Hacker News new | comments | show | ask | jobs | submit login
Apple CarPlay (apple.com)
503 points by lele0108 1299 days ago | hide | past | web | 364 comments | favorite



Yay, more touch screen in cars. More ways to distract, frustrate and confuse users who are operating a big box of steel weighing 3 tons and going at speeds the human brain has never evolved to appropriately deal with.

1) Not only are touch screens a very poor interactor in the first place [0], but why do you think planes and other complex machinery have stuck with physical controls? For operating complex vehicles/apparatus, you just cannot do better than tangible controls. Knobs, switches, sliders can be operated without looking at them while giving rich tactile feedback, they have no modes = 0 risk for confusion, you know where they're going to be located on your dashboard regardless of what you're doing, etc.

2) Self-driving cars cannot come fast enough, and every single innovation in the car industry that does not go towards electric self driving cars is just useless fluff at this point. Seriously- then you'll be able to fiddle all you want with your phone, drink, travel while sleepy, arguing your spouse, whatever you want- we'll be saving tens of thousands of lives every year [1], and the secondary social benefits will be fantastic (less cars produced since they don't have to sit on a parking lot 99% of the time, people won't have to spend a year's worth of wages just to buy a car (and then a significant chunk to maintain it), etc.). If society were a game of Civilization, I'd be putting all of my resource points towards the "Self driving cars" achievement.

Of course the insurance companies, car manufacturers, oil companies, etc. don't want that to happen- but seriously, fuck those guys. The benefits on human society at large here are so significant that there is no room for caring about the feelings of greedy old white men.

[0] http://worrydream.com/ABriefRantOnTheFutureOfInteractionDesi...

[1] http://en.wikipedia.org/wiki/List_of_motor_vehicle_deaths_in...

(the views in this post are a bit on the extreme side- but that's how interesting conversations get started :-)


>Yay, more touch screen in cars. More ways to distract, frustrate and confuse users who are operating a big box of steel weighing 3 tons and going at speeds the human brain has never evolved to appropriately deal with.

This reduces distraction.

You have to remember to take into account human behavior. In the real world, people get into their vehicle and use their phone's map application, as well as text, phone, and music applications. Not only do they do this, they do it in overwhelming numbers.

This product takes common behavior that people engage in (and will continue to), and makes it safer and less confusing to engage in while driving.

Vehicles have knobs and touch-screens now. This tech doesn't change that dynamic, -it works with the UI elements already present in order to reduce fumbling around with your phone and tiny touch-targets on your mobile device.

>2) Self-driving cars cannot come fast enough, and every single innovation in the car industry that does not go towards electric self driving cars is just useless fluff at this point.

That's besides the point. They aren't here yet, and they won't be in the immediate future. Making driving safer in the meantime isn't fluff.

> there is no room for caring about the feelings of greedy old white men.

What a terrible, low-value comment to make in this forum.


In 2011, 3,331 people were killed in crashes involving a distracted driver. 3,267 in 2010. [1]

When you say "This reduces distraction." you mean compared to holding the phone and doing those actions, right?

The way I would account for human behavior is to intelligently disable certain functionality as the car is moving. Compared to someone else's life, how important is that text message? Even dictating a text message takes some cognitive load.

Let's not mix the word "safer" and "distracted driving".

[1] http://www.cdc.gov/motorvehiclesafety/distracted_driving/


> "The way I would account for human behavior is to intelligently disable certain functionality as the car is moving. "

This is about the only way to enable intelligent control over device functionality. Absent this additional link, a modern cell phone can't tell a driver from an occupant.

With this, Apple and the car companies offer people a carrot so that they pair their device [1] and now they can limit the distractions: you can control the apps that are available on the car interface, you can pare down the notifications they're receiving (enable DND mode by default while driving), you can disable direct use of the paired device while the car is in motion, etc.

[1] The carrot being GPS and music-selection features already available on many modern car dashboards.


>The way I would account for human behavior is to intelligently disable certain functionality as the car is moving. Compared to someone else's life, how important is that text message?

Thereby ensuring a situation where the person takes their eyes off the road and looks at their seat for their phone, picks it up, and starts reading and replying to texts while barreling down the road in a high velocity chunk of steel.

That's what happens now. Apple's technology aims to prevent that.

>Even dictating a text message takes some cognitive load.

Yes it does. It also is much safer than doing it on your phone as people can and will continue to do, even in states where it is illegal. Apple isn't a legislature. They can't tackle unsafe driver behavior by crafting laws, but they can do so via software offerings.


If you intelligently disable in-car functionality (such as GPS when the car is moving), people will reach for another device to get that functionality. It's naive to think that constraints will automatically change users' desires.


I believe Apple is sending out the wrong signal about driving+other activities at the same time.

Sure, people have already picked up bad habits, but that makes it ok to encourage that behavior if they are going to do it anyway?

Maybe Apple does make certain activities less dangerous, but it's still making access to distractions easier and I think that's a bad thing.


Not true. Driving a car with bluetooth, I now see how wonderful and useful this feature is. I don't have to take my eyes off the road and I don't have to hold anything to my ear. Its very very seamless.

If you can't talk while you drive, then you shouldn't have passengers either.


Yes, but you're still distracted, and that's the problem.

The obvious difference between a phone call and a conversation had with someone in the car is that the person in the car is also placing their life at risk. Unlike the person who isn't there, they can (a) see and (b) intervene when you're starting to drive like an idiot.

It's not that a live conversation is any less distracting. It's just that it comes with a built in safety mechanism.


> In 2011, 3,331 people were killed in crashes involving a distracted driver. 3,267 in 2010. [1]

That figure should be normalised to make the comparison fair. Presumably, there were more cars on the road in 2011 than in 2010.


So essentially, phones should just turn themselves off when moving? I find it highly unlikely that such a law would get any traction (implementation seems rather difficult - by the time it's fully deployed, we'd have been better off putting efforts into self-driving cars). Companies have no chance of implementing this optionally: Why would I buy a car/phone I can't use as a passenger? Deaf passengers would be doubly upset, as the presumably-allowed voice-control features wouldn't be accessible to them.


A good compromise on distractions and 'more inputs' is voice-based UI. Honestly, car companies and OEMs must be focusing on better voice-based UIs for cars rather than distracting touch screens. The Sync system in Ford cars is a well-intentioned example that doesn't work that well.


My pet peeve with Sync is they disable navigation input via touch screen when car is moving. Even when a passenger tries to enter it. Instead they force the driver to use their clunky voice-recognition to do this. Which is much more distracting.

But this is a minor quip compared to how MS implemented Sync Services. It's 2014 and they use a voice(!) line from a pre-registered cell phone to connect to their services.


That's a very American way of thinking about things.

I assume you're American (or British), because voice-based systems are nearly useless for non-native english speakers.

I'm a fluent english speaker, but due to my non-american accent, and the fact that my language is not one of the top 10 used in the world I still cannot operate Siri (or other voice systems) properly, and the dictation feature is nearly useless to me.


I'm a heavily accented non-American, non-British (though I can claim English is my native language since it's been my language of instruction since Kindergarten).

Google's voice recognition system, and of late, Siri both work fine for me. Sync, iPhone 'voice control', etc. are next to useless. In certain cases, I've tried to adapt. For example, my immediate family, rather than have their names or what I call them in my other native language, I've named them using common English words. So I can just say "Call Dad", and the speech recognition system works.

When I say that OEMs and auto-makers should work on improving voice recognition systems, I mean the entire processing system: sound acquisition, signal conditioning, and machine-learning based speech recognition.


From personal experience, a fixed location screen; my dash; is more distracting than my phone which can be moved into my line of sight.

That out of the way, my VW will duplicate much of what appears on the nav screen into a small display between the speedometer and tachometer. Having used a few heads up displays in cars I am not if I really want to go that route.

Still, when moving the car should be displaying relevant information only. I would much rather have a talking car than one trying to show me


It's worse than low value, it's flat-out racist.


>why do you think planes and other complex machinery have stuck with physical controls

Avionics has been moving away from tangible indicators and controls for some time. Honeywell has already announced that is working on "touchscreen-enabled avionics for the next-generation air transport and business aviation cockpits" [1]. I also believe SpaceX runs a largely virtualised command centre.

>Self-driving cars...every single innovation in the car industry that does not go towards electric self driving cars is just useless fluff at this point

Imagine a bond that pays, into perpetuity, 100% of the profits earned by auto manufacturers worldwide on non-self-driving cars. How much would you be willing to pay for this bond? Likely a non-trivial amount.

Wanting to get to Mars doesn't make a Moon mission "useless fluff". The incremental steps add to our learning and help finance, even if indirectly, more ambitious goals. One could interpret Apple's move as an incremental step towards a future where a car's value is inside the cabin, not under the bonnet.

[1] http://www.aviationweek.com/Article.aspx?id=/article-xml/avd...


Granted, the things that touchscreens replace in avionics are the stuff that you need your eyes on to operate in any case (navigation computer etc). And they often replace input systems that were far, far worse (read http://philip.greenspun.com/flying/garmin-gps for a funny review of the old Garmin avionics).

And in the air, you have a little more time to look at the instruments without risking crashing into a meeting car or driving off the road. I suspect (hope) that the current trend with touchscreens in cars will be a fad, and that manufacturers will realize that some features (climate control etc) are best implemented with traditional controls.

