Google Scales Back Self-Driving Car Ambitions

[Kabeyun]

I see your point, but there is a difference: Apple is a hardware maker in need* of a new product line. Google is not, and this always seemed like more of an experimental side project for Google.

*This need may be mostly perceived. They're still raking in cash through the iPhone.

I see your point, but there are lots of similarities too. You're speaking to strategy, and the need thing is arguable. But at their core Apple is a thing-making company while Google is a data company (after all this Android time they just released their first phone). It seems more natural for Apple to strike out in an experimental direction with a new thing than Google. Frankly I've been asking for an Apple Car for a decade, but what I probably wanted was a CarOS plus Apple design. I think the only reason it didn't seem absurd for Google to make a car was the natural focus on the data involved rather than the vehicle itself. Both Apple and Google seem to have realized this, which it why both seem to have migrated to systems rather than vehicles.

 

[v0lume4]

Actually, the back-pedalling may have been Tesla's fault because of the negative publicity generated by AutoPilot.

The problem with self-drive is that it won't be ready until it can demonstrate that it is virtually infallible. Rational thinking may suggest that it only has to be as good as the average driver to produce a major advance in road safety - but unfortunately, rational thinking won't have much of a role in the public and legal acceptance of self-driving.

For a while, every incident involving a self-driving car is going to be international front-page news. The first court case over a third-party death caused by a self-driving car is going to be a media circus. Adoption of self-driving cars is going to depend on changes in the law on liability and insurance (are you going to get into a self-driving car if you're liable for the consequences of bugs in Tesla/Google/Apple's last software update?)

Reality is, our pathologically risk-averse society has developed an irrational blind spot when it comes to letting a fallible human control a ton of metal hurtling around at 70mph (the accident rate is actually quite low, considering - turns out that ape brains evolved to tarzan through the trees are really good at it). That free pass isn't going to be extended to robots.

I think we're currently at the point where self-parking, lane-keeping, cruise control, auto braking etc. are at the limits of what can safely be done without causing driver complacency. The industry now has to make a quantum leap[1] to a system which allows the driver to kick back and start playing Pokemon - because typical drivers are not going to pay attention to the road and stand ready to intervene while a car drives itself. What we have now is a good proof-of-concept (and maybe a slightly better lane assist) - it still has a long way to go before mass acceptance. There are significant technical challenges and significant cultural/social challenges.

I'm sure it will come eventually, but its not going to be pulling in profits next year.

[1] (in the correct sense of the term - i.e. no intermediate states)

Exactly this. Well said.

I don't want a self-driving electric car.....I just want more choices for great electric cars with extended range and low prices. I like....no I LOVE driving my own car (Leaf) and just want better. It's going to take a lot of convincing for me to let a car drive me. I'm such a control freak I don't want any other person than me to drive....much less 1's and 0's. I'm sure self-driving cars will come eventually but the time is not now. So Google and Apple, try working on perfecting the car itself, then worry about the software a little further down the line.

Haha! I'm the same way! I LOVE driving! To be honest, a feature I wouldn't mind having is auto pilot for the interstate (a la Tesla), but in nearly every other circumstance I'd rather drive myself. I love the flexibility. And it's just fun.

 

[Zirel]

Childish response much? First you reply to an obvious joke with a rant, and then you want to argue about something that we don't even know exists. Do you have any proof Apple is even making a car?

Dude, deal with it.

Apple will make a self-driving car.

Google will not.

 

[v0lume4]

Waiting to see if comments here are as observant of Google's folly as these about Apple's only two months ago.

At the risk of sounding like a real downer, the difference between Apple and Google is that I never even believed in Apple's self-driving initiative to begin with.

 

[jmgregory1]

Until self-driving cars develop a conscience, I fail to see how they're going to impact the automotive industry. There are just far too many instances where a human has greater control over a given situation. How would it communicate with other drivers at a roundabout? Or give way in narrow roads when you're behind a parked car? How would it fight to change lanes when Mr BMW or Mrs 4x4 are far too busy to slow down and let you out? The list is endless.

Self driving cars will only really work when all cars have the ability to communicate with each other. Otherwise there will situations where a human's stupidity will overcome the software / hardware's ability to maintain control.

