hope this is sarcasm, otherwise you're incredibly short-sighted. People with long commutes would love to not have to waste time driving to their jobs, losing productivity. They could be working or reading or doing something else productive, especially in traffic jams.

People that are drunk and beyond reach of transit options would love driverless cars.

People that are on family trips would probably love to not have to pay attention to the road. Play games with their kids, etc.

I enjoy driving but I used to have to drive to the same stupid places and I hated it

 

[whoisit]

Steveha7[/url]":1uk69se9]Is there really a demand for a driverless car? Driving is the fun part and is why I got a license. This may help handicapped people and if kids can ride in them without an adult. Other than that, I don't see much demand.

Driving can be fun, if you are on an open road, and there's no traffic. But, sitting in stop and go for 2 hours just to make a 10 mile trip is mind numbing. Add to that drivers, who are angry, tailgating, distracted, changing lanes without signaling, or chemically impaired, and it's down right dangerous. I really think automated cars would help out a lot in congested cities, especially since they can communicate to each other, and use GPS and route around traffic slowdowns and accidents*.

*Routing to other main arterials/roads designed for higher traffic volumes. Not residential streets.

 

[bugsbony]

srs[/url]":11gsy09h]Not being able to see the road (and obstructions) due to the sun is far from "not perfect".

Leaving aside the sensor failure:

1) Why didn't the car stop when there was a sensor failure?

and

2) Why aren't there redundant systems using different technologies acting as backups?

This accident was on "autopilot", badly named, but it is not autonomous, the driver is not supposed to look away from the road or to move his hands away from the wheel.

 

[Boskone]

Infinity4011[/url]":2tjjxqf2]I think it's worth noting that cars didn't start out as safe as they are today. My first and second vehicles did not have airbags, or ABS. My current vehicle doesn't have traction control or cruise control. Every iteration of autonomous vehicle that gets put on the road will be better than the previous, but no car design is perfect. And the most imperfect part of any car is the driver.

Agreed. I think the question is less "Is Tesla's autopilot safe?" than "Is Tesla's autopilot safer than a currently-normal car?".

As a Texan with a decent commute and fairly regular multihour highway drives, I think the answer's probably "Yes, when the driver's not being an idiot.". I see a lot of instances of people zoning out and drifting lanes, or sliding speed (up or down) and then correcting when they close on someone or another driver passes them; I'd like to see a lane-following/cruise-control system that does something like play a loud ping to alert the driver when there's a "driver-induced incident".

That said, I'm uncomfortable as yet with the idea of full automation. I do think it's something we really should develop slowly and carefully. Like, start with lane-following (and cruise-control), add in things like auto-(un)parking, then maybe "parking lot summoning" (it'd be spectacular at some airport park-n-rides), then maybe segue to highway automation, and so on. Come up with a clear development plan, but test each stage before releasing it to the public, then monitoring public use for a year or so before the next stage.

 

[Jim Z]

Steveha7[/url]":3mkrpg6m]Is there really a demand for a driverless car? Driving is the fun part and is why I got a license. This may help handicapped people and if kids can ride in them without an adult. Other than that, I don't see much demand.

#1 the group of people who actually enjoy driving (i.e. enthusiasts) is a small percentage of the car-owning public (and they think they're way bigger and more important than they are.) the average person buys a car because they "need" one, not because they like driving.

#2 I'm an enthusiast (I even work in the industry) but I have to admit that the bulk of the driving I do (commuting) is shitty and I hate it. It's the kind of driving that makes even me hate a manual transmission at times. I would rather let my car deal with the traffic on the interstate then for me to sit there hating people left and right.

(I'd actually rather move to commuter rail, and save the driving for non-commuting time. hell, given where people tend to live and work in the Detroit area, you could run rail along all of the major interstates and state highways in the tri-county area and stand a good chance of connecting a hell of a lot of us from home to work, and v.v.)

 

[Boskone]

Statistical[/url]":wsawmh8p]

Steveha7[/url]":wsawmh8p]Is there really a demand for a driverless car? Driving is the fun part and is why I got a license. This may help handicapped people and if kids can ride in them without an adult. Other than that, I don't see much demand.