It's sad when my 10 year old car has a UI on that part that allows me to use it without taking the eyes off the road.


Having previously worked in avionics I'm a little annoyed that many of the things he complains about are a combination of overstrict FAA rules and the immense cost required for certification for some of the functionality he wants. For instance, you can't warn for terrain unless you're level A certified on almost all airframes - and there's a differences between TAWS A, which lets you warn with "Terrain - Pull Up" and TAWS B, where you can only warn with "Terrain". Approach plates require level B certification as well, although for some reason the FAA lets some people get away with releasing them for iPads with no certification whatsoever (they generally have a disclaimer about "not for navigation purposes" but the FAA used to frown upon that sort of thing).

There was a collision on a runway a while ago that we passed around the news article for, because it could have very easily been prevented by features we had, but were not certified for the airframes that had the collision. It's understandable that we want to ensure that the new glass works just as well as old instruments, but it's frustrating when customers very clearly want functionality that we're not allowed to provide.


This is why experimental airplanes often have far superior avionics these days: you can put anything you want in one, without any certification. The result is cheaper and better alternatives.

I believe there are efforts underway to change the rules to avoid this "making the perfect the enemy of the good" effect: http://en.wikipedia.org/wiki/Small_Airplane_Revitalization_A...


That's interesting. Not sure if it's a "jurisdiction" thing, as in if it's a part of the airplane, they go into the certification regime?

For instance, laptop PCs running Windows and iPads are certified for EFB use now (I would assume for approach plates as well), but I'm not sure what the limitations are, or what certification level that is.

Btw, what did you work on in avionics, and what made you move to a different area of work?


Cars have a greater tendency to crash into other things than planes.


Rockwell Collins (big avionics maker) has a good touch screen product line that has been gaining traction in the commercial aviation market. Rockwell Collins Pro Line Fusion: http://www.rockwellcollins.com/prolinefusion/


You ignore the fact that this is meant to be controlled by Siri and, yes, tactile knobs and switches in your car. It says it right there on the linked page. Little to no distraction needed.


You ignore the fact that this is meant to be controlled by Siri

Nothing to improve road safety like elaborate arguments with a speech recognition engine...


C'mon, this is Hacker News. Let's try to look beyond the current state of technology and think about its potential instead.


You'll acclimate to it like you do everything else in life. People learned to drive just fine after coming out of horse drawn carriages and wagons. Don't act like people are not adaptable to innovation and change.


You think Siri is bad, try the ones built into 5 year old cars.


You don't even need to look that far back. Cars being released as new today have horrible speech recognition compared to what you can get with siri/google now.


Too late. The cat is already out of the bag and everyone is already doing all of this anyway, except awkwardly holding their phone up to their face at the same time.


>The benefits on human society at large here are so significant that there is no room for caring about the feelings of greedy old white men.

So insurance, oil and car companies are filled with stodgy, greedy, old men who we are not to care about, but I guess SV is filled with cool, hip, innovative people?

Get your head out of the sand. I'll take the innovation of Ford and Toyota over Facebook and WhatsApp anyday of the week.


How is this the top comment?

'greedy old white men' should justify dead on its own, but it also completely misses the point.


Because 1) he typed a lot and 2) cited some sources 3) early on when the link was submitted. That's instant upvote criteria.


Because it is the only comment.


because patriarchy


> going at speeds the human brain has never evolved to appropriately deal with.

Not to discuss your other points, but who are you to say the human brain has evolved or not to deal with X or Y ? The human brain is actually very capable of dealing with multiple sensory signals, and the fact that most people can drive cars on motorways without crashing in each other's every 2 seconds suggest that the human brain is perfectly up to the task.


He is accidentally technically correct. Saying that "the human brain evolved to do x" implies that evolution has purpose or intention, which is not the case.


> He is accidentally technically correct. Saying that "the human brain evolved to do x" implies that evolution has purpose or intention, which is not the case.

But to say that human behavior evolved to meet a specific environmental requirement isn't a purpose, it's an outcome that allowed the genotype to procreate. It's a gray area and one must be careful speaking of goals, but if one is addressing a specific adaptation to meet a specific environmental requirement, it's all right.


Sure, but then you could apply the statement to anything new enough that there hasn't been anywhere near enough time for adaptations to evolve. The human brain certainly didn't evolve to read books, or do computer programming, or fly airplanes, or play chess. The trick is that the human brain did evolve (according to your interpretation of the phrase) to solve general problems and maintain strong and complex social bonds, which happens to lend itself to many things, including reading books, playing chess, etc.


At the current fatality rate among young drivers I wonder how many years it would take us to notice statistically significant genetic changes in the population that could be attributed to automobiles.


Or, more interesting, a correlation between innumeracy and traffic deaths. It turns out that the distance required to bring a car to a stop (after reaction time) is proportional to the square of the vehicle's speed. This is something that most people don't know, and that young drivers ignore at their peril.

After the driver reacts and presses on the brake pedal, on dry, level pavement, a car requires about 20 feet to stop at 20 miles per hour, 80 feet at 40 miles per hour and 180 feet at 60 miles per hour (all approximate distances).

Simple equation (reaction time not included): y = v^2 / 20

v = velocity MPH

y = stopping distance in feet

Another reason to learn some mathematics and science (it's based on the kinetic energy of the moving car). :)


Is that relevant? Most fatalities are not because the car drove straight into a stationary object directly in front of them on the road. So, total stopping distance doesn't come into play as you describe it. Collisions mostly happen because of the actions of car drivers who are going roughly the same speed as you.

Not saying that speed is a factor, just that stopping distance isn't the thing. Perhaps you should discuss kinetic energy instead?


> Is that relevant?

Only when a misjudgment of stopping distance is a factor, in one way or another. I'll bet that the car driven by actor Paul Walker's friend -- the car that crashed and killed both of them -- was an example where the friend thought he had enough distance for a safe stop. But at the estimated speed -- "more than 100 MPH" -- more than 500 feet would have been required and just wasn't there, before a sharp turn they didn't manage.

> Most fatalities are not because the car drove straight into a stationary object directly in front of them on the road.

Yes, but all such accidents involve cars hitting each other or objects along the road, or rolling over during efforts to stop them, and similar causes. All these examples involve energy management, and in each case, the driver couldn't safely dissipate the car's kinetic energy.

> Collisions mostly happen because of the actions of car drivers who are going roughly the same speed as you.

Not in any number of cases, like crossing over the center divider, or hitting a roadside object, or losing control in bad weather, or fog collisions, etc. etc..

And even if two cars have a relatively low-speed collision while both traveling at high speed, then the problem becomes controlling the cars after the collision and avoiding hitting other cars or obstacles, many of which are are stationary.

> Not saying that speed is a factor, just that stopping distance isn't the thing.

Of course it is. If you want to avoid crashing into another car, you have to maintain a safe distance. If your car is carrying too much kinetic energy, you can't dissipate the energy in advance of getting intimate with the other car, or a tree, or a center divider through which you roll your car uncontrollably.

> Perhaps you should discuss kinetic energy instead?

Say what? How did you miss the kinetic energy equation I included in the post to which you're replying? I even included the words "kinetic energy" to avoid any possible confusion.


Stopping distance is extremely important. Here are a few examples of typical car accidents:

- a driver in foggy weather crashes into a traffic jam

- a car crosses an intersection at a red light, is hit by a car from the other direction

- car hits a child running on the street

- driver looses control, hits a tree at 60mph

- reckless driver tries to overtake, hits car coming their way

These types of accidents are the most lethal, and they are all due to stopping distance (and could be avoided by driving at a lower speed)


> - a driver in foggy weather crashes into a traffic jam

This is a particularly dangerous and often misunderstood case, because staying several car lengths away from the car ahead won't do any good if that car suddenly collides with a stationary car in front of it that's hidden by the fog. When the car you're following comes to a sudden stop, all the rules about safe following distance go right out the window. This is why professionals sometimes recommend just getting off the road when fog prevents a reasonable amount of visibility.

> These types of accidents are the most lethal, and they are all due to stopping distance (and could be avoided by driving at a lower speed)

Very true.


No. Evolution is just a phenomenon, but its successful products, such as brains, have purposes. Eyes are there for seeing. They came about, in a sense, by coincidence, but they don't work by coincidence. There's no conceptual error in saying "a brain evolved to do X".


Products do not have purposes. Eyes can be for seeing, but can also be for eating or throwing at things. Atoms don't have tiny XML tags describing the purpose of an object they're part of. It's the user who attributes purpose to a tool.

You can't say "brain evolved to do X", you can at best say "brain is well suited to do X".


You are trying to make it harder to communicate for no benefit. The distinction you're advocating is not useful, to anyone. But the idea "the purpose of eyes is seeing" is quite productive, and faultlessly grounded in reality.


His comment makes out like everything in nature is restricted to 15mph.


Looks like you forgot to scroll down.

The second thing on the page:

Control with a word. Or a touch. Or a twist.

CarPlay features Siri voice control and is specially designed for driving scenarios. It also works with your car’s controls — knobs, buttons, or touchscreen. And the apps you want to use in the car have been reimagined, so you can use them while your eyes and hands stay where they belong.


