BMW Showcases Self-Driving Concept Car

SmartAboutThings writes: We've just been given a glimpse of what the future of motoring could look like, with BMW showing off its latest concept car, and it's self-driving. The Vision Next 100 was unveiled on Monday, at a ceremony celebrating BMW's 100th birthday, at Munich's Olympic Hall. This comes just a few days after BMW made official its intentions of competing with Google to build software for Self-Driving cars.

The Vision Next 100 has two driving modes, a driver mode and an autonomous mode, or 'ease' as its known. In driver mode the car operates mostly like cars do now, except the BMW indicates the ideal driving line and speed, but when the car is set to autonomous mode the steering wheel retracts and the two front seats turn to face each other. Perfect for two people to have a chat, and if it's only you in the car, put your feet up and relax.

Comments

self-promoting, too

[Anonymous Coward]

Is it 3D printed in space ?

Re:self-promoting, too

[aliquis]

Is it 3D printed in space ?

Where else?

Can the AI act like a BMW driver?

[Anonymous Coward]

It would be nice if the AI can act like a BMW driver. No turn signals, swerving into lanes if there is a gap wide enough... then slamming on the brakes to show supremacy. Oh, and making sure the AI's algorithm when parking is to take as many parking spaces as possible.

Of course, to protect the sides, add e-ink paint, so the vehicle looks like it already saw the business end of a key.

Re:Can the AI act like a BMW driver?

[jfbilodeau]

http://9gag.com/gag/aVW1LGM/yo...

Re:Can the AI act like a BMW driver?

[prefec2]

Jerk mode cost extra. And when you want to drive that way you can do it yourself. As a BMW owner it should be natural to you to do so ;-)

Re:Can the AI act like a BMW driver?

[fisted]

That strategy wastes fuel, not saves.

Re:Can the AI act like a BMW driver?

[Dutch Gun]

The last BMW driver I noticed not only took up a handicap spot near my bank's ATM, but carefully straddled his car over both handicapped zones, so as to make sure one of those pesky handicapped peons wouldn't get too close to his car. In fairness, it was during off hours, so there was unlikely to be someone who needed it, but the thing that killed me was that the handicapped spots were actually no closer to the ATM than the normal spots, which were also open. For some people, I guess the world truly does revolve around themselves.

Still, I don't think this is exclusive to BMW or other luxury car drivers. I think it's just human nature to get even more annoyed when we see that it's not only an asshole, but a rich asshole, and we tend to remember those examples.

Re:Can the AI act like a BMW driver?

[Grishnakh]

Apples and oranges. People who suck at driving and drive 15 under the speed limit and are generally inattentive aren't assholes the way that aggressive drivers are. They both suck, to be sure, but it's two totally different types of suckage. The OP posted what he observed to be the common behavior of BMW drivers, which is totally different than what you have observed to be the common behavior of Subaru and Prius drivers; you're likely both right.

Some of this is probably geographical, however. Where I am, I haven't seen such annoying behavior from Subaru or Prius drivers at all. However, I do see a *lot* of people driving 10 under the already ridiculously low speed limits we have here in Virginia, but not in those cars. I think they're usually in some crappy American car.

maybe you shouldn't be driving.

Unfortunately, this applies to the majority of the driving population. That's why we desperately need autonomous cars.

Re:Can the AI act like a BMW driver?

[fisted]

It's called gravity assist. You gun it down a hill then coast up the next.

Which still gives a net loss of efficiency, thanks to non-linear air drag, mainly. Plus, it should be fairly obvious that you're not going to gain more energy by going down a hill than you'll be expending on the climb of the next hill of the same height, regardless of what speed(s) you go.

That said, unless I'm misremembering it, ICEs efficiency peaks under high load, so by coasting uphill you're missing out on an opportunity to operate your engine at max. efficiency, even.

Re:Can the AI act like a BMW driver?

[arf_barf]

I'n not a BMW owner but I would rather take that AI than one based on a Prius owners driving pattern ;-)

Funny thing is, in my area, the local population consists mostly of BMW and Prius owners...a deadly combination if you ask me...

For Ubber!

[fbobraga]

Joint-venture?!?

Re:For Ubber!

[grahamsaa]

I accidentally rated this comment as informative, so I'm posting here to nullify that.

If only

[monkeyxpress]

If only making a few renderings and CG animations was all that was required to make those driverless car things.

Re:If only

[dwsobw]

It is a real car (they showed it on their "birthday" party). But it is really hard to tell what footage is CG and what is not.
Even the (partially) real footage looks unreal: https://www.youtube.com/watch?...

I don't know

[fyngyrz]

My first reaction to the seats turning to face each other was "so if the driver has to take control.... huge delay." My second reaction was "...and if you hit something, you'll be oriented sideways, which seems like a really bad idea."