Nothing fun about a commute or I would say any fun wears thin after the first 1000 times. I would gladly pay $20K extra for a self driving car today even if it was limited to only self driving on the highway (user took over at on/off ramps).

I wouldn't go $20k extra for an autonomous car, but semiautonomy and a $30k pricetag is why the Tesla 3 is tempting.

I can't readily afford a second car, though, so I won't buy one. :|

 

[Alfonse]

JButler[/url]":ewmucdq5]

Rosekind said automation would “save people’s lives”

Again, is there real evidence proving/projecting this? Or are we just assuming things will work out? I'm talking about Level 3 and above where drivers are not expected to pay constant attention to cover for system limitations and bugs. I also highly doubt Tesla's Autopilot in current form reduces accidents.

To be honest, I think true automation driving systems of today are probably much safer than the current half-assed "automatic-enough-that-people-stop-paying-attention" that you find in cars now.

 

[SixDegrees]

Boskone[/url]":3k70h1is]

Statistical[/url]":3k70h1is]

Steveha7[/url]":3k70h1is]Is there really a demand for a driverless car? Driving is the fun part and is why I got a license. This may help handicapped people and if kids can ride in them without an adult. Other than that, I don't see much demand.

Nothing fun about a commute or I would say any fun wears thin after the first 1000 times. I would gladly pay $20K extra for a self driving car today even if it was limited to only self driving on the highway (user took over at on/off ramps).

I wouldn't go $20k extra for an autonomous car, but semiautonomy and a $30k pricetag is why the Tesla 3 is tempting.

I can't readily afford a second car, though, so I won't buy one. :|

Agree. We'll be waiting decades for complete autonomy - despite Mr. Musk's overbrimming optimism - but partial autonomy is already here, and I'll happily take more of it.

Some yutz came within an inch of broadsiding me earlier today because he was so intent on his phone he wasn't paying attention to where he was going - which was straight into my stopped car. I would love for a system that jammed on the brakes at 100% grip instantly and hard enough to trigger his airbags and jam his phone deep into his face. Idiot bastard.

Although I would very much love to be chauffeured everywhere by a completely autonomous car, I'll settle for incremental approach to that goal.

 

[SixDegrees]

JButler[/url]":120aj2t7]

Rosekind said automation would “save people’s lives”

Again, is there real evidence proving/projecting this? Or are we just assuming things will work out? I'm talking about Level 3 and above where drivers are not expected to pay constant attention to cover for system limitations and bugs. I also highly doubt Tesla's Autopilot in current form reduces accidents.

When traction control was introduced in German cars by Mercedes, accidents for those cars immediately fell by 15% compared with other vehicles. The government wisely made such systems mandatory. Similar improvements were noted years earlier for ABS - also, by the way, introduced in Germany, by Mercedes. It's probably too early to tell how automated braking systems work, but we should get preliminary results soon, and there's no reason to think there will be anything but another improvement over cars unequipped with such systems.

There's really no doubt at all that such automation reduces accidents, or that reducing accidents saves lives and reduces injuries. Comparisons for specific technologies are either already available, or will be very soon. There is no reason to think they will be anything but an improvement.

 

[foxyshadis]

Wow, I have to say that I'm shocked at pleased that NHTSA would take this position. There have been times in the past when they've been adamantly against the implementation of new technology (although SAE tends to be the bigger problem), and the support and weight of government regulations behind it will hopefully improve funding and accelerate availability on the road.

Of course they're going to look into every accident for a while to see how it can be improved, but they know better than anyone else how much pain and death human drivers cause, because they're the ones who put out those numbers.

Someday I'm sure we'll have the fight over whether humans can be allowed to drive at all, but maybe it'll mean licensing more like the UK than the current "a few hours in a class then around the block twice" driving instruction the US gives.

 

[Ooks]

When car autonomy reaches the point that I can call for a ride rather than have to spend 40k + insurance and maintenance to own one, I'm all in.