> It also works with your car’s controls — knobs, buttons, or touchscreen

It'll work with the knobs/buttons in the cars that have them and integrate them with whatever functionality they have (I have an old boss who had a Porsche Cayenne with a large screen controlled entirely with side buttons and a small scroll thingy, which was super awkward as the device's interface seemed designed with a touch screen in mind), but it seems that the car industry is largely moving towards touch screens in cars, with brands like Tesla leading the movement. I wouldn't be surprised if innovations like these encourage an even faster move towards touch screens in cars- Apple sure loves its touch screens.


Do you seriously think we'll have widespread self-driving cars — and not just the bare bones technology, but ones that we can ride in without any control at all, even while intoxicated — within a timespan short enough for every other advancement in car technology to not matter?

I am almost 100% certain that we won't see that kind of thing even within the next few decades, if ever. Self-driving cars have way too many hurdles to overcome before they reach that point. I mean, Europe hasn't even moved over to automatic transmission yet.


I mean, Europe hasn't even moved over to automatic transmission yet.

That's because automatic transmission is the worst and invented by the devil.

More seriously, using an automatic transmission results in trade-offs, for instance fuel usage was the historical number 1 sticking point with automatics guzzling up the petrol compared to manuals. Modern engineering has closed that gap but there are still issues related to control - ironically a move to fully autonomous cars would solve most of them (principally in non-autonomous cars drivers in manual transmission cars can use knowledge of future road conditions to make gear selection decisions that an automatic transmission cannot make as they have only a limited input set)


I used to think automatics were horrible (I live in the UK, drove manual for 15 years) then I got one...

In terms of fuel economy, isn't it just a matter of time before automatics become as economical (then probably more economical) than manuals? Even without anticipating future events? Driving efficiently feels like something a computer could optimise for and that a human has no natural advantage doing.


Without being able to determine future road conditions then automatics can't guarantee being in the "right" gear for the up coming section of road. Be that due to traffic, grading, weather, banking, sharpness of corner etc.

My go to example is the insanity of junction 15 of the m8, that loop onto the west bound motorway is a 270 degree corkscrew rising 30 feet with a diameter of 55 yards onto a 5 lane motorway where I need to go from the left hand lane to the right hand lane (as there are entrance and exit lanes on both the inside and outside of the motorway at that point). A current automatic engine wouldn't know what was the right thing to do in that situation ( https://www.google.co.uk/maps/@55.8672951,-4.2358853,374m/da... )

To determine all of that you'd basically be at the level of self-driving car. Which would be awesome.


>> Without being able to determine future road conditions then automatics can't guarantee being in the "right" gear for the up coming section of road. Be that due to traffic, grading, weather, banking, sharpness of corner etc.

Latest Rolls-Royce Wraith does exactly that - it uses the GPS map data to change gears in advance as it is approaching a turn in the road. I am sure that this technology will be available in ordinary cars soon enough.


An augmented version of this technology (applying known map data plus sensor data) is a huge piece of the self-driving car.


Or just buy a Nissan Versa today: http://www.edmunds.com/fuel-economy/five-myths-about-stick-s...

Manuals have better fuel usage then automatics is a myth for recent cars, no fancy self driving needed.


I just purchased a Toyota Yaris. The official figures for the 1.3L engine give 5.7L/100km for manual and 6.3L/100km for the automatic.

Now in a high-end car with a sequential gearbox this is different, but for a small town car when you'll be doing a lot of starts and stops you'll see the losses on the hydraulic coupling.


That's based on the official figures. No-one gets the listed MPG ratings in real life and automatics suffer worst than manuals for under-shooting the official mpg rating.


Your GPS already knows where you want to go, what the road ahead looks like and often even the current traffic situation. I imagine writing some semi-smart software that hooks up that info to the automatic gearbox to give it hints about what to do wouldn't be outside the realm of possibility and no doubt a lot easier than self-driving.


There are cars which already do exactly that: http://www.dvice.com/2013-3-7/new-rolls-royce-looks-ahead-gp...


I'm a die-hard manual fan, so you have my sympathy here. That said, I think current automatic technology is at the point where it can handle your go-to example with no problem. Yes the responsiveness of your base-level Toyota/Honda/Ford automatic transmission is going to be lacking versus a manual, but step up to a Mercedes or BMW double clutch automatic gearbox, or just go straight to the top of the automatic heap with Porsche's PDK transmission, and I think you will find all of them quite performant, even without paddle input.

To preempt the "but you have to own a porsche for that!" responses, I'm not saying this is cheap, I'm just saying it is not limited by the available technology, just cost.


I have a feeling being able to determine immediate near term road conditions is a 1.0 feature for self-driving vehicles?

Waze-style technology + GIS + networking with other vehicles + weather data (if you look, there is already tons of small weather monitoring stations all over major). thoroughfares).

I mean many vehicles within the last handful of years have automatic traction control that can react faster to a real time change in road conditions than any human could ever hope too. A lot of the stuff I mentioned above is just extending that foresight out further so as to not be so jarring.


It pretty much has already happened. Most cars that offer both versions are very similar in terms of mileage for automatics and manuals - the difference is usually only like 1 or 2 mpg, with the automatics actually coming out on top at times because the car makers can add more gears. And realistically, most drivers aren't as good at using manual transmissions as they like to think they are, which also does a lot to cancel out the theoretical improvements a manual transmission offers.


A traditional automatic with a torque converter will always be less efficient I but the newer dual clutch models can probably match (or even exceed through better gear selection) the efficiency of manuals.


You'd think so but I get 7mpg more than the best rated Toyota hybrid which is supposedly the most efficient thing out there.

Bear in mind I drive a 6 year old 1.3L diesel Fiat van converted to a people carrier which is as aerodynamic as a shed. It also has 3 children, a buggy and a pile of crap in it.

Perhaps a computer could optimise this better but I doubt it. The computer and automatic box just goes when you push the pedal down. The driver isn't connected to it and feeling what it's doing and understanding the power curves.


> You'd think so but I get 7mpg more than the best rated Toyota hybrid which is supposedly the most efficient thing out there.

You sure about that? Because I easily get about 70 mpg on my Prius (C class) when I'm supermiling in the city. Granted it's not as efficient on the highway, and unfortunately I do have to do a lot of highway travel. I can sustain a good a 55mpg average at the end of the month though. I plan to start borrowing my wife's whenever I'm on the highway, I think this way I can average mpg in the 70s overall when I do this.


I get 58 solid. Prius is quoted at 50 here in the UK. If I'm on a flat road at 40-50mph (north circular) I can get a solid 80mpg average. Just have to stay easy on the throttle and not change speed.


You do realize that you're comparing a diesel car to a standard petrol-hybrid right? Diesel engines are ~33% more efficient per-liter (and usually that much more expensive).

The number for a Prius-hybrid you'd really care to look at would be the city driving numbers as that's where you get all the advantages of regenerative breaking. Running a Prius at 100+ km/h it just turns back into a standard petrol fueled vehicle carrying drained batteries.


Yes. It's about the end game, not the tools.

I've found diesel engines and vehicles to be considerably more reliable and therefore less expensive. Fuel is between 5-7% more expensive here in the UK so things are good there.

The city driving numbers are about the same I found. I regularly drive right into London's congestion charging zone - at least 1-2 times a week.

51 is the expected MPG in city driving scenarios. I can achieve that with no particular effort in London traffic. I don't drive like a crazy person and hang back regularly - that is all.

My vehicle doesn't have piles of expensive electronics, batteries and things to go wrong either (look at recent threads about Toyota stack overflows for example).

The most expensive thing on a Diesel that can go wrong is the particulate filter, EGR valves and timing chains and these are cheaply regenerated and replaced respectively.


Not really, an automatic transmission always adds some amount of weight over a manual transmission.

And this is why electrics are eventually going to win. A motor optimized for a single RPM speed driving a generator doesn't need a transmission. Even if the rest of the drive chain was as efficient, the loss of weight would drive the change over time.


For at least 15 years or so, that's just cargo cult BS Europeans who haven't tried automatic transmission cars say.


I drove a new Vauxhall Corsa with automatic transmission a couple of months ago. It was the second worst driving experience of my life.


What an utterly un-useful statement without any detail. Why was it a bad driving experience? What vehicles are ranked first and third?


Very low cost automatic transmissions are terrible - the gap between gear shifts goes on for ages. The Vauxhall Corsa is an example of this; for the American market check out a Smart Car.

Is this really a good argument against automatics though? The worst example of anything is usually going to be pretty bad.

As a matter of current economics, yes, if you have a very tight budget and want a good driving experience, manual is going to be better. But automatics are as fast and as efficient as manuals a little higher up the market. And it will trickle down soon enough.


Also worth noting that most of the modern super-cars don't even have a true manual (that is, a stick-shift and a user-operated clutch). They're mostly electronically controlled dual-clutch setups since a computer can shift faster than any human can. The 2015 BMW M3 was even rumored to not have a manual option at all, and it's known for being a driver's car. I believe they're keeping it around for now, but there's some worry that the clutch/stick-shift arrangement is going away in the future because the electronically controlled transmissions are that much faster.


Well, I find the Vauxhall Corsa with manual transmission is lovely to drive so I am comparing the experience between the two. So sure, I might not have been looking at the best-of-the-best when it came to automatic gears but I was comparing like-for-like.