Hmmm.

Re:I don't know

[Ogive17]

Not to mention how uncomfortable it will be to bump knees with the other passenger. I'm not sure how much room they think they have inside the car but very few cars on the road now have the interior space to effectively pull this off.

Re:I don't know

[N1AK]

It's all about balancing risk and benefit. If drivers only have to take control once every couple of hours and then being able to use the couple of hours better likely outweighs the small delay in those circumstances; from a safety perspective it is unrealistic to expect someone to be poised ready to act for hours straight while the car drives itself so the drivers ability to react to sudden dangers the car can't handle is very low anyway.

From a crash safety perspective if travelling in a self-driving car in 10 years time was say 90% less likely to result in an accident that could cause injury then you'd likely still be vastly safer travelling in a less safe position then than you are travelling in a safer position now. I'm not sure where the cut off is but given that our time is limited at some point you have to put a value on how you spend that time.

Re:I don't know

[drinkypoo]

My first reaction to this car was BMW's shoddy PR approach. They simply shit all these little video clips around 1 minute long onto the internet, and various automotive outlets reposted them to youtube. My second is that you could not possibly build a car which let the occupants swivel around or put their feet up on the dash, for safety. There are actual laws which your design would fall afoul of.

Re:I don't know

[dwsobw]

Well they also actually showed it on their "birthday" event, their PR is probably betting on everyone (or at least the automobile-centrics, e.g. journalists) watching the CG commercials thinking that can not be possibly real and then seeing the real thing.

Also they are most likely driving this prototype only on private property, much laxer laws there ...

Re:I don't know

[AthanasiusKircher]

My first reaction to the seats turning to face each other was "so if the driver has to take control.... huge delay."

This is one of those things that everyone thinks, but seems completely unreasonable if you think about it a little more deeply. I know that self-driving cars right now require the operators to be attentive and take over when necessary -- because they are still being "trained" in their early phase.

However, this is simply not a reasonable thing to expect the general public to do with a self-driving car. The first mass-market car that truly claims to be "self-driving" should NEVER require human intervention to "take control" suddenly. The car needs to be able to handle all situations adequately, and if it can't (e.g., case of impending major mechanical failure), it should have a backup system that will at least exit the current traffic pattern to the side of the road and come to a halt safely.

Anything less is a huge safety problem.

The idea that a human driver will just "take over" control in an emergency is incredibly stupid and naive. Think about how many accidents are caused by drunk drivers. And most of them are caused because drunk drivers have impaired reaction ability -- and a few tenths of a second is an issue.

Now imagine some rich guy who buys a BMW to drive him to work every day. He'll bit sitting there drinking coffee and reading a newspaper, and you somehow expect him to suddenly "take the wheel" when the AI can't cope? Forget about seats turned sideways or whatever -- you're talking a couple seconds minimum before (1) he realizes there is a problem, (2) manages to get rid of his newspaper, coffee, and breakfast sandwich before he can, (3) hit the button to take control, and (4) actually adjust his body to the controls. And with that amount of delay and breakfast bits flying around the car, there's no guarantee that a human driver would even be able to respond well and make safe maneuvers for a few seconds even after he takes control.

Three seconds at 65 mph is almost the length of a football field. Once we get to the point that drivers expect a car to drive itself MOST of the time, we simply cannot rely on the idea that drivers will be alert and ready to take over at a moment's notice. Moreover, I think it will be actually MORE dangerous to have a human driver take over in most such scenarios when unprepared and emerging from distraction.

So, actually, some ideas about the BMW design are better -- assuming that the AI actually functions flawlessly (which I think we're still a long way from). But once we get that level of AI, it's probably better to encourage people to act like they're not really a "driver" anymore.

My second reaction was "...and if you hit something, you'll be oriented sideways, which seems like a really bad idea."

We already have plenty of buses, shuttles, long limos, etc. which have some sideways oriented seating. I realize those seats aren't generally at the front of the vehicle, but still -- if AI is not demonstrably better than most bus or limo drivers in completely autonomous driving, then it shouldn't be allowed on the roads.

Re:I don't know

[CastrTroy]

This is my biggest problem with self driving cars at the moment. They still aren't good enough that you can take a sleep or watch a movie while they drive you around. You are still expected to be ready to take over at any moment. I really don't see too many advantages of such a feature. I'm much more likely to be paying attention if I'm actively engaged in controlling the car. I think that many people would try to watch a movie or read a book anyway, which would make the self driving car more dangerous than if they were just driving it themselves.

Re:I don't know

[PolygamousRanchKid]

Not to mention how uncomfortable it will be to bump knees with the other passenger.

Not to mention how comfortable it will be to jump bones with the other passenger.