 

[Jeremy1026]

How would it communicate with other drivers at a roundabout?

You don't. Thats the point of roundabouts. While there is a car in it, in a position that blocks you from entering, you don't enter. When there is a gap that is safe for you to enter it, you enter it. Once in the round about you do not stop.

 

[merkinmuffley]

Smart move, the margins on cars are thin.

 

[Fender2112]

At the risk of sounding like a real downer, the difference between Apple and Google is that I never even believed in Apple's self-driving initiative to begin with.

I always believed Apple's agenda was more about developing the software platform and licensing it to auto makers. Same with Google.

 

[theluggage]

Once in the round about you do not stop.

...obviously not driven in the UK recently Most of the big roundabouts are sprouting traffic lights because the roundabout idea just doesn't work above a certain traffic volume & when modern cars can take them at 40mph. Not so bad unless you using the only entrance that isn't light controlled, your turn relies on other traffic not shooting orange lights and there's some dismount-for-nobody cyclist who insists on cycling around 5-lane roundabouts.

 

[Kabeyun]

At the risk of sounding like a real downer, the difference between Apple and Google is that I never even believed in Apple's self-driving initiative to begin with.

Why? They have far more experience than Google actually building products, and hired an army of automotive experts to work on the idea. Sure, Google has a good maps database and some early concept proofs made the papers, but it takes a lot more than that to build a car.

Honestly, I think the big benefits of these efforts are to slap Detroit in the face with a little future reality, and to help these companies double down on their automotive systems focus, which they're now both doing. Pity, because I'd have loved to see a Cupertino-designed car. They don't win design awards for nothing.

 

[alfonsog]

I wonder if a self driving car can ever successfully drive around Naples, Italy. Never seen craziness like that, my cousin drove us around... completely ignoring every traffic signal, stop sign, driving wrong way down one way roads with cars double parked along each side with signs every 5 feet saying no parking. If someone actually stops at a light or stop sign honking and everyone mashing together ignoring any lanes with scooters taking up all available space between cars...

Also here in Florida with the 90+ year old drivers that still drive although they have dementia and no one barely can do a 4 way stop correctly and forget about roundabouts most people have no clue.

And us Americans who can't adopt the metric system and constantly elect politicians against our best interests are never going to accept car socialism.

 

[Apple Knowledge Navigator]

You don't. Thats the point of roundabouts. While there is a car in it, in a position that blocks you from entering, you don't enter. When there is a gap that is safe for you to enter it, you enter it. Once in the round about you do not stop.

This is how they should work, but not everyone follows the rules as I'm sure you're aware. Particularly during rush hours.

 

[Harmonious Zen]

Surprising to me that two of the biggest names in tech have abandoned larger ambitions. Maybe goes to show that manufacturing a car from the ground up really isn't that easy. Which makes what Tesla has done all that more impressive.

 

[Pootmatoot]

Nailed what exactly?

All tesla has shown is that they can make an electric car. If they can deliver on their orders, that is.

Even their self-driving tech is nowhere near prime time.

The most important thing Tesla has done is pushing forward battery tech. It's lifting so many fields.

 

[TurboPGT!]

Maybe they found out that not only would no one ever buy a car powered by Google, hardly anyone is interested in autonomous driving, period.

Only the tech world is so fascinating by it, and tech companies with too much money to invest in such projects are dumping it down this hole, without asking whether or not they should.

Auto manufacturers and the general public, have little-to-no interest in autonomous driving at all. Until you win over both of those, you're precisely no where.

 

[ksnell]

Actually, the back-pedalling may have been Tesla's fault because of the negative publicity generated by AutoPilot.

The problem with self-drive is that it won't be ready until it can demonstrate that it is virtually infallible. Rational thinking may suggest that it only has to be as good as the average driver to produce a major advance in road safety - but unfortunately, rational thinking won't have much of a role in the public and legal acceptance of self-driving.

For a while, every incident involving a self-driving car is going to be international front-page news. The first court case over a third-party death caused by a self-driving car is going to be a media circus. Adoption of self-driving cars is going to depend on changes in the law on liability and insurance (are you going to get into a self-driving car if you're liable for the consequences of bugs in Tesla/Google/Apple's last software update?)