Maybe in another 10 years the automatic experience at my price range will be better but at that point we're 25 years on from coldtea's period where European cargo-cult bullshit started.

Apparently.


How much of it was placebo? How much of it was "feeling out of control"?

"second worst driving experience" has nothing quantifiable.


I had absolutely no preconceptions going into using the automatic - I would even go so far as to say I was looking forward to it as I'd never driven an automatic before.

I would think my number 1 complaint about the driving experience was the occasional (but not predictable) incredible lag between pressing the accelerator and the car going faster and the ridiculous over-revving when I presume the car got confused about what gear it thought I should be in.


>I would think my number 1 complaint about the driving experience was the occasional (but not predictable) incredible lag between pressing the accelerator and the car going faster and the ridiculous over-revving when I presume the car got confused about what gear it thought I should be in.

I've never seen the lag (much less "incredible") in any model I've known.

I've indeed seen the over-revving sometimes. That said, how is the over-revving any issue compared to the whole not having to even worry about manual gear shifting benefit?

It's like going to the US on a boat, because airplane food is bad. Yes, it might be, but it misses the whole point of getting there faster, easier, safer and with less concerns.


> I've never seen the lag (much less "incredible") in any model I've known.

Check out a Smart fortwo on car2go sometime and be amazed :)


Driving a car with a cheap or slow automatic transmission is a bit of an art - after a while you start knowing when and how to floor the accelerator just to tell the car that you'll need it to shift up in a couple of seconds.

It feels somewhat counter-intuitive as normally you press the pedal to speed up, but when going uphill you sometimes have to really step on it just to maintain speed.


Yeah dude it's totally because we haven't tried them. It's not because they're 'intelligent' pieces of shit with virtually no control or anything.


While in motion, a driver should maintain 100% of their focus on two things: direction and speed. Having to juggle another pedal and lever to make sure upcoming speed change would be suitable to the driver is unnecessary and a distraction. Modern cars perform gear changes good enough, at the worst. If you have a CVT, it is even better (but very expensive to buy/repair, I acknowledge that).


Changing gears manually is you paying a lot of attention to speed and direction. Once the first part of getting comfortable changing gears is past, you as the driver change gears based on paying attention to engine noise, speed, driving conditions, other vehicles.

You end up paying a lot more attention to everything, vs being largely on autopilot with automatic transmission. Or at least having the option of being a lot more on auto pilot with auto transmission.

Of course this varies driver to driver, but having driven both types for about 10 years, I personally find mysel way more engaged with the manual vehicles.


>It's not because they're 'intelligent' pieces of shit with virtually no control or anything.

Really, and you need control of what and for what?

Anything quantifiable?


You should try a proper DSG gearbox, they will change your mind and they're even faster and more economical than manual gearboxes.

I have one in my VW Golf!


But DSG isn't a traditional auto. It's still clutch based, you just don't control the clutch yourself.


Automatic transmissions have a clutch that's operated automatically. DSGs have two clutches for quicker gear changes. That's the only major difference.


Standard automatic transmissions don't have a clutch disc. They have a torque converter.


Virtually all modern automatic transmissions use a "lockup torque converter" for fuel efficiency. The lock-up clutch is activated once the transmission is fully coupled, and at that point the efficiency of the transmission is essentially identical to that of a manual transmission.

http://en.wikipedia.org/wiki/Torque_converter#Lock-up_torque...


Oh how I wish this were the language we all adopted.


> I am almost 100% certain that we won't see that kind of thing [self-driving cars] even within the next few decades, if ever.

Come on -- self-driving cars are already legal in California --

"Self-driving cars now legal in California" : http://www.cnn.com/2012/09/25/tech/innovation/self-driving-c...

-- and the military are testing self-driving trucks to prevent IUD casualties in hostile areas --

"Army tests self-driving trucks for the next big war: " http://autos.yahoo.com/blogs/motoramic/army-tests-self-drivi...

-- so your doubts are anachronistic to say the least.

> Self-driving cars have way too many hurdles to overcome before they reach that point. I mean, Europe hasn't even moved over to automatic transmission yet.

Hey -- wake up and smell the Cappuccino. :)


As I understand it current self driving car technology fails whenever the situation becomes sufficiently complex (eg. icy/wet roads, some other unexpected traffic situation) and require the driver to take control at this point. My inference from that was that the 'simple' problems have been solved, but to have technology that can safely drive an intoxicated person home would require solving another set of much harder problems. Ie. the ninety-ninety rule.


> My inference from that was that the 'simple' problems have been solved, but to have technology that can safely drive an intoxicated person home would require solving another set of much harder problems.

But they already have cars that drive themselves, without human interference. Google (and other companies') engineers have been driving them all over the country on test runs. Obviously if the car will drive itself to a specified destination autonomously, the physical state of the occupant is no longer an issue.

> Ie. the ninety-ninety rule.

Yes, fair enough -- I'm sure there are still edge cases, but the OP said, with respect to this technology, "... we won't see that kind of thing even within the next few decades, if ever."

The turning point will be when the technology's edge cases are less dangerous than a human driver's edge cases. That won't take very long.


But they already have cars that drive themselves, without human interference. Google (and other companies') engineers have been driving them all over the country on test runs. Obviously if the car will drive itself to a specified destination autonomously, the physical state of the occupant is no longer an issue.

The physical state of the occupant is an issue if there are situations where the occupant might need to regain control of the vehicle. Unless the autonomous car can handle all potential situations the occupant will still need to be in a fit state to drive the vehicle.


> The physical state of the occupant is an issue if there are situations where the occupant might need to regain control of the vehicle.

All I can say is that the goal of self-driving cars is to take the occupant out of the picture entirely. I think a lot of public safety officials will welcome this change. Eventually, there will be two classes of vehicle occupants -- those who are qualified to touch the controls, and those who aren't.

For an earlier social model, look at elevators. At one time, to have an elevator in your building, you had to hire an elevator operator. Then there were elevators that the building's occupants could operate if they were careful and knew what to avoid. Now elevators are safely operated by anyone.

Same idea.


  > when the technology's edge cases are less dangerous than a human driver's edge
  > cases. That won't take very long.
Just a couple of decades.


A vanishingly small amount of time, really.


That seems about right. For American drivers, maybe a bit too long. :)


It seems likely to me that automated cars could do the following with incremental, predictable progress: 1) Drive with good road conditions on pavement in areas that have accurate maps. 2) Recognize when conditions change and the car has a lower chance of functioning correctly. 3) When 2 occurs, pull over and yield to the human, or turn around and go back to another location.

This isn't the fantasy of "get wasted, get driven home, no matter what," but it seems good enough to ensure people are safe. Maybe you can get home, or maybe you sleep it off on the side of the road, or maybe your car drives you to a motel.


Humans fail regularly even when the situation is not especially complex! e.g., not paying attention and rear-ending someone at lights on a straight, dry road during the day.


That's of course true. However if a human driver causes injury through error he might expect to be held liable. If an autonomous vehicle responds incorrectly to an unanticipated driving scenario and causes injury, the manufacturer will likely be subject to a lawsuit.


And if you thought the "sudden unexpected acceleration" hysteria recently was bad, just wait for the first time someone dies in or because of a robocar


In Europe, virtually every brand and model of car can be bought with an automatic gearbox as well as manual. So it's not that they (the manufacturers) have not moved over yet - it is just that they (the drivers) generally prefer manual.


That is by choice. I live in the UK. Personally, I don't see the point in driving manual, but most drivers in the UK enjoy using stick shift. In the context of a self-driving car, that doesn't matter.

The self-driving car technology is also very viable to use right now. I have seen Google self-driving cars on the freeways driving next to me. It would not surprise me if they were being sold within the next 5 years, commonplace within 20-30 years.


I personally just love driving a manual car. Its not for any practical reason, I just get a kick out of shifting perfectly, the feel of the engine and the control that I have.

That said, if there's a self driving car, you bet Id rather use that.


>> I just get a kick out of shifting perfectly, the feel of the engine and the control that I have

Funny how I feel exactly same way when I drive an automatic.


"Europe hasn't even moved over to automatic transmission yet"

After having a couple of cars with automatic gearboxes (including a 4.4L BMW - so lack of power wasn't the problem) I find that I am much happier with a manual gearbox.

NB I'm not resistant to new technology in cars - my current car has automatic parking and I use it a lot and normally leave the lights on auto as well.

[Note I am in the UK]


second. I don't know what it is with it exactly, but automatic shifting kills any pleasure in driving I still have. All the ones I tried seem to react to pedal movements too slowly, which I find sometimes even dangerous because I'm not used too it, and I cannot use the engine to do light to intermediate braking anymore.

and normally leave the lights on auto as well.

Why don't you just leave them always on anyway? Or is that not allowed in the UK?


It used to freak people out, others would flash their headlights at you if you had them on in the daytime. Now all new cars automatically put sidelights on during daytime anyway, I think its a (pretty sensible) EU regulation.


Down-shifting puts extra stress on transmission and engine components which are far harder to replace than brake pads.

If you're heading down a long slope, sure, use the lower gears to add control.

If you're just trying to come down from your amazing commute freeway race into the clover leaf... use the brakes for braking. Most people downshift incorrectly and add unneeded wear to their clutch.

There's few enough sticks out there already... don't leave me one with a beat up clutch.