Forget your Mile High Club . . . this one will be the difficult Beanie Baby to get in the set . . .

Re:I don't know

[pr0fessor]

This is one of those things that everyone thinks, but seems completely unreasonable

Right it's entirely unreasonable to believe that unforeseen failures can occur those things never happen. It's been decades since the last time a car had a re-call because it caused a dangerous and possibly fatal failure. It's not like people, birds, and other wild life have ever ran out in front of cars without warning and I have never seen a new car broken down on the side of the road.

Re:I don't know

[MindStalker]

I seriously doubt they will expect the driver to take over at a moments notice. The cars vision seems to be good enough that it can stop in reasonable time for any reasonable issue. The only time I'd expect a car like this to want to drive to take over is construction/badly marked roads. The car should be able to have enough information to be able to force the drive to take over a good half mile before any such issues arise, or at worse case slow to a stop until the driver has taken over. No need to be "instant".
I think a good passthrough system would be a remote "call center" where a driver could remotely drive the car through such areas or to a gas station or other resting place as well.

Re:I don't know

[dwsobw]

GP's claim seems to be that there should never be a reason for the driver to take control, i.e. that the car can autonomously handle people, bird, other wildlife and any condition the road can be in. Sure if that system is faulty then the passengers are in trouble, but that is not in any way a fundamental difference from the way it is now. Just one of the software systems in your car is a bit more complex. Probably not by that much since you already have all those fancy assistance systems anyway (e.g. line keeping assistants).

Re:I don't know

[Ol Olsoc]

Not to mention how uncomfortable it will be to bump knees with the other passenger. I'm not sure how much room they think they have inside the car but very few cars on the road now have the interior space to effectively pull this off.

Could be great for car secks tho'!

Re:I don't know

[koan]

Shannon Hamilton: That's it. You're dead, mallrat! I'm gonna fuck you up beyond repair!

Brodie: Ladies and gentlemen, this tall drink of water headed my way is a pillar of the shopping community who informed me earlier today of a nefarious plan of his to screw my girlfriend in an extremely uncomfortable place.

Gil Hicks: What... like the back of a Volkswagen?

Re:I don't know

[fisted]

This is my biggest problem with self driving cars at the moment.

Yes, it's really a shame they're on the market for so long already, yet /still/ suck ass.

They still aren't good enough that you can take a sleep or watch a movie while they drive you around.

Fully agree. It's a pain in the ass to sit there, every morning, watching my car drive me to work but not being allowed to distract myself. Why are they selling those cars already? Obviously the technology isn't there yet. Come to think of it, you also still have to manually remove the dead hookers from the trunk. How autonomous is that?!

Re:I don't know

[Ol Olsoc]

The idea that a human driver will just "take over" control in an emergency is incredibly stupid and naive. Think about how many accidents are caused by drunk drivers. And most of them are caused because drunk drivers have impaired reaction ability -- and a few tenths of a second is an issue.

To listen to the proponents of autonomous cars speak, people get in a fatal accident every time they get behind the wheel, and we're all dead now.

Spare us, just a little, eh? Untold millions of miles are driven by meatbags without incident, and unless autonomous cars are never ever going to fail, never ever going to ever be involved in any incident forever, at best, they are merely an incremental improvement in safety.

Re:I don't know

[Ol Olsoc]

This is one of those things that everyone thinks, but seems completely unreasonable

Right it's entirely unreasonable to believe that unforeseen failures can occur those things never happen. It's been decades since the last time a car had a re-call because it caused a dangerous and possibly fatal failure. It's not like people, birds, and other wild life have ever ran out in front of cars without warning and I have never seen a new car broken down on the side of the road.

You touch upon an important thing. It seems to be that the Autonomous worshippers believe that there is no hardware in a vehicle - only software that will apparently be the first 100 percent flaw free software ever.

Re:I don't know

[AthanasiusKircher]

Right it's entirely unreasonable to believe that unforeseen failures can occur those things never happen.

Of course failures can happen. What I'm saying is putting a so-called "self-driving" car on the road which can fail in such a way as to require a person to take over SUDDENLY is a major safety hazard, both for the "driver"/passengers and for those cars around it.

It's not like people, birds, and other wild life have ever ran out in front of cars without warning

If an AI can't respond to such things, it should NOT be allowed on the road. Period.

And if the "failure mode" in these scenarios is to suggest that the distracted businessman reading the newspaper while eating his breakfast is supposed to suddenly take the wheel... well, that's just not reasonable. And you can say, "Oh, but these 'drivers' should be alert," but that's just not how people are going to use a "self-driving" car in the real world.

If the distracted guy on his commute who takes 3 seconds to take the wheel is going to do better than the AI in avoiding "people, birds, and other wildlife," then I definitely would not buy such a car, and I would propose that such an AI is a TERRIBLE driver and should be banned from the roads.

and I have never seen a new car broken down on the side of the road.