Reality is, our pathologically risk-averse society has developed an irrational blind spot when it comes to letting a fallible human control a ton of metal hurtling around at 70mph (the accident rate is actually quite low, considering - turns out that ape brains evolved to tarzan through the trees are really good at it). That free pass isn't going to be extended to robots.

I think we're currently at the point where self-parking, lane-keeping, cruise control, auto braking etc. are at the limits of what can safely be done without causing driver complacency. The industry now has to make a quantum leap[1] to a system which allows the driver to kick back and start playing Pokemon - because typical drivers are not going to pay attention to the road and stand ready to intervene while a car drives itself. What we have now is a good proof-of-concept (and maybe a slightly better lane assist) - it still has a long way to go before mass acceptance. There are significant technical challenges and significant cultural/social challenges.

I'm sure it will come eventually, but its not going to be pulling in profits next year.

[1] (in the correct sense of the term - i.e. no intermediate states)

Two things:
1. The frenzy will really come when the single update causes multiple crashes on the SAME turn, blind hill, roundabout, etc. in the same day because the car reads it differently.

2. Studies show the majority of people feel they are better than average drivers (lol - statistically impossible) so it will take a better than average autonomous car for individual people to trust it can do better than they can.

 

[tatonka]

Dude, deal with it.

Apple will make a self-driving car.

Google will not.

Apple will not make a self driving car .. and I am quite certain it was never their plan to begin with. Just like Google they would like to be the company providing the technology and software to do so, potentially .. we don't even know that for sure.
That said, I am happy hear arguments why you think they will.

And yes, Google will make continue working on self driving cars. The article actually states exactly that they made a move towards incorporating their technology into a real car, only that the current and foreseeable legislation will not allow fully autonomous cars for years. So they adapted .. sounds smart to me.

 

[dan110]

Google is another company that has gone off the rails. They were so busy cow towing and boot licking to Hillary this past two years, they forgot they were in charge of leading multibillion dollar companies.

 

[tatonka]

Google is another company that has gone off the rails. They were so busy cow towing and boot licking to Hillary this past two years, they forgot they were in charge of leading multibillion dollar companies.

Gone off the rails where exactly? And what exactly has Clinton to do with it .. or rather what has Google done that was so strange?
This is just another hollow rant that says exactly nothing at all.

 

[Kabeyun]

Google is another company that has gone off the rails. They were so busy cow towing and boot licking to Hillary this past two years, they forgot they were in charge of leading multibillion dollar companies.

Yes. That's exactly what it was. So amazed more people don't realize that. Thank you for the insight. I'm off to break up a child p**nography ring in a DC pizzeria.

 

[miniyou64]

Lol at these horse and buggy fans.
Self driving cars are coming. And I can't wait.

 

[Kabeyun]

Lol at these horse and buggy fans.
Self driving cars are coming. And I can't wait.

...until The Grid decides you're not going where you want to.

 

[Zirel]

Apple will not make a self driving car .. and I am quite certain it was never their plan to begin with. Just like Google they would like to be the company providing the technology and software to do so, potentially .. we don't even know that for sure.
That said, I am happy hear arguments why you think they will.

And yes, Google will make continue working on self driving cars. The article actually states exactly that they made a move towards incorporating their technology into a real car, only that the current and foreseeable legislation will not allow fully autonomous cars for years. So they adapted .. sounds smart to me.

Google only has failures in their repertoire, except for the ads and their search engine, but that's because they hype that so much and nobody cares to compete.

 

[tatonka]

Google only has failures in their repertoire, except for the ads and their search engine, but that's because they hype that so much and nobody cares to compete.

Yeah right .. another series of empty generalizations with no factual basis.

 

[MadeTheSwitch]

How would it communicate with other drivers at a roundabout?

Why does it need to? It's not like drivers communicate with each other now, other than an occasional turn signal which a car could read through cameras. But if it is really an issue, there are cars on the drawing board which will have such car to car communication. Although I think the real issue is someone hacking that protocol and creating havoc.