Irrelevant amounts of so. Any driving school teaches using engine braking for increased fuel efficiency and decreased wear of brakes. Manual clutches still last for far longer than even high-end automatic ones.


  > I mean, Europe hasn't even moved over to automatic
  > transmission yet.
For the good reasons.


> I mean, Europe hasn't even moved over to automatic transmission yet.

Haha. 1) Driving automatic cars is so boring. 2) it makes your body unbalanced constantly since you use a single foot, therefore your back will ache over long periods of driving. 3) You cannot adjust your speed as fast with an automatic car, we are stuck with binary controls while manual cars provide all the flexibility you need in real, and critical driving situations. All serious cars (like sports cars) are by far usually manual.


I disagree with both points so much it hurts. Driving automatics is fun, especially if you are in a car with a massive and a powerful engine - the sensation of it dropping just to the right gear,and then holding up to just the right amount of RPMs is exciting, something that only pro drivers can achieve in cars with manual transmission.

As for the second point - in critical situations on the road, I would rather keep my eyes on the road and both hands firmly on the steering wheel - not manipulate gears and clutch.

And I also disagree with the notion that automatic cars don't give you control - with enough experience,you can force precisely the gear you want in an automatic by operating the throttle in a right way. They actually give you way more control - if I find myself at a need to suddenly accelerate, I just press the throttle all the way down and then the transmission just drops down two gears,or one, and does it in less than a second. I am pretty sure that 90% of drivers would not be able to do is as quickly/smoothly/efficiently in a manual, selecting the right gear, declutching and starting at the right rpm - automatic transmission does that for you.

I never had an automatic stall, they don't do jarring shifts between gears, and my left leg is not about to fall off from holding the clutch in a slowly moving traffic jam.

And "serious" cars like sport cars mostly come with an automatic transmission without the manual as an option. You know why Bugatti Veyron comes with an automatic? Because no human would be ever able to change gears as quickly as needed with that car.

And interestingly, the entire off-road market is moving towards automatics as well. You know why? Because you can't burn out a hydrokinetic clutch while starting uphill or in difficult conditions.

I've driven many many cars in my life, and while there are situations where I can see how the manual transmission is "enjoyable", automatics are just objectively better, with new ones offering lower fuel consumption, better acceleration, smoother gear changes and relative lack of maintenance in comparison to manual transmissions.


All but one of the cars I've owned have had a manual transmission, and I bought a 2013 GTS with a stick (even though the PDK-equipped model is faster 0-60, and certainly more fuel efficient).

But I reluctantly disagree with you. Almost all serious sports cars -- hardly "boring" -- now come with automatic transmissions: * All Ferraris are now automatic * All McLarens are now automatic * Porsche's top street car, the 911 Turbo S, is PDK-only * Porsche's top track-ready car, the 911 GT3, is PDK-only * The fastest hypercars, the LaFerrari, the McLaren P1, and the Porsche 918, are all automatic (I recall all 7-speed dual clutch)

Even Lotus, the one company you'd expect to keep the manual faith, on lightness grounds if nothing else, made the Evora available a few years ago with an automatic transmission. And Porsche has been quoted as saying the manual transmission may no longer be available after the 991 is phased out: http://www.autoguide.com/auto-news/2012/02/porsche-911-manua...

Unfortunately manual transmissions now make the car slower. A stick is becoming an enthusiast's conceit, C&D's campaign notwithstanding, and I hope they're still offered, but I'm now considering a dual clutch automatic for my next car.


It should be noted that there's a difference between the old-tech viscous coupling approach and the cutting edge dual-clutch sequential transmission.

The old-tech viscous coupling was lossy and difficult to control. Even when you can manually tell the trans to shift (as on a 2009 Mazda5), there is a time delay in the shifting, then you wait for the hydraulic system to confer torque and finally drive the wheels/engine brake.

The modern DCT that high-end cars are using is a direct mechanical connection and uses robotic shifting to do so instantaneously. Shifting is faster than a human can shift an H-gate, and the PID controls on gear selection have better programming etc. due to both the high end nature of the cars and the better systems they're being integrated into.

There is a world of difference between what we colloquially term an "automatic" (slushbox torque converter) and the modern DCT.


Well said! And a good explanation. Though I'm not sure if the DCT can properly be termed "cutting edge" anymore; Porsche's PDK started appearing in 911s that went on sale six years ago (2009 model year, on sale in 2008).


Well driving with an automatic in London traffic jam is much more comfortable than not being able to shift beyond 2nd gear ;-)

I have a Golf with a 7 speed DSG gearbox and trust me, in sport mode, you can have amazing and smooth acceleration.


I can confirm that the 7 speed DSG gearbox in the Golf will get you up to speed at an amazing and smooth pace!


Regarding point 2, long freeway drives will still be just as unbalanced with either transmission, or if you have cruise control, more balanced.

Regarding point 3, is that really true to an extent that would actually matter for ordinary driving?


In europe, we don't have long freeways. I can drive from one end of my country to the other in a day, and most people don't ever do such long drives.


Tiptronic, best of both worlds.

(Apart from the back ache)


Note: Europe doesn't spectacularly drive through shop windows, through petrol stations and car parks at least 5 times a day though (source: live leak, youtube etc)


I think you'll be surprised at how quickly this type of thing is evolving. Google is putting an immense effort into solving this problem. I don't see self-drive cars as a solution for rural areas, but I do see a massive demand for cars in a city, replacing taxis and car clubs.

I can even see it being a sort of a 'mini-bus service', for example, noting that you and 4 others are doing a pickup in an area and doing a school drop off, offering to split the cost of the journey.

The point is that individual car ownership doesn't really 'work' in a city too well.

Getting to the point where this is socially acceptable shouldn't take too long. The reduction in cost base and the fact you can sit in the car and work/play should make it very attractive.


I'm from Nebraska originally, and I see a bigger market for self-driving cars in flyover country than in an area like SF Bay.

Lots of long, straight drives with far less vehicles on the road probably even make it more realistic to accomplish bc it's a less dynamic environment. I know 20+ who drive an hour both ways to work...and on the morning drive you can do it and only see a handful of cars if it's early.


I have a car with Adaptive Cruise Control (it has a fricking radar on it's head) and it rocks for long straight roads.

Add adaptive lane assist and your half way there ;)


This is definitely a cultural difference.

In many countries in Europe driving a car is considered a pleasure and a skill.

Driving manual vs. automatic is considered the same as playing a guitar vs. playing Guitar Hero.


Wonder if that has something to do with size of America relative to most countries in Europe. Americans just simply have to drive more and for longer distances. Driving doesn't seem so fun in the midwest.


Well, car enthusiasts here in the US also consider driving a pleasure and a skill, and I too vastly prefer shifting myself when driving at speed. On the other hand, I find driving in heavy traffic with a manual transmission to be incredibly tedious. That's why a dual-clutch transmission is my favorite - manual performance and control when I want it, automatic convenience when I don't.


I think your gut-check here will turn out to be inaccurate. We will have self-driving cars, and they will come soon. Ultimately, you need:

1. Cars capable of making decisions real-time, on the road, given unpredictable stimuli 2. A network of things, between the cars on the road, to allow optimization and homogeneity of traffic 3. Regulatory environments permissive of autonomous vehicles.

All of this is well under way and I'd say conservatively we'll see robust networks in less than a decade.


> All of this is well under way and I'd say conservatively we'll see robust networks in less than a decade.

I'd say you are being quite optimistic here. Such major changes will be heavily monitored and regulated, and regulations will make it very hard to market any of these vehicles before a long time. 10 years go by pretty fast, I don't think changes are going to that fast and radical. More likely in the scale of 15 to 30 years. But we'll see.


Sage advice from Mr. Gates in his book "The Road Ahead":

"We always overestimate the change that will occur in the next two years and underestimate the change that will occur in the next 10. Don't let yourself be lulled into inaction."

There are myriad examples of the amazing leaps our civilization makes in a decade, particularly in technology.

I really don't think my prior post is unduly optimistic, given the pace of the last decade in the auto industry, the continual convergence of the various problem spaces, and the incentives to be the first-to-market with the remaining building blocks.


Think about them being allowed in private areas (theme parks) or certain districts (Vegas or a city's CBD, etc). That will ease things in against those stonewalling it all.


> Do you seriously think we'll have widespread self-driving cars — and not just the bare bones technology, but ones that we can ride in without any control at all, even while intoxicated — within a timespan short enough for every other advancement in car technology to not matter?

Yes. Without question. It's a process that has an exponential adoption curve to it.

The benefits from self-driving cars are so enormous its ridiculous.

Productivity goes up, so for any entrepreneur, a self-driving car will become a necessity (no more wasting an hour of unproductive work to get to the airport, for example). Pollution goes down because self-driving cars have to be electric. Self-driving cars can go recharge themselves at a nearby central charging station instead of requiring expensive charging stations at every company. Parking can be decentralized (somewhat, eventually you have to cope with self-driving cars causing a mini rush hour when they come get their passengers). You don't need "long term" parking anywhere. You can take your bike to work, but call your car in if it rains.

It might not cause the revolution that the Internet has, but it's gonna be damn close.


> It might not cause the revolution that the Internet has, but it's gonna be damn close.

Ubiquitous autonomous roving robots will lead to a revolution at least as big as the Internet, but probably bigger.