Different scenario. Re-read my post. If a major failure happens, there needs to be a backup system that can get the car to the side of the road and to a stop safely. Clearly, in your scenario, the human driver was able to do that. Why precisely should we demand less of an AI "driver"?

I didn't say that mechanical failures would be impossible on future cars -- but the AI should be designed in such a way (with various backup systems) that it's basically the LAST thing that will fail... and if it does so, such a failure should never be sudden and in the middle of a highway.

In all but the most catastrophic of failures, the human should NEVER have to suddenly take control. It's dangerous, and it's a stupid idea.

Re:I don't know

[AthanasiusKircher]

To listen to the proponents of autonomous cars speak, people get in a fatal accident every time they get behind the wheel, and we're all dead now.

Just to be clear, I am most definitely NOT a "proponent of autonomous cars." If they ever get up to the standards I'm proposing, I'd consider them -- but I think such AI is still quite a ways off.

Also, according to the CDC, approximately 1 out of every 100 people in the US will visit a hospital due to a motor vehicle accident in a given year. Approximately 1 in 10,000 people will die in such an accident. They are a leading cause of serious injury and death among accidental causes.

So, no -- we're not all dead. But car accidents happen... a LOT. Self-driving cars need to be programmed to deal with scenarios that are likely to cause them, not flash a warning and hand off control at high speed to an unprepared distracted human.

Re:I don't know

[jimbolauski]

This car is a concept car and is no where near a production model. There are two things concept cars are used for one is to put in radical designs that get people talking about your products. The other is to learn how to integrate new features into the car.

Re:I don't know

[SlaveToTheGrind]

Now imagine some rich guy who buys a BMW to drive him to work every day. He'll bit sitting there drinking coffee and reading a newspaper, and you somehow expect him to suddenly "take the wheel" when the AI can't cope? Forget about seats turned sideways or whatever -- you're talking a couple seconds minimum before (1) he realizes there is a problem, (2) manages to get rid of his newspaper, coffee, and breakfast sandwich before he can, (3) hit the button to take control, and (4) actually adjust his body to the controls. And with that amount of delay and breakfast bits flying around the car, there's no guarantee that a human driver would even be able to respond well and make safe maneuvers for a few seconds even after he takes control.

I think you've perfectly articulated why this kind of design -- allowing the "driver" to essentially lay back, pop open a newspaper and ignore the road -- is a Really Bad Idea.

Cars already have bits and pieces of autonomous capability today -- radar cruise control, lane control, etc. But in all of those cases, the driver has not only the the ability to instantaneously take over when the driver sees that the technology is making a poor choice, but the expectation to do so. Airplanes are even more autonomous -- the vast majority of any commercial flight you take these days is controlled by autopilot, yet there are humans -- plural -- sitting at attention for the inevitable moment when something goes wrong. And that's up in the big, largely empty sky. Why in the world would we want it to be any different for millions of cars tearing around in close proximity?

Re:I don't know

[N1AK]

This is my biggest problem with self driving cars at the moment.

There's been an argument that more automation in driving has made drivers more careless and distracted. There may be some truth to it but thus far the safety figures are still trending down so it's not a massive concern to me. Clearly there's a risk that we enter a dangerous period where cars are almost good enough to drive themselves but still get into dangerous positions they can't get out of safely on a regular basis but I don't think this is inevitable. Most self-drive functionality so far is based on someone driving and the car intervening where necessary, auto-braking if you're about to hit something for example. A likely extension of this is that there will be cars on the road in the next couple of years that could drive autonomously safely but which are limited to stepping in where there is a clear safety issue the driver isn't responding to.

Re:I don't know

[AthanasiusKircher]

I think you've perfectly articulated why this kind of design -- allowing the "driver" to essentially lay back, pop open a newspaper and ignore the road -- is a Really Bad Idea.

And that's precisely why I've been arguing here for years that AI cars a lot far off than many people think. Because until we get to a situation where the kind of design under discussion is safe and practical, true "self-driving cars" won't be feasible.

Airplanes are even more autonomous -- the vast majority of any commercial flight you take these days is controlled by autopilot, yet there are humans -- plural -- sitting at attention for the inevitable moment when something goes wrong. And that's up in the big, largely empty sky. Why in the world would we want it to be any different for millions of cars tearing around in close proximity?

The problem with this analogy is that airline pilots are paid to ensure the safety of hundreds of people on their aircraft. Their reason for being there is to ensure that the autopilot works and if it doesn't to take control.

The reason most people get into a car is because it takes them somewhere -- not because they are paid to drive or care deeply about being a safe driver.