Especially if lawmakers get forward thinking enough to convert HOV lanes to autonomous lanes. Imagine a string of cars following each other at incredibly short distances at incredibly high speeds. Would certainly solve a lot of the current issues with highway capacity.

 

[SixDegrees]

Dietz[/url]":15belkc3]So, are these same regulators going to shield car companies from civil liability when imperfect automation causes driver fatalities?

Yep, didn't think so.

Uh, why should they? If a vehicle can be shown to be defective right now, its manufacturer will be liable for injuries that result from such defects. Why is driver automation special in any way? Manufacturers will be found at fault just as readily as if their brakes failed due to improper design or manufacturing.

See also "Takata".

 

[Hot Jupiter]

in a time when auto fatalities have been up 8 percent since 2014.

Completely out of context...

To date?

2014-2015?
2014-2015.5?
What are the most recent numbers?

Is that up 8% in just the number of deaths or the number of deaths per mile driven?

Isn't context important here?

 

[Derecho Imminent]

Has any insurance corp committed to insuring these cars?

 

[SixDegrees]

adamrussell[/url]":mexqhq1t]Has any insurance corp committed to insuring these cars?

Well, they insure every car with automated systems already on the road, from ABS to traction control through blind spot monitoring and alerts to backup cameras and monitoring to...

 

[SixDegrees]

Hot Jupiter[/url]"g5fp4mp]

Boskone[/url]"g5fp4mp]

Statistical[/url]"g5fp4mp]

Steveha7[/url]"g5fp4mp]Is there really a demand for a driverless car? Driving is the fun part and is why I got a license. This may help handicapped people and if kids can ride in them without an adult. Other than that, I don't see much demand.

Nothing fun about a commute or I would say any fun wears thin after the first 1000 times. I would gladly pay $20K extra for a self driving car today even if it was limited to only self driving on the highway (user took over at on/off ramps).

I wouldn't go $20k extra for an autonomous car, but semiautonomy and a $30k pricetag is why the Tesla 3 is tempting.

I can't readily afford a second car, though, so I won't buy one. :|

Why would anyone buy a fully autonomous car?!?

They would be in a fleet and dispatched as needed.

Could be. Then again, why would anyone buy a non-autonomous car, using the same reasoning? Probably because despite its basic features, cars vary tremendously in other amenities, and also serve very different purposes.

For my purposes, a car-on-demand would probably work out just fine - as long as usage costs were similar to ownership - and were far lower than what it currently takes to use a cab, which is functionally equivalent. For many, though, such a system would probably not be optimal.

In any case, that's a long-term end game, at best. Such a result is many, many decades away at the earliest.

 

[Boskone]

edwould[/url]":38yxcqko]“autonomous doesn't mean perfect,” but that “we need industry to take the safety aspects of this very seriously.”

No dumbo, your the regulator, we need YOU to take the safety aspects of this very seriously because it is YOU we will sue when we discover injury and death from automakers who PUSH

If the manufacturers didn't push, eventually someone would sue due to the accidents resulting from the lack of automation.

Don't be a dumbass by letting the perfect be the enemy of the good. A lot of stuff we do is "good enough" because it's still vastly better than the option. I mean, if you get a hernia the repairs leave a weak spot prone to further hernias. Should we just leave hernias untreated because there's not a perfect solution yet? No, because that would just be stupid.

The same holds true for automation. We shouldn't wait until it's 100% perfect (because it never will be; reality's an annoyingly complicated place), but rather deploy automation features as they become superior to leaving the features out.

 

[JButler]

Alfonse[/url]":1hc4py8e]

JButler[/url]":1hc4py8e]

Rosekind said automation would “save people’s lives”

Again, is there real evidence proving/projecting this? Or are we just assuming things will work out? I'm talking about Level 3 and above where drivers are not expected to pay constant attention to cover for system limitations and bugs. I also highly doubt Tesla's Autopilot in current form reduces accidents.

To be honest, I think true automation driving systems of today are probably much safer than the current half-assed "automatic-enough-that-people-stop-paying-attention" that you find in cars now.

What makes you think we even have this "true automation" systems today? I'm not talking about autonomous vehicles for "easy" situations like parking lots, factory floors, or a few suburban blocks on a clear day. I'm talking everyday pubic road driving conditions.

 

[valkyriebiker]

Steveha7[/url]":13r42sih]Is there really a demand for a driverless car? Driving is the fun part and is why I got a license. This may help handicapped people and if kids can ride in them without an adult. Other than that, I don't see much demand.

The advantages to driverless cars are nearly limitless. The *only* negative is whatever fun one may lose.

And if we ever reach a mandatory post-driver society, then an entirely new set of advantages come into play, such as: higher speeds, far less following distance, some reduction in lane width (e.g. more lanes), elimination of traffic signals, elimination of virtually all onsite parking including at your home in favor of fewer mega lots, renting your car out for taxi-duty if you wanted (Tesla is already thinking along those lines), auto-sharing clubs.

There's bound to be advantages I'm not even thinking of. Most of this is probably beyond my lifetime, sadly.

 

[SixDegrees]

Hot Jupiter[/url]":1rzyeee8]

adamrussell[/url]":1rzyeee8]Has any insurance corp committed to insuring these cars?

Why would a 'perfect' car need insurance?

The whole premise is crazy...

A perfect car... ...bah

Trains run on rails and they wreck all the time.

Well, not all the time. In fact, they wreck very, very, very much less often than cars do.

 

[Boskone]

Hot Jupiter[/url]":2tap3lix]

Boskone[/url]":2tap3lix]

Statistical[/url]":2tap3lix]

Steveha7[/url]":2tap3lix]Is there really a demand for a driverless car? Driving is the fun part and is why I got a license. This may help handicapped people and if kids can ride in them without an adult. Other than that, I don't see much demand.

Nothing fun about a commute or I would say any fun wears thin after the first 1000 times. I would gladly pay $20K extra for a self driving car today even if it was limited to only self driving on the highway (user took over at on/off ramps).

I wouldn't go $20k extra for an autonomous car, but semiautonomy and a $30k pricetag is why the Tesla 3 is tempting.

I can't readily afford a second car, though, so I won't buy one. :|

Why would anyone buy a fully autonomous car?!?

They would be in a fleet and dispatched as needed.

Why would I want a fleet car shared with random joes? Rentals are bad enough, and they get cleaned after every use. A "just in time fleet" would have cars going directly from one user to the next, complete with puddles of vomit or whatever other nastiness.

I'd rather have my own vehicle who's condition I can guarantee.

 

[pr0t0]

I've been thinking about autonomous driving lately. I think it's going to start with semi trucks ('lorries' for my friends across the Atlantic). I'm envisioning that the right lane on highways will be designated as the 'autonomous' lane, and trucks and passenger cars will be free to use that lane without driver intervention.

You know who's probably paying very close attention to this? The airline industry. Imagine if you could hop in your car at 10pm, go to sleep, and be at your destination when you wake up. You and several passengers, and your luggage, will be able to travel (over land, anyway) pretty comfortably:
1. without the need to get a taxi/uber to the airport or pay to park your car long-term
2. without the need to stand in long security line
3. without the need to have every orifice of your body inspected because you have 4oz of shampoo instead of 3oz
4. without any restrictions of what you can take with you
5. without paying hundreds to thousands (say a family of four). It all comes as a bonus to purchasing an autonomous vehicle

Assuming destinations less than 8-10 hours away by car; I'd travel a LOT more if I could do this.

 

[Jim Z]

JButler[/url]":2naiiyyc]

Rosekind said automation would “save people’s lives”

Again, is there real evidence proving/projecting this? Or are we just assuming things will work out? I'm talking about Level 3 and above where drivers are not expected to pay constant attention to cover for system limitations and bugs. I also highly doubt Tesla's Autopilot in current form reduces accidents.

it's going to rely on vehicle-to-vehicle (V2V) communication, so your car "knows" what surrounding cars are doing, and are going to do.

but even then, an autonomous car doesn't drive drunk, tired, isn't texting friends or taking selfies, isn't all pissed off and driving like an asshole, isn't some 16-year-old asshole who thinks the roads are race tracks, etc.

 

[Alfonse]

pr0t0[/url]":1706ne5u]You know who's probably paying very close attention to this? The airline industry. Imagine if you could hop in your car at 10pm, go to sleep, and be at your destination when you wake up. You and several passengers, and your luggage, will be able to travel (over land, anyway) pretty comfortably:

People have said a lot of things about sleeping in cars. "Comfortable" is probably not one of them.

Granted, neither is an airplane trip.

Also, you're going to have to wake up a couple of times to fill it up with gas.

That being said, I could see companies buying a few autonomous cars that have more legroom and so forth, so that they could send executives relatively short distances overnight without having to pay airfare. Though they'd still need a quick hotel stay to shower and so forth before their meetings.

 

[Ooks]

Hot Jupiter[/url]":2kgp91q9]

Boskone[/url]":2kgp91q9]

Statistical[/url]":2kgp91q9]

Steveha7[/url]":2kgp91q9]Is there really a demand for a driverless car? Driving is the fun part and is why I got a license. This may help handicapped people and if kids can ride in them without an adult. Other than that, I don't see much demand.

Nothing fun about a commute or I would say any fun wears thin after the first 1000 times. I would gladly pay $20K extra for a self driving car today even if it was limited to only self driving on the highway (user took over at on/off ramps).

I wouldn't go $20k extra for an autonomous car, but semiautonomy and a $30k pricetag is why the Tesla 3 is tempting.

I can't readily afford a second car, though, so I won't buy one. :|

Why would anyone buy a fully autonomous car?!?

They would be in a fleet and dispatched as needed.

Exactly! Car ownership will become unnecessary for most users. I would much prefer to use a subscription-based service than be forced to take on ownership of an autobot.

 

[Jim Z]

Dietz[/url]":1nm6g3c3]So, are these same regulators going to shield car companies from civil liability when imperfect automation causes driver fatalities?

Yep, didn't think so.

Liability is one of those interesting questions which will need to be answered. Who's responsible for the car? The owner? The manufacturer? What happens when an autonomous car and a human-controlled car collide?

the Tesla case is interesting, because AutoPilot is likely only a tertiary factor in the collision. just strictly going on the event as described, the fault for that crash seems to be (in descending order):

1) The truck driver for not yielding the right of way (I believe he was ticketed)

2) The Tesla owner for not being in control of his vehicle,

3) AutoPilot for not being able to "see" the broad side of a 40-foot trailer.

in Michigan, with a two-vehicle crash the investigating officer can assign fault to either party, or assign "shared fault" by noting both parties took a "Hazardous Action" contributing to the collision. Had this happened in Michigan (and had the Tesla owner survived) it's probable the truck driver would have been cited for a "2600.Failed to yield right of way," and the Tesla driver would have been popped for "2300.Fail to have car under control."

 

[Luddite Curmudgeon]

srs[/url]":3e4exzh3]Not being able to see the road (and obstructions) due to the sun is far from "not perfect".

Leaving aside the sensor failure:

1) Why didn't the car stop when there was a sensor failure?

and

2) Why aren't there redundant systems using different technologies acting as backups?

The redundant system is you, the driver. Using Mark 1 eyeballs connected to the Mark I brain that commands the Mark 1 limbs to do something.

 

[sryan2k1]

srs[/url]":1gts6pw4]Not being able to see the road (and obstructions) due to the sun is far from "not perfect".

Leaving aside the sensor failure:

1) Why didn't the car stop when there was a sensor failure?

and

2) Why aren't there redundant systems using different technologies acting as backups?

#1 No sensor failed. It simply encountered a known deficiency in what it could sense. It was never claimed to be perfect.

#2 There are several (Ultrasonic, Radar, and Optical(Camera)), but again they are not perfect, which is why Tesla says you always need to pay attention. They could add LIDAR (Musk doesn't want to) and a whole bunch of other stuff and then the car could cost twice what it does.