Like the printing press, the Internet made distances shorter, but in a low-fidelity way since reality has much higher bandwidth. With self-driving cars, suddenly “real” distances become smaller since 12 hour trips become routine†. The differences between, say, a middle-class Oklahoman, Texan, Missourian, Coloradian, et.al. start to disappear, and this scenario will play out all across the U.S. and other large regions (Canada? India?).

†If we raise the speed limit to 120 mph (mostly afforded by a change in energy policy), then I could trek from Oklahoma to NY or LA every weekend.


For some strange reason it is users in Europe that prefer manual transmission, since basically every car is available also with automatic transmission also.

If anybody cares, here is my theory (at least for Germany, where I am from): Driving lessons are very expensive in Germany and you have to decide if you want to learn automatic or manual transmission. If you learn on automatic transmission, you are only allowed to drive automatic cars. If you learn on a manual, you are allowed to drive both. The prices for the drivers license are the same. So, most people will learn to drive stick. Since they are then only used to driving stick, they are actually more comfortable with than with automatic. I have met quite a few people who I had to explain how to drive an automatic (it's obviously super easy if you already know how to drive a manual). So, I think it's want you learned first that you will stick with.


> For some strange reason it is users in Europe that prefer manual transmission, since basically every car is available also with automatic transmission also.

The more likely cause is that gas in Europe is far far more expensive than it is here, and manual transmissions are historically more efficient than automatics (may have changed with hybrid e-CVT and newer drivetrains).

Also possibly an explanation is that engines in Europe are far smaller than here (to improve gas mileage) and manual transmissions gives the ability for a driver to play with the torque curve and delay shifts to improve immediate speed/responsiveness.


>Europe hasn't even moved over to automatic transmission yet.

I wonder if there's been a study that looks at the safety of manual vs. automatic transmission vehicles. This coming from a guy who made it a pre-req that his future wife drive stick...


Basic slow models are availble now , so that's a good sign:

http://ideas.4brad.com/navias-induct-officially-sale-250-000


The three cars I've owned in my lifetime all had 90% physical controls, with a small little digital screen that would display something like the radio station you're on, the temp, the strength of air, treble/bass, etc, etc.

There was absolutely zero tactile feedback on all of them, as their were dedicated buttons that had to be cycled through to reach a state....and the only way you could know if you were getting to what you wanted was if you looked at the tiny little digital display.

Want preset radio station #x? Jam the preset button a few times.

Want to warm the passenger side a bit? Jam a button to select that side, then select heat, the turn it up.

That's on a Honda, a Cadillac & a Toyota spanning a decade in years....so I'm guessing you must be referring to pre-1990 vehicles that literally had dial tuners and manual button to control valves...and 5x less options for stuff that could even be adjusted?? Or did you speak too soon?


http://ipocars.com/imgs/a/b/n/u/s/citroen__citron_citron_ber...

Here's an example of a car from ca. 2005 that has simple, manual controls for climate control. If the passenger is too warm, there's a separate button for opening the window :)

Some newer cars also have separate dials for heat on each side: http://static.autoexpress.co.uk/sites/autoexpressuk/files/st... ..but of course they still have pushbuttons for fan speed.


This is a fundamental mistake of UI design that will just keep being repeated forever in favor of "clean" design.

The word for what is missing from these interfaces is "affordances": http://en.wikipedia.org/wiki/Affordance


My 2002 Toyota Corolla might (does) have relatively few options for what you can adjust, but it certainly doesn't have cycling buttons requiring you to look at a tiny digital display. It doesn't have a display at all. The buttons are all labeled with what they do. The climate control is dials. Want to warm the passenger side a bit? Your option is to warm the whole car. Set the temperature dial to warm and the mode dial to (most likely) floor.

Your pre-1990 estimate is way, way, embarrassingly off, at best.


Hell, my 2013 Subaru BRZ has dials and buttons for the A/C. Granted, there's a display that shows you the temperature, but if you want it warmer, you don't have to look at it unless you really want to set it to a specific temperature. If you want it warmer, just turn it a few clicks to the right. Same goes for the passenger (it's dual-zone). There are loads of 2014s out there that don't have the automatic climate control, even. It's the same, old system that you see on cars from pre-1990.

Contrast this to the sound system, which only has a volume knob. Jumping to the next track requires you to hit a small touchscreen button that offers no tactile feedback. It's frustrating to do it while driving.

IMO, there's a good middle-ground here. There are a lot of settings and features on my sound/navigation system that I don't need to access all the time, or really at all while driving. Bass is too high? I can wait until I come to a light. That stuff can remain touchscreen driven since it offers more flexibility in the design and allows you to build in a lot more features than you'd have otherwise without cluttering your dash with a ton of buttons and knobs. But stuff that I mess with pretty often, like skipping to the next song, should be a physical button. Other head units get this right.


My stepfather is a pilot.

| but why do you think planes and other complex machinery have stuck with physical controls

They haven't. Most new planes and retrofits are moving to touchscreens. It allows the pilot to display only the important information on screen at a given time. Like the airport map, for example.


But the yoke, pedals, etc are not touch screen.


so is driving wheel and brakes


huh? Are you claiming that they are in the case of this car interface?

Also, this: http://en.wikipedia.org/wiki/File:Airbus_A380_cockpit.jpg


I may be a dying bread, but I love driving. I love driving fast cars. I love racing (legitimate racing on legal tracks) and I love the thrill of being one with a powerful machine. Someone said recently that sports cars may well be the last drivable cars to exist, and I'm ok with that. If I'm not in the mood to drive and I just want to get from point a to point b then fine, a self-driving car is awesome, but I get a lot of joy, excitement, and pleasure from driving my car around. I'm not in a big city where it's an issue and there are lots of open roads to go down. I take great pride in how I drive and the example I represent when driving. I take it seriously and I wish that more people would.

So no I don't think that every new innovation is bad and useless. There will always be auto racing and fast cars. There's nothing else like it.


Good job on ignoring the main point of this pitch - that it integrates the phone with in car controls, obviating the need to use the touchscreen when on the move.


>>(the views in this post are a bit on the extreme side- but that's how interesting conversations get started :-)

They are. I know this is an extremely unpopular opinion on HN, but I would like to drive my own car, thank you very much. I don't mind automatic mode on long journeys on the motorway, but other than that, I want to drive myself - and I hope that automatic driving will not be forced upon us during my lifetime.


Thank you. The only times I hate driving are long road trips and at night in the rain.


> Yay, more touch screen in cars.

It's odd though that the number of auto accidents in the US has not increased with the increased use of in-dash touch screens and SmartPhones. Deaths from auto accidents have also been on the decline. It seems like people are doing a pretty good job adapting.

2000 2001 2002 2003 2004 2005 2006 2007 2008 2009

13.4 12.5 18.3 11.8 10.9 10.7 10.4 10.6 10.2 10.8


Population adjusted vehicle miles driven has also decreased during that time period: http://www.businessinsider.com/vehicle-miles-driven-2013-2

There's a lot of factors that go into reducing auto accidents, so that correlation doesn't really tell us anything about the danger of driving with touch screens in the car.


It tells us something, just not something anyone seriously wondered about. We can be pretty sure that vehicles with touchscreens don't get into (for example) 10,000x as many accidents as vehicles with manual controls, as that should be visible in the statistics.

What it doesn't tell us is whether or not touchscreens pose any significant additional danger.


> It seems like people are doing a pretty good job adapting.

Or something else is making it safer


>>why do you think planes and other complex machinery have stuck with physical controls?

They're not, Garmin has offered not only custom touch screen flight decks[0] but they also have an ipad app that can be flown with.[1]

[0]https://buy.garmin.com/en-GB/GB/aviation/flight-decks/g3000-... [1]https://itunes.apple.com/us/app/garmin-pilot/id340917615?mt=...


I don't know what white men have to do with this. I bet a bunch of the people who are opposed to that are middle eastern!


Dude I up voted your comment, but seriously with the white man racecard?


> greedy old white men

Please stop doing this


It depends on each manufacturer's implementation. The ever-growing complexity needs to be managed, both in learning curve as well as overall placement and design of features.

For one, BMW does this right. The large screen is controlled by the iDrive knob in the middle, which rotates with haptic detents, is pushable, and has 4-directions on it. All of the major controls are still present with physical controls: radio volume, presets, dual temp control, fan speed, vents, cruise control, phone, etc. The only time you would regularly use the iDrive is for navigation, which is always going to be kludgy because of the complexity of putting in addresses. I welcome voice and touch here.

http://www.moibibiki.com/images/bmw-4-series-interior-12.jpg


iDrive is terrible and distracting. Never has changing a radio station been so dangerous.


Yes, which is why they give you hard buttons for presets and up/down scan.


Only after 2012 as far as I can tell. I don't know if that finally makes the system sane but it certainly wasn't previously.


Try the Lexus car-mouse. It's far better than the "knob" systems (Mercedes, BMW, Audi). You still get haptic feedback (magnets, how do they work?) and the motion of the controller corresponds directly with the screen display.


I remember seeing that at the local auto show last year. I couldn't help but think that if you see a mouse in the car they did something wrong.


It looks gimmicky. But it's actually very functional.


"but why do you think planes and other complex machinery have stuck with physical controls?"