And already we're seeing how distracted people are willing to be WITHOUT "self-driving" cars. People text, people talk on phones... people eat, shave, apply makeup, whatever. They do this knowing full well that they are supposed to be controlling their vehicle too.

And somehow people imagine that after we create a "self-driving" car where 97% of the time they DON'T need to pay attention, that drivers will still be like attentive airline pilots, ready to take over at a moment's notice?

Sorry, that's just not going to work in the real world. Whether or not you want it to happen, the vast majority of people are going to be reading the newspaper or their tablet while eating breakfast when they are supposed to be "driving" a self-driving car.

That's what will happen. And thus the only mainstream AI cars which should be put on the market are those which cannot depend on the driver taking control suddenly as a normal "failure mode."

Re:I don't know

[pr0fessor]

Sorry, autonomous or self driving cars just piss me off, I already have people texting while driving hitting my parked cars on more than one occasion. The technology we have is not ready to take over driving for these people but I see more and more stories and talk about it.

I'm afraid one day I'm going to have cars crashing through my living room and the not driver will still have not looked up from candy crush.

Re:I don't know

[AthanasiusKircher]

I actually agree with you. I think true self-driving cars are farther off in the future than most people think, precisely because of the reasons I articulated. Until cars can be THAT autonomous, they won't be safe.

Re:I don't know

[Dutch Gun]

Commercial aircraft are different because they're carrying hundreds of people at a time, and any serious failure is likely to result in scattered, flaming wreckage and death on a massive scale, not to mention the loss of millions of dollars in aviation hardware. In other words, the stakes are much higher than for that of a personal passenger vehicle on the road.

Short of head-on collisions at freeway speeds or driving off a cliff, most vehicular accidents are quite survivable in modern airbag-equipped cars, especially if the computer is at least able to start braking ahead of time to some reasonable degree.

Re:I don't know

[pr0fessor]

The problem is we already have people causing accidents because they are distracted by cell phones and what not, giving them a self driving car based on technology that isn't up to the task because it won't be at a price they can afford is just silly.

Re:I don't know

[pr0fessor]

I think that in order to make a safe self driving vehicle we will have to rethink how we imagine that vehicle operating.

Re:I don't know

[tnk1]

You miss the point. He's not saying it is unreasonable because nothing bad is ever going to happen.

It is unreasonable because if something bad does happen, it will already be outside the range where a human could successfully intervene.

The scenarios where a human taking manual control would be better would be the first scenarios that they would have to ensure the computer dealt with as well, or better than a human, when it comes time to release the vehicle to the general public.

Note I did not say that automated cars would always do better. If a pedestrian does something completely unexpected, like decide to commit suicide in front of your vehicle, it is very possible that the automated car would run that person over. But then, so would the human driver.

In the end, however, when it comes to braking, speed control, and attentive driving, a computer wins every time. That alone is enough to ensure a higher safety rate among vehicles.

And don't get me wrong, I don't want to stop having the option of driving my BMW. But it would be really nice to have the option to have it drive itself when I need to do rush hour, or take long trips.

Re:I don't know

[Lumpy]

My BMW X3 was the same price as a fucking Chevy. I think you need to look at car prices and see how the poor man's Chevy and Ford are now priced only for rich people.

A freaking Civic.. the worlds iconic economy base car for the poor person is over $18,000.00! That is utterly insane.

Re:I don't know

[SlaveToTheGrind]

And that's precisely why I've been arguing here for years that AI cars a lot far off than many people think.

Agreed -- likely infinitely so. It's an interesting network effect problem, where roads mostly or completely filled with autonomous cars likely would be much safer on balance, but where there are only a handful it's unclear that net safety goes up, and it may even go down (witness Google's recent conundrum over their AI being "too law-abiding").

It's not clear to me that people will have the collective appetite to push through that first phase to try to reach the second, if that's even possible to reach without the government effectively mandating autonomous cars, which in turn would require an improbable amount of political will.

Another inherent assumption in all this is that pricing would eventually get down to a "reasonable" level. First, "reasonable" is going to be perceived a lot differently by people at various points on the income spectrum. Second, if the human driver truly has no responsibility for what the car does, it seems like most liability would ultimately end up flowing upward to the manufacturers/mechanics and be internalized in the purchase/maintenance price, putting it that much further out of reach/further down on the priority list of many people.

Re:I don't know

[tnk1]

One assumes that you're either going to get the adequate software, or they're not going to sell the automatic option at all. If the price is too high for a safe self-driving car, they're not going to get any self-driving car.

If there is one thing I am happy about, it is the fear of self-driving cars ensuring that no one dares to cut corners on it.

Look, humans can be really good drivers, but they're also extremely shitty drivers. And it can shift from one to another with the same driver, based on their circumstances. Just this morning, some mother in her SUV was driving really slowly and failing to turn right on red because it looked very much like she was fussing with her baby or something else. She was *not* paying attention to the road, and of course, that's pretty par for the course around here.

Other times you have slow people in the fast lane, yapping on their phones without even as much as handsfree. Just removing those people from driving responsibilities would do wonders for traffic, as well as aggressive driving and road rage. And if you took the road ragers out of the loop and put their cars on full auto, their overblown reactions to similar scenarios would also improve safety.

Yes, I don't want to end up being a passenger in some automated pod, but if I can go on autopilot for awhile, instead of having to experience the "thrill of the open road" whether I want to or not, I would love that.

Re:I don't know

[pr0fessor]

I don't think we will have a safe self driving car like we might imagine them, we would have to rethink much of what we have imagined and possibly redefine our concept of a car.

Re:I don't know

[Grishnakh]

The problem is that drivers **cannot** take control, unless it's pre-planned. You can't expect drivers to have their hands on the wheel, ready to take over at a moment's notice. It's simply impossible; if you require that, humans will fail nearly every single time it happens. Humans just can't do that.

What *can* work is the car acting autonomously in certain places (like on highways), and then requiring the driver to take over outside of those situations, *as long as* the driver is given ample warning to be ready to take over. So if you're taking a 1-hour drive with 45 minutes on the freeway, you could have the car take over on the on-ramp and drive autonomously on the freeway, and then require the driver to take over on the off-ramp, giving plenty of auditory warning ("Arriving at off-ramp in 3 minutes, please be prepared to resume manual driving"), and then falling back to pulling over on the off-ramp if the driver has fallen asleep or something.

But any rules requiring drivers to be able to take over at a moment's notice **will** be a failure and cause catastrophic results. If autonomous vehicles are not ready to act autonomously without the need for manual takeover, then they don't belong on the road, or they need to be set up to be restricted to certain places or situations (like freeways, that really should be pretty easy to automate).

Re:I don't know

[Grishnakh]

Other times you have slow people in the fast lane, yapping on their phones without even as much as handsfree. Just removing those people from driving responsibilities would do wonders for traffic, as well as aggressive driving and road rage. And if you took the road ragers out of the loop and put their cars on full auto, their overblown reactions to similar scenarios would also improve safety.

You'd improve safety on both ends. The nincompoops wouldn't be holding up traffic or causing wrecks through their carelessness or idiocy, but also the road ragers would be mostly eliminated because all the things which make them rage (the behaviors of the nincompoop drivers mainly) would be eliminated. After all, the road ragers are usually set off by something, whether it's a slow driver in the fast lane, being cut off, another aggressive driver flipping them the bird, etc. Automated driving would eliminate all that stuff, so people wouldn't have much of a reason to get mad in the car any more.

As for the "thrill of the open road", good riddance. That was a fun thing back in the old days when there weren't many cars, and sometimes it's still a fun thing if you can find the rare place/time when there's no other cars on the road (3AM in a rural area?), but it's pretty much gone. Heck, even 3AM isn't fun to drive because there's always a bunch of cops out there looking for an excuse to pull you over.

The biggest obstacle I see to automated driving is actually the police. How are local police departments going to handle it when people aren't even driving their own cars any more, and they aren't committing actual infractions? How are they going to finance largesse in small towns with excessive ticket revenue? How are the police going to be able to pull people over for Driving While Black? We might even see automakers suing local governments left and right (and winning) for harassing their customers when the customers weren't in control of the vehicles but got pulled over anyway for some falsified infraction, which the vehicle is easily able to disprove by its logs and recorded video.

Re:I don't know

[Grishnakh]

I completely disagree. Human drivers are already so horrible that it's hard to do worse with autonomous vehicles, even if they have errors sometimes.

However, peoples' fear of autonomous vehicles and lack of understanding about death rates (it's better to have 1000 deaths/year because of software errors than 30,000 deaths/year because of stupid human errors) will delay the deployment of fully-autonomous vehicles, I'm sure.

But what I do believe will happen is that we'll see partially autonomous vehicles, namely on freeways. Freeway driving is far simpler than driving on surface streets anyway since the traffic flows in one direction and there's no stop lights. The only things you really have to worry about is traffic jams and unexpected obstacles (like retreads and other debris, or wild animals). So what we'll probably see before too long is cars which take control from you when you enter a freeway, and then let you resume control when you exit the freeway. You won't be expected to take over at a moment's notice (that would be stupid, as you said), and they'll come up with algorithms for dealing with the problems it's possible you'll see on a freeway. You'll be warned ahead of time that you need to be ready to take over on the exit ramp, and if you're unconscious then it'll default to just pulling over in a safe spot.

But going from that to fully autonomous on all roads will probably take a while.

Re:I don't know

[Grishnakh]

The Civic is not a poor person's car. If you're on a budget and want a Honda, the Fit is the car for you: it's smaller and cheaper and crappier than the Civic.

Just because the Civic carries the same name as the low-end car Honda sold in the 70s doesn't mean it's aimed at the same market.

And if you're really poor, new cars probably aren't a good choice for you anyway. That's why we have used cars. The average age of a car on the road now is over 11 years. Poor people are driving the cars middle-class people used to drive 10 years ago. If you can't afford $18k for a new car, then you should be driving a used car that you can get for $3-8k. There's plenty of great cars in that price range.

Re:I don't know

[arf_barf]

You are spot on. Here are my two cents:

1. Most driving decisions require context and I would venture that it's not possible to acquire context within a second or two for a driver that was in distraction mode
2. I think that the current AI generation is already better then most drivers on US roads.
3. From a human driver's pespective, I would like to see an identifier on a vehicle if it's in autonomous mode

Re:I don't know

[delt0r]

And its like most of the things that run out in front of cars get run over. Whats your point. That the magic of human cognition will somehow change the laws of physics? That they can somehow see into the future if the force is strong with them and despite reacting thousands of times slower somehow change the cars direction through sheer willpower alone?

Basically even on a geeks tech website we have a bunch of Luddites that are irrational when it comes to car they can't drive. Guess what your a shit driver and the machines can't replace you soon enough.

Re:I don't know

[delt0r]

I think your being irrational because you believe that human cognition is somehow magic. It is not. Machines will be better drivers than 99.9% of current drivers in the near future, and better than all humans a little while after that.

Re:I don't know

[pr0fessor]

Machines fail all the time and the more complex the more likely it will fail regardless of whether it is a better driver. My computer does math many times faster and more accurately than I do but not when the hard drive fails then it doesn't even boot and a self driving car will be the same.

Re:I don't know

[delt0r]

Yea cus having a human at the end of all that complex mechanical machinery we already have is somehow magic and make failures not serous. Like when a human has a tire blowout. Or if the breaks fail, human cognition will intervene and change everything. NOT. We already have life critical software/machinery everywhere. Traffic lights, fly by wire, brake systems, aircraft hydraulics, air traffic control systems, medical equipment,.. etc.

Concept Car, like Windows 1.0: Vaporware

[evolutionary]

Oh...Shiny Ones! Google's is a lot less fancy. but...they have real cars. :D

Re:Concept Car, like Windows 1.0: Vaporware

[GrumpySteen]

Windows 1.0 was not vaporware. It was ugly, ran slowly and everyone kinda hated it, but it was a real product that was put up for sale.

Re:Concept Car, like Windows 1.0: Vaporware

[evolutionary]

See article:

http://www.techradar.com/us/ne... They were trying to sell it before it was ready and had to wait 2 years before an actual product. Microsoft put up a LOT of things for sale that weren't ready. Windows 1.0 wasn't even the worse. Try Windows ME (worst on record), Windows Vista (almost as bad). The best thing Microsoft ever designed were their keyboard (really nice!)

Oh, and like a driverless car, Windows 7+ will also take you for a ride (if you don't turn off Windows Update using the Services dialog): Whether you want it or not, whether you've agreed (outside of installing meaning you agree to the 10k+ word EULA), if Windows Update is left running, you will get a "free" upgrade to Windows 10! Complete with spyware (MS insists is "telemetry" data) that you CANNOT shut off (though you can turn some elements off in Enterprise apparently but not all).

Wow, who would have thought MS was so related to this topic, it turns out...it really is. The government agencies will of course put in a backdoor for "telemetry data" and "law enforcement". Makes me wish I could get a car from old day Russia, which had NO computers and you could repair yourself without proprietary software. Imagine a virus in your cars, or ransomware.

Re:Concept Car, like Windows 1.0: Vaporware

[dwsobw]

The BMW car itself is also real. They actually showed it on their "birthday" party ...
It is really hard to tell what is CG and what is not but this seems to be a actual photo: Vision 100

Re:Concept Car, like Windows 1.0: Vaporware

[GrumpySteen]

Vaporware is a product that is announced and/or sold without ever being made. Windows 1.0 was, in fact, inflicted on the world. It was a terrible product which was shipped late, but it existed.

Correction

[wonkey_monkey]

BMW showing off its latest concept car, and it would be self-driving if it actually existed, which it doesn't

FTW.

In other news I will be unveiling my concept airplane/car/submarine/helicopter/spaceship.

Re:Correction

[dwsobw]

The car is real. I am unsure about the self-driving part and what is CG and what is not ...
But you can find footage of their birthday party that is at least partially real and shows that car driving around ...

It's insightful

[fbobraga]

not informative: is a guess (and a fun one!), not a fact :-)

Correction to the correction

[wonkey_monkey]

FTFY.

Oops.

Incredible Car

[dwsobw]

That car looks really interesting even without the self-driving part.
The wind-shield of the wheels of the car (side view) actually stretches (streched right) and relaxes (unstreched right) to get an nice drag coefficient (cW) of 0.18-0.19.

gov needs to say no EULA for autodrive stuff t FAA

[Joe_Dragon]

gov needs to say no EULA for auto drive stuff the FAA does not let airbus or boeing pull that shit with autopilot software.

Also how will that hold up in court when some 3rd party victim is hurt / killed by a auto drive car?

Forward-thinking

[Gennerik]

This is where forward-thinking entities (Businesses, governments, regulatory bodies) should be getting together to come up with "future-proof" requirements in order to operate on roadways. Businesses should be working to set the standards of operation and how, if any, inter-car communication should occur. Governments and regulatory bodies should be setting down the standards for safe operation, and what requirements must be met by all automated cars in order to be sold.

Instead, it seems like we've got a bunch of car makers trying to push their own agenda and ways to control an automated car. There doesn't seem to be any spirit of cooperation or desire to set standards. Governments seem to be lagging, still trying to decide if they should allow automated cars instead of trying to push forward and realize that in some form it is going to happen. So instead we're going to have cars that can't communicate between each other, have no standardization, and no set requirements. And the governments are going to let it happen, because they can't move fast enough to react to anything newer than a few years in the past.

Re:ideal crash position

[fisted]

Why? Cars can also be rear-ended, you know. Come to think of it, the odds on whether the hit comes from the front or the back should be exactly 50:50.

Re:ideal crash position

[fisted]

Come to think of it a bit more, probably there's a slight bias towards the front since there should be more head-on collisions than people rear-ending with the rear end. I'd assume the impact of this on the 50:50 chance to be negligible, though

Too much, too soon?

[westlake]

when the car is set to autonomous mode the steering wheel retracts and the two front seats turn to face each other

This means that the driver can't possibly have the situational awareness or the time needed to intervene should autonomous mode fail for any reason.

Re:Too much, too soon?

[Overzeetop]

If the autonomous driver module is capable of independent operation, and yet unable to react fast enough to avoid a collision, there is no chance that a driver will be able to. A failure of the autonomous system wouldn't be apparent until it is too late for a human to react. And by failure, I mean a failure of both the autonomous system in operational mode, as well as a back-up system which is designed for fail-over control (and removal from the roadway/traffic).

Two years too late BMW...

[Lumpy]

Tesla has had self driving cars for two years.... all they had to do was push out a software change.

So it seems that BMW is desperately trying to catch up to the ONLY car company actually innovating.

No need for fast reaction.

[DrYak]

My first reaction to the seats turning to face each other was "so if the driver has to take control.... huge delay."

Cars which are already on the streets now (including cars by BWM themselves, but also other constructors like Volvo) are already able to brake on their own to avoid a collision.
The driver is never expect to suddenly jump at the wheel (or at the brakes with the current generation of assisted-driving cars) in a split second.
But the driver might be required to do some intervention eventually.

E.g. with modern cars: You're on the highway, there's a traffic jam in front of you. You don't need to put back your feed on the pedal in a split second. The car's "forward collision avoidance system" and "adaptive cruise control" will detect the obstruction and progressively slow down and bring the car to a stop on its own. You're only required to interact to tell the car to resume its course (either touch the accelerator once the jam clears or push the "resume" button on the adaptive cruise control).

This could translate well with future autonomous vehicle.
In case of emergency, a robot car should be able to take a safe emergency behaviour (to avoid the danger).
e.g: safely pull out to the shoulder/emergency stopiing lane.
Then once there, it can safely wait for the human to drop the phone/wake up/whatever and do the necessary intervention.

My second reaction was "...and if you hit something, you'll be oriented sideways, which seems like a really bad idea."

In my humble opinion, that's a much bigger problem.
Lots of emergencies might require sudden decelerations. Not only collision (as you suggest) but also sudden emergency braking.
And being sideways sound as a not very safe position.
Would be better if both front seats turned 180 to face the back seats, offering a nice enrionment to chat among all the 4 passengers. (I've seen concept cars demoing this kind of stuff).
This backward position would be ideal in case of braking.
But then you'd be exposed to objects flying around (whatever was laying on the central table at the moment the car hit the brakes).

Please hurry

[slipped_bit]

EVERY BMW driver needs a self-driving car.

Do the indicators work?

[stoatwblr]

If so, it must be a robot in command. BMW drivers don't know how to use them.

Self Driving Uber?

[tmjva]

So will BMW market self driving Uber cars?

(I missed that story if it was reported previously.)