Mainly because of technology. But more modern systems are doing away with this.

" For operating complex vehicles/apparatus, you just cannot do better than tangible controls."

You're not controlling the vehicle.

" Knobs, switches, sliders can be operated without looking at them while giving rich tactile feedback, they have no modes = 0 risk for confusion, you know where they're going to be located on your dashboard regardless of what you're doing, etc."

Physical controls in airplanes have caused confusion. Wrong switches have been activated by mistake several times (either because someone bumped into them, or switched the one next to the one intended)


I agree it's a bad idea to have touch devices within an arms length of a drivers seat, even if the device can be interacted with also via voice and physical buttons.

It would would be simple to not allow touch interaction while driving. So before you start and if you pull over you can text/call/setup navigation which is easy to do on the touch screen and hard to do with a few buttons.

Once moving the touch screen should dumb down to stupid mode with only one valid interaction: hitting the screen hard a few times should silence the navigation voice (a feature I have found missing in most current models).


Did you even read the article linked to?


We already know how to make transportation exponentially safer; simply don't drive. It's the No. 1 thing you can do to lower your risk of near-term mortality.

I'm not joking around here either. I firmly believe we need safer cars, and self-driving cars would be a great step towards that, but the real issue here is how car-dependent our society is. If more trips could be done on foot or via public transportation, the U.S. would be a much safer place.

Our reliance on car driving makes other forms of transportation unsafe too. The risk of riding a bike is not that you'll fall or hit another bicyclist, but rather that a car will hit you (and probably not even be cited). The biggest risk to walking is that a car will hit you in a cross walk or on a sidewalk (will trying to avoid hitting another car), just because. And we're so cavalier about killing people with cars.

We don't need have to wait (or hope) for technology to save us. We can build better places to live today. Places that are safer and healthier. And that largely starts by making towns and cities more walkable.

When I walk and take the train to work, I can send texts, read news stories, listen to podcasts and check Twitter, all without endangering people's lives. And no futuristic technology is needed to accomplish this.

If society we're a game, I'd be putting my resources to building walking neighborhoods and better public transportation. Even if you have self driving cars, cars are still bad for your health from the lack of exercise and the pollution.


I don't really understand the "greedy old white men" part of your comment... What do you mean specifically?

By the way, touch screens in avionics are gaining a lot of traction in the commercial aviation market. See one such example here: http://www.rockwellcollins.com/prolinefusion/


"white men"? Why malign the minority in tech, people of European ancestry.


I agree that certain dashboard functionality are better handled by knobs, switches, and sliders. However if you look into the webpage above, CarPlay only supports Phone, Music, Maps, and Messages, the kind of features that are better handled by touchscreen. As for Messages, it can read the text messages for you.


Also Podcasts, Beats Radio, iHeartRadio, Spotify, and Stitcher.

I'm curious if Apple will be controlling what apps are allowed to work with CarPlay to keep people from putting millions of Flappy Bird clones on a screen that drivers shouldn't be looking at. My guess would be yes, but all it says for now is "Stay tuned for even more supported apps coming soon."


Touch screen in cars is debatable, but for the record, Apple's solution allows the use of physical knobs and buttons in the car to hook into CarPlay.



A touchscreen isn't tangible?


it is, but it's really limited in kind of tactile feedback it can provide. (haptic feedback).


One word. Siri.


Not that it's any surprise, but it's nice to see an emphasis on Siri. Voice control is very well suited to car interfaces: the car is a relatively-quiet, private space where you can be heard by the interface but not bother a stranger, and voice is as hands-and-eyes-free as it gets (not that this is the only challenge with car interfaces – divided attention from the road will only be solved with full autopilot).

What I would love to see going forward is real consideration for the optimal mix of buttons, touch-gestures, and voice control. A promising design [1,2] is the combination of mode-selection buttons and a thumb touch pad on the steering wheel. Left thumb holds down one of several modes, right thumb makes a simple gesture (no visual attention needed) and boom, you've got a lot of options for controlling things without taking your eyes of the road.

[1] graphic representation: http://i.imgur.com/sHhDLKZ.jpg

[2] pilot study: http://pro.sagepub.com/content/57/1/1643.short


Voice control works fine in Windows Automotive since at least 2008 [1].

Remember both Microsoft and Apple are using the same voice technology and sound libraries from Nuance Communication [2]. You can change the radio, mp3-playlist, do phone calls, listen the SMS, etc., though the internet enabled services are missing in older editions.

[1] http://en.wikipedia.org/wiki/Windows_Embedded_Automotive

[2] http://en.wikipedia.org/wiki/Nuance_Communications#Partnersh....


Are you sure it's the same tech? It very well could be, but scroll down on that page to mergers and acquisitions, Nuance seemingly buys every company it can in that space. So they actually have a wide variety of tech and perhaps deals associated with them that predate acquisitions.


Yes, it's based on former DragonDictate: http://en.wikipedia.org/wiki/DragonDictate

The following product: http://en.wikipedia.org/wiki/NaturallySpeaking

Microsoft Auto uses it http://www.nuance.com/company/news-room/press-releases/ND_00... and http://en.wikipedia.org/wiki/Ford_Sync

And Apple's Siri is Nuance Dragon NaturallySpeaking cloud service.


If I were an auto manufacturer, I'd be pushing Apple to open up the standard for interfacing this with phones. iOS and iPhones are great, but are still only used by a fraction of potential customers. Why limit your market so Apple can score some lock-in?


I believe car manufacturers already have a cross platform standard: http://en.wikipedia.org/wiki/MirrorLink

Carplay seems to be just Apple reinventing the wheel.


Carplay seems to be just Apple reinventing the wheel.

To be fair, that's basically Apple's business models. Just add some proprietary standards and some software patents and you are good to go.


This comment is a bit dismissive. Can you really claim that for iPod, iPhone and iPad?

Sure, every invention is based on work of others, but Apple has still made huge improvements on multiple areas.


This comment is a bit dismissive. Can you really claim that for iPod, iPhone and iPad?

MP3 players were already pretty mainstream when the iPod came around and it didn't really add much. And the iPad was merely an iteration on the iPhone.

But every rule has an exception. I'll grant that Apple definitely redefined what a smartphone was with the iPhone.


I, for one, found the click-wheel to be an innovation strong enough to switch from other MP3 players. I have, and had at the time of the iPhone release, a large music collection, thanks to my interest in jam bands that allow for legal downloads of recordings from their concerts. I tried many mp3 players, both before and after the release of the iPod. None matched the ease or speed of navigation provided by the iPod. There's a reason that, after the release of the iPod, no other MP3 player had more than a tiny sliver of market share. The click wheel was an innovation. Yes, they patented it, so nobody could mimic it. But to claim that it was not an innovation is ignoring its functionality and novelty at the time it was released.


> MP3 players were already pretty mainstream when the iPod came around and it didn't really add much.

That's like saying that cell phones with color screens were already pretty mainstream when the iPhone came around and it didn't really add much.

I think you have forgotten just how different the iPod was.

Let's add the Mac to this party. Do you think that the Mac was just reinventing earlier standards on the market?


> That's like saying that cell phones with color screens were already pretty mainstream when the iPhone came around and it didn't really add much.

> I think you have forgotten just how different the iPod was.

It wasn't very different. Jesus Apple fans talk a lot of shit. The first iPhone couldn't even install apps, stop this ridiculous rose tinted view.


I don't know who this Jesus Apple guy is, but I don't own an iPod, iPhone, or iPad.


The iPod didn't add much, when you're a techie looking at technical specs.

What the iPod added was fewer features (yes, that's a good thing when you do it right) and usability. It did what people wanted, and in a way that regular people could actually use. For reasons I can't comprehend, a large number of technical people continue to be unable to see the value in that, but it does provide immense value, and that's why the iPod was so popular for so long.


> MP3 players were already pretty mainstream when the iPod came around and it didn't really add much.

In the world of techies, perhaps, but it hadn't made much of a dent per-iPad in the mainstream consumer market. http://www.clickz.com/clickz/news/1695501/ipod-mp3-player-ad...


Marketing and gaining widespread adoption seem to be the areas that Apple excels at. In most cases where something is heralded as "revolutionary" by Apple, it has been done before. Maybe not as well or as polished, but usually it has been done before.


It's pretty hard to see the iPhone and iPad as anything other than a revolution. Similar to how Windows was a revolution in desktop computing (whether you liked it or not).

"Gaining widespread adoption" was never an area Apple was very good at. They were always good at creating consumer-focused PCs. The initial line of MacBooks and iMacs are good examples of this — minuscule user base compared to Windows, but built to Apple's standards of what a computer should be. Their initial aim for the iPhone was to capture 1% of the market, they hardly expected the success they received.


The Ipad definitely wasn't a revolution.


It was marketed as one though. And to be fair it was the first tablet to gain widespread adoption.


What is your criteria for "revolution" in the tech industry?


Even if all they do is do things better and more polished than others, that's something any company would do well to emulate.

And sometimes doing things simply in a "more polished" way is revolutionary.

The early, pre-iPod mp3 players were an absolute mess when it came to navigating your music. That doesn't show up on a spec sheet, but "creating the first mp3 player that lets you actually navigate your music in a simple and enjoyable manner" is pretty freaking revolutionary in my book.


Yes and no. I mean you cad hardly say the iPod shuffle was a revolution in music navigation. They just marketed it right.


It was a pretty good product for a lot of people! A lot of people, like me, just want to shuffle some music while we exercise and we're willing to trade fine-grained control for the fact that the device is as light as a feather.

It also has really respectable sound quality and volume, which is not always the case (to put it gently) for teeny tiny electronics.


I agree 100% with the first two sentences - but their strong point is exactly that they finish their products too. It is important to execute projects well (enough) and to add polish too - and Apple excels at that. (to be fair, their UX is far from perfect IMHO... but it is apparently good enough)


Saying that Apple is good at "gaining widespread adoption" is kind of like saying that J. K. Rowling is good at "being read".


If J. K. Rowling had only copied other books and stories, then yes, that would have been a good description.


You can't expect this to be taken seriously.


Well... yeah. Remember FairPlay? http://en.wikipedia.org/wiki/FairPlay


Yeah, I'm absolutely aware of this; it just seems to be so wasteful from engineering point of view.


Well, if you look at Airplay, they had a technology that worked well for _years_ until the "leading standard" caught up. They do support other standards such as A2DP though.


A drop in the bucket. Thinking just about code, the amount of redundant code or code that will never ship or will only ship to a few people or will be replaced in 5 years must be enormous


The steering-wheel to be precise :)


From the brief amount of information available so far, CarPlay doesn't seem to preclude the in-car system from supporting other mobile phone OS's.

For instance it seems possible to equip a vehicle with, let's say, the Open Automotive Alliance's in-car Android system that can use Project Mode with Android phones and also implement CarPlay for when iPhones are present. After all, CarPlay appears to be dependent on the phone's OS for all displays and whatnot, with CarPlay just being the protocols to link the in-car system to iOS in the necessary manner.

This is just speculation, we can't know for certain until some of the technical details emerge, but it seems unlikely that car manufacturers would limit their systems interoperability to one platform.


Seems like it mainly depends on Apple's licensing terms.


True.


Or (probably more Apple-like approach) make it stand-alone and interact well with iPhones and leave it open to Android support later on or maybe start by offering the kind of support for android other mid-high end systems offer.

So the thing is usable without a phone: GPS, music, etc. You can answer calls for any bluetooth enabled phone. Advanced functionality like siri would only work with an iPhone.

This is a new market with new use-cases. Starting by making the basics work well is a good starting point.

I think that what iPhone (and Android shortly after) did for phones can be done for a bunch of things, including conventional PCs. Apple is uniquely capable of targeting the high end but they are not the only ones capable here. They'll need to execute very well.


Or, they can continue making their head units an upsell and offer a tightly integrated experience to people who want it, while also making a little extra money on the sale. They've been doing it witn nav upgrades for years.


Seems unlikely this will be standard equipment on any car (for several reasons, but mostly the one you stated).


There is no auto manufacturer, or manufacturer of any kind, really, that is in a position to push Apple.


Wait, so this works with the iPhone 5, yet you have to plug in your phone to make it work? What is this, 2003? I get in my car, the radio syncs up to my WP8 phone via Bluetooth, and if I get a text message I can have the car read it to me. I can dictate the response, or choose to ignore it, or call back the person, all with my voice. It can play music from my phone, including internet radio. I can call up an address search, and the Maps app will give me voice guided turn by turn directions.

All of this on a Windows Phone device that stays in my pocket, connected to a single-DIN aftermarket stereo without a fancy LCD, in a 2000 model Toyota. Recent cars with advanced navigation head units can do much, much more than mine. I think Apple is a bit behind the curve here.


There is not a phone on the market that will project an interface over Bluetooth to a display in the car. What curve are they behind exactly?


Do you really think it's mirroring the phone's display onto the car's screen? That would be pointless, as you don't want dozens of irrelevant app icons cluttering your car's display. You want the basic controls that are useful in a car, with large targets that can be hit with a glance and later with muscle memory.

If you really think they are just mirroring the display, then you didn't even look at the linked Apple page.


My guess is that the iOS SDK has been expanded to allow for a second screen that is output over lightning. So no, it's not mirroring the phone screen, but displaying another screen powered by the phone. This would seem to fit with it being iPhone5 and above only as lightning is designed to have these sorts of custom digital output.


Correct. At a minimum, it's a secondary display. But it's also deep integration into both ends over the Lightning cable (hence the iPhone 5 requirement).


Yes, this is the case. External screen support was added a few years ago.


That makes a lot of sense, thank you.


I'm sure you can do all that with bluetooth on the iPhone too.

The point here is the touchscreen appears to be an extension of the phone.


Again, I don't see why that needs USB tethering, apart from keeping the phone charged. If it's just an extension screen, with a sensibly small number of large icons relative to driving, it can be done (and has been done in the past) via wireless links. Much like the Sony Smartwatch, Pebble, or Galaxy Gear, the external screen is just a remote control for certain specific functions.

But this is Apple, so perhaps they want to make sure the interface only works with their phones. That makes sense as they are designing the in-car interface in the first place.


Bluetooth doesn't support enough application throughput to enable streaming things like maps (the map solution you're referring to uses on-board maps, not Apple's native iOS maps). Everyone would love to not require a physically connected phone, but it's not possible right now.



To be blunt I am pretty capable of driving without any maps in the car, or at ignoring instructions that appear incorrect or dangerous. And while Apple Maps might still get it wrong more than other maps, I've never used a mapping service that hasn't gotten some things wrong.

I do hope it supports other mapping apps though, or if not I hope Apple gets pressured to support that. And I don't really see why they wouldn't, it makes their product better.


The GPS get it wrong very often too. There is an enormous amount of magic going on behind the scene to put a car on an actual road on the map.

When I'm driving on the highway with spotty coverage, I very often hear "Turn Left in 50 m" when the magic fails and the GPS (mobile phone running TomTom in this case) now thinks I teleported on a side road.

Similarly, you could get yourself lost for hours if you follow your GPS blindly in a small European city. Actually, you would probably get into a traffic accident before that.

When I see people comment on GPS, I can only imagine that in 20 years, you will see headline like "Family had to be rescued after their autonomous car failed to exit Walmart Parking lot for 5 hours and ran out of gas. Google made no comment."



Firstly if the road sign says no entry don't go down it because the sat nav says to, there are plenty of mistakes in all sat navs and maps and even if there weren't there are changes they can't keep up with, roadworks etc.

I wondered what sort of international airport can you drive onto the runway of without being stopped so I looked it up:

http://dot.alaska.gov/faiiap/pdfs/GA_Control_Surf.pdf

It's the Alaskan sort secured by yellow lines and red signs. Now I'm not saying Apple isn't at fault here but I would put greater blame on the airport. I would expect if not manned or electronic gates with an intercom at least there should be a barrier the of some sort to act as a warning. I'm not sure signs and lines are enough for something like an airport.

I guess being Alaska manning gates/opening windows or getting out of cars are seriously unappealing activities in the Winter.


I don't understand why Apple hasn't purchased Foresquare.


Foursquare doesn't have its own maps.


Correct, they use mapbox, which uses Open Street Map under the hood.


What puzzles me is that there has been extremely little improvement to Siri since the first introduction 2.5 (!) years ago - at least, none that I have noticed. That is a long time for there to not be major updates or to release her from beta.

I would have expected/hoped that it would be possible to simply silently improve the server-side algorithms because of all the data being sent in and analysed. But so far it's been a big disappointment. Google Now is far better and more useful.


>> "What puzzles me is that there has been extremely little improvement to Siri since the first introduction 2.5 (!) years ago - at least, none that I have noticed. That is a long time for there to not be major updates or to release her from beta."

Siri exited beta with iOS 7. As for updates there have been quite a lot but nothing major. They added new more human sounding voices. They added new sources (Wikipedia, Twitter). They added other functions like returning calls, playing voicemail, adjusting brightness. That was just iOS 7. iOS 6 added more I believe.


Do you mind providing examples of use cases re Google Now vs Siri? Not disagreeing with you - just haven't owned an Android device in the last year to know.


You can use Google Now on your iOS device (I guess I'm assuming you have an iOS device). Just download the Google App for iOS.

https://itunes.apple.com/us/app/id284815942?mt=8

At least I think that's Google Now. It looks similar though maybe it's missing some features?

Of course with it not being integrated into the experience it's not quite the same. (can't appear on your home screen etc...)



this talks about some of the added features in the app over time http://forum.xda-developers.com/showthread.php?t=1961636


I think that making it an extension of an existing phone was a good choice. When I heard that Apple was planning some sort of device for cars, I was imagining how much of a pain it would be to update apps and keep things in sync with a device in the car. Sounds like it's essentially a second-screen for the iPhone (plus some built-in controls).


Ah finally, no more atrocious pressure-sensitive touch screens. I wouldn't be surprised if this feature alone is sufficient to increase demand for supported brands in large enough quantities to make other manufacturers take notice.


Am curious who is in charge on manufacturing the device.. Both car manufacturers and Apple have history of wanting build the "whole widget."


The home, menu and volume controls on the left don't look like something Apple would make.


Those main images are from a 2013/4 honda Civic.

More

Applications are open for YC Winter 2018

Guidelines | FAQ | Support | API | Security | Lists | Bookmarklet | DMCA | Apply to YC | Contact

Search: