Perhaps because Apple is American and the U.S. is crazy about cars, this has been a blind-spot for Apple. They really, really should not be encouraging the end of the planet. Apple really ought to be investing in eliminating the need to travel.
Every IT company in existence is eliminating the need to travel.
And as this report clearly shows, no one outside Apple has a clue what they're planning. I think they're looking for some sort of autonomous driverless car rental thing. Cars are expensive to own and spend a good 80% to their time in car parks or driveways. I don't think Apple sees any mileage(?) in contributing to that.
This makes no sense. This use would be so trivial as to be a complete waste of talent, time and money. If this is, instead just a way to get experience, then it could be useful, but the article doesn’t seem to be saying that.
i also have strong doubts that Apple could have a wide success with a real CarOS. There are several entrenched competitors in this field. And yes, I know that Apple has faced entrenched industries before. But it really needs to be remembered that Apple made its success in those industries with software based hardware, not just software. Why would car makers want Apple’s OS? If Apple is as restrictive about sending info back to the car makers as they are with other clients, such as the Ad agencies, then the car makers will have the same reaction. Which is to move away. And if Apple doesn’t, then Apple will lose its voice as the holder of privacy for its customers.
So without an actual vehicle for this OS effort, I don’t see it going anywhere, unless fleets take a shine to it, and that seems very un Apple-like.
Reading the article it seemed as though there were team members wanting to reinvent the automobile and then reality set in and Bob Mansfield was brought in to bring the project back down to earth. But I don’t see how it goes anywhere unless Apple eventually develops its own vehicle.
EDIT: John Gruber had his own take (which sounds like he might have heard some things from employees):
“Shelved” is an accurate word, but I think many people have interpreted it as meaning that Apple has given up on designing its own vehicles. My understanding is that it’s more like “Let’s get the autonomous shit down first, and worry about designing vehicles to put it in after that.” Eat the steak one bite at a time rather than all at once.
This makes the most sense. If you don’t have the software nailed there’s really no point to a car.
Hardware car guys at Apple just sit around until the day where their autonomy is perfected? I highly doubt Apple are that foolish. No, they’ll work on autonomy and simultaneously will have teams around the globe, as they do, working on other areas such as design (how it’ll work) and drive trains and a whole host of other difficult car technologies. Awaiting for the day when their autonomy is perfected before beginning the hardware would be silly. Better cut that steak into 4 pieces and have four people each have a smaller-than-full-steak bite.
This makes no sense. This use would be so trivial as to be a complete waste of talent, time and money. If this is, instead just a way to get experience, then it could be useful, but the article doesn’t seem to be saying that.
i also have strong doubts that Apple could have a wide success with a real CarOS. There are several entrenched competitors in this field. And yes, I know that Apple has faced entrenched industries before. But it really needs to be remembered that Apple made its success in those industries with software based hardware, not just software. Why would car makers want Apple’s OS? If Apple is as restrictive about sending info back to the car makers as they are with other clients, such as the Ad agencies, then the car makers will have the same reaction. Which is to move away. And if Apple doesn’t, then Apple will lose its voice as the holder of privacy for its customers.
So without an actual vehicle for this OS effort, I don’t see it going anywhere, unless fleets take a shine to it, and that seems very un Apple-like.
Reading the article it seemed as though there were team members wanting to reinvent the automobile and then reality set in and Bob Mansfield was brought in to bring the project back down to earth. But I don’t see how it goes anywhere unless Apple eventually develops its own vehicle.
EDIT: John Gruber had his own take (which sounds like he might have heard some things from employees):
“Shelved” is an accurate word, but I think many people have interpreted it as meaning that Apple has given up on designing its own vehicles. My understanding is that it’s more like “Let’s get the autonomous shit down first, and worry about designing vehicles to put it in after that.” Eat the steak one bite at a time rather than all at once.
This makes the most sense. If you don’t have the software nailed there’s really no point to a car.
Hardware car guys at Apple just sit around until the day where their autonomy is perfected? I highly doubt Apple are that foolish. No, they’ll work on autonomy and simultaneously will have teams around the globe, as they do, working on other areas such as design (how it’ll work) and drive trains and a whole host of other difficult car technologies. Awaiting for the day when their autonomy is perfected before beginning the hardware would be silly. Better cut that steak into 4 pieces and have four people each have a smaller-than-full-steak bite.
Beware of anything that cites 'sources close to the project'.
If Gruber hasn't mentioned it then take it with a pinch of salt.
Personally I can't see a really good reason for Apple to actually develop a car. To what end?
They make far more profit with consumer and pro-grade electronics and computing. Ain't no way in Hades that they can realize a 30+ margin on a car all things considered. Tooling, patent royalty payments, on-going component support for several years after last sale, maintenance facilities, warranty issues, distinct (and IMO almost certain at some point) possibilities of high-profile, headline-grabbing, legally-mandated recalls that damage Apple's brand, and toss in liabilities and lawsuits over injuries and possibly deaths. Other than "hey me too" why would they continue to pursue a product in a market where no profits at all are possible and with so many downsides?
Personally I think Apple's management has realized that building a car in the foreseeable future doesn't make sense for Apple as a company. Nice thought and well worth investigating since even if a car is never built the learning Apple gained from putting some effort into it can pay dividends in other areas like AR, mapping, robotics, display technologies, etc., but I believe Apple producing a car of their own is more problematic than practical as things stand. Even the traditional automakers who have been working towards this for years and years are struggling with where the profits will come from with future transportation, and how autonomy will fit in actual practice. At some point everyone will have to be if not on the exact same page at least in the same chapter for all of this to work. It is not a space where Apple will be able to dictate terms.
IMHO Apple will not be building their own car from the ground up.
If Apple is focussing on software for their first car, that could be because they have decided to purchase (via a friendly or hostile takeover) an existing car company outright. Most of the big car companies have a total market cap of $40-$80 billion. Apple could even buy out Toyota for $180 billion. But I would rather see Apple shell out a measly $1 billion for McLaren Automotive (though I suspect they have their eyes on BMW.) McLaren's cars are the "Apple of automobiles", in my opinion. Right now the cheapest McLaren, the MP4-12C, is about $200K, and Apple would have to work a little on reducing that price. In 23 years of Top Gear, only nine other cars have gone higher on their charts than the entry-level MP4-12C, one of which, the very top one, is another McLaren, the 675LT.
I'd favor Apple acquiring McLaren over BMW or anyone else just for the engineering expertise. Sure, McLaren doesn't have the manufacturing facilities that BMW has, but McLaren's F1 racing expertise is not grown on trees. They concentrate on gasoline engines but have a hybrid and just unveiled a pure electric vehicle.
IMO, if Apple acquired McLaren, it would be like the PA Semi acquisition - the PA Semi expertise vaulted Apple's iDevice performance to the top.
Personally I can't see a really good reason for Apple to actually develop a car. To what end?
They make far more profit with consumer and pro-grade electronics and computing. Ain't no way in Hades that they can realize a 30+ margin on a car all things considered. Tooling, patent royalty payments, on-going component support for several years after last sale, maintenance facilities, warranty issues, distinct (and IMO almost certain at some point) possibilities of high-profile, headline-grabbing, legally-mandated recalls that damage Apple's brand, and toss in liabilities and lawsuits over injuries and possibly deaths. Other than "hey me too" why would they continue to pursue a product in a market where no profits at all are possible and with so many downsides?
Personally I think Apple's management has realized that building a car in the foreseeable future doesn't make sense for Apple as a company. Nice thought and well worth investigating since even if a car is never built the learning Apple gained from putting some effort into it can pay dividends in other areas like AR, mapping, robotics, display technologies, etc., but I believe Apple producing a car of their own is more problematic than practical as things stand. Even the traditional automakers who have been working towards this for years and years are struggling with where the profits will come from with future transportation, and how autonomy will fit in actual practice. At some point everyone will have to be if not on the exact same page at least in the same chapter for all of this to work. It is not a space where Apple will be able to dictate terms.
IMHO Apple will not be building their own car from the ground up.
You need to stick to watching Google, my friend. You clearly haven't been watching Apple.
This makes no sense. This use would be so trivial as to be a complete waste of talent, time and money.
This sounds like something the pundits have said about pretty much everything Apple has done.
Really? When has Apple ever developed software for other people’s hardware? I guess maybe Motorola ROKR and we know how that turned out.
Constantly, but that's not the point. The point is suggesting that Apple starting small is rudderless because you can't see how it can grow into something more or you envision that Apple has to come out with something epic when pretty much everything they do starts with a manageable foundation.
Name for me a non-Apple consumer hardware product I can buy that uses Apple software.
This makes no sense. This use would be so trivial as to be a complete waste of talent, time and money.
This sounds like something the pundits have said about pretty much everything Apple has done.
Really? When has Apple ever developed software for other people’s hardware? I guess maybe Motorola ROKR and we know how that turned out.
Constantly, but that's not the point. The point is suggesting that Apple starting small is rudderless because you can't see how it can grow into something more or you envision that Apple has to come out with something epic when pretty much everything they do starts with a manageable foundation.
Name for me a non-Apple consumer hardware product I can buy that uses Apple software.
Why are you ignoring the first sentence of his reply. "That's not the point." Soli's not arguing that Apple is/should be taking a software-only approach. He's just saying "wait and see; time after time people have looked foolish when they pooh-poohed Apple's strategies based on initial glimpses."
No need to rush things at all. There will be self-driving car fatalities and the car/software makers will be blamed in the media. I'd hate to see Apple have to explain why a code error or unanticipated scenario sent someone to the mortuary.
Death by force unwrapped optionals. Tesla and Uber already went through some crash incidents:
Bus/taxi/train drivers can be prosecuted for accidents or fatalities so if the vehicle software is found to be to blame then I'd expect accidents to result in lawsuits for the manufacturers. There are a list of payouts here from a law firm due to vehicle accidents, they are usually in the hundreds of thousands of dollars but go up to a few million:
The vehicles will have data to say exactly what happened. They will have all manner of safety features, even fully automated vehicles can have an emergency system that causes the vehicle to move quickly to a safe state. The AI software can be tested in virtual environments.
Software simulations can run billions/trillions of scenarios through the control software before it's deployed. They can simulate the data the vehicle sensors would get in different weather conditions, they can put human drivers in the simulation. Apple might have been scanning the roads to build this kind of simulation. Even in the event of accidents, the software can evolve to never allow them to happen again, human drivers will keep making the same mistakes because humans have to keep learning how to drive every generation that comes along so there's always a new supply of inexperienced and aggressive drivers, a new supply of elderly drivers, tired drivers etc.
The number of vehicle fatalities is fairly low worldwide (1.25m) with 90% happening in developing countries:
The number of injuries is higher at 20-50m out of 1b vehicles. Replacing 1b vehicles is a pretty big task. Tesla is only going to manage ~100k/year just now. Using ride-sharing is easier to make an impact because people only have to pay a journey fare to benefit rather than the full cost of a car and deal with charging. Ride sharing and taxis can transition millions of people to driverless overnight.
That's where providing software alone, like CarPlay ( https://www.apple.com/ios/carplay/available-models/ ), can work out better. They can do both though like how Google makes phones and Android for other phones. Apple's vehicles would be among the best ones made and set a standard while their software can be used with partners. They'd have to be involved in the manufacturing even with 3rd parties to integrate the sensors properly and control the different vehicle classes.
Large cities should really be more proactive with this and closing down routes to be only used by automated vehicles. For example, have a road that cuts right through the city and only allow automated vehicles on it and people use the vehicles to get from one side of the city to the other. Then they can use taxis or public transport beyond that. The more that it reduces traffic and accidents, the more they can expand it until entire cities are only automated. This will also reduce infrastructure costs because automated cities can operate with much fewer vehicles and it cleans up the cities from having huge parking facilities and cars littering the sides of the roads.
This makes no sense. This use would be so trivial as to be a complete waste of talent, time and money.
This sounds like something the pundits have said about pretty much everything Apple has done.
Really? When has Apple ever developed software for other people’s hardware? I guess maybe Motorola ROKR and we know how that turned out.
Constantly, but that's not the point. The point is suggesting that Apple starting small is rudderless because you can't see how it can grow into something more or you envision that Apple has to come out with something epic when pretty much everything they do starts with a manageable foundation.
Name for me a non-Apple consumer hardware product I can buy that uses Apple software.
1) You have iTunes and Quicktime, to name just two, that have been on Windows for around 16 and 20 years, respectively. Quicktime for Windows development ceased last year but iTunes is going strong. Then you have countless "other people’s hardware," as you specifically put it, that Apple purchases to use in their final products. Intel processors, Qualcomm to cellular chips, and Samsung NAND are just three such examples.
2) As @randominternetperson astutely comments, my point is to just wait and see instead of asserting Apple doesn't know what they're doing and/or won't do something based on a rumour. Many here said Apple would never release a digital personal assistant for the home and argued that if that was a viable market they would've beaten Amazon, despite history showing that Apple is rarely first. History also shows they're willing to change up how they approach a product when the dynamics are different.
"Some sought to build the software off Apple's Swift programming language, while others wanted to employ C++."
No freaking way was THAT a debate. Was Fortran the 3rd contender?
I don't know enough about either language to know if this could be a real debate—unlike the previous internal challenge to grow the iPod's Pico(?) OS into the Apple tablet OS that eventually became the iPhone or to reduce macOS nee Mac OS X—but my gut reaction is that C++ over Swift is odd. The only way that makes any sense to me is if Apple has bought a well developed system that is built off a comprehensive C++ codebase.
One thing that was often referenced with Blackberry buying QNX is that it was a "true" realtime OS that was in nearly all automobiles. It looks like it built using C, C++, Embedded C++, and/or Java. Does anyone know what Tesla uses?
This makes no sense. This use would be so trivial as to be a complete waste of talent, time and money.
This sounds like something the pundits have said about pretty much everything Apple has done.
Really? When has Apple ever developed software for other people’s hardware?
CarPlay.
That's a good point, but possibly debatable. At the very least Apple did have to develop a system that can work through the car, but the OS and UI are still never licensed out to the automobile makers as it all resides on the iPhone. Personally, I'd consider that "for other people's hardware," but then we could not say that about WiFi routers, Bluetooth and USB accessories for using standards that work with limitless devices?
PS: I've alway been curious just how much extra effort auto makers had to put in to make Android Auto work after Apple created a foundation for CarPlay to work.
Cook called the self driving car " the mother of all AI projects". I think that says it all. I don't think they are really serious about designing the actual car yet because unless/until they have the software down it will be premature. If they designed the car now it would certainly not be what they want by the time they have the software ready. So, it's obvious to me why they will stick with the software for a while until the picture is a little clearer.
Perhaps because Apple is American and the U.S. is crazy about cars, this has been a blind-spot for Apple. They really, really should not be encouraging the end of the planet. Apple really ought to be investing in eliminating the need to travel.
I can pretty much exist on my couch surrounded by and wearing Apple products already.
This makes no sense. This use would be so trivial as to be a complete waste of talent, time and money. If this is, instead just a way to get experience, then it could be useful, but the article doesn’t seem to be saying that.
We don't need anyone to tell us what is obvious. It makes perfect sense.
I agree that it makes "sense"...because I want it. I want to be stunned and amazed at what a clean slate and years of R&D and Apple finesse add to a car.
It also makes no "sense"...because it's a waste of engineering and software for the few AI choices that would have to be made to drive between one Apple facility and another. No one has educated me on the world-winnning approach Apple will bring to what amounts to utility software, especially in the face of a hundred other competitors.
Personally I can't see a really good reason for Apple to actually develop a car. To what end?
They make far more profit with consumer and pro-grade electronics and computing. Ain't no way in Hades that they can realize a 30+ margin on a car all things considered. Tooling, patent royalty payments, on-going component support for several years after last sale, maintenance facilities, warranty issues, distinct (and IMO almost certain at some point) possibilities of high-profile, headline-grabbing, legally-mandated recalls that damage Apple's brand, and toss in liabilities and lawsuits over injuries and possibly deaths. Other than "hey me too" why would they continue to pursue a product in a market where no profits at all are possible and with so many downsides?
Personally I think Apple's management has realized that building a car in the foreseeable future doesn't make sense for Apple as a company. Nice thought and well worth investigating since even if a car is never built the learning Apple gained from putting some effort into it can pay dividends in other areas like AR, mapping, robotics, display technologies, etc., but I believe Apple producing a car of their own is more problematic than practical as things stand. Even the traditional automakers who have been working towards this for years and years are struggling with where the profits will come from with future transportation, and how autonomy will fit in actual practice. At some point everyone will have to be if not on the exact same page at least in the same chapter for all of this to work. It is not a space where Apple will be able to dictate terms.
IMHO Apple will not be building their own car from the ground up.
You need to stick to watching Google, my friend. You clearly haven't been watching Apple.
So... ?
You've not explained why you think they're definitely building their own complete car from the ground up. I explained why I think they're not and the reasoning behind it. Don't take the easy way out with some vague comment that anyone could make without thought. So again, to what end?
Personal transport is going through a huge transformation right now. Future car will be a 90/10 split between a comfortable seating pod with lots of entertainment and a tiny power train that you hardly notice. Software for engine management, navigation, and infotainment will be what differentiates brands. Think super lux pods in first class sections of airplanes. Would love to sleep my way to and from Vegas. Apple will be part of this.
Personal transport is going through a huge transformation right now. Future car will be a 90/10 split between a comfortable seating pod with lots of entertainment and a tiny power train that you hardly notice. Software for engine management, navigation, and infotainment will be what differentiates brands. Think super lux pods in first class sections of airplanes. Would love to sleep my way to and from Vegas. Apple will be part of this.
I wonder if the majority of people simply stop buying cars. The convenience of tapping an app and getting a driverless car to collect you and drive you to your destination, without any of the maintenance and ownership concerns that come with owning a vehicle may prove too tempting to most. Especially if these car pods become quite paired down and mechanically simply. So regular people rather than opting for the best entertainment may opt for some cheaper solutions, while being advertised to and getting privacy raped: electrification, autonomy, ride hailing — as Tim has said, and Elon is doing.
Comments
Every IT company in existence is eliminating the need to travel.
And as this report clearly shows, no one outside Apple has a clue what they're planning. I think they're looking for some sort of autonomous driverless car rental thing. Cars are expensive to own and spend a good 80% to their time in car parks or driveways. I don't think Apple sees any mileage(?) in contributing to that.
If Gruber hasn't mentioned it then take it with a pinch of salt.
They make far more profit with consumer and pro-grade electronics and computing. Ain't no way in Hades that they can realize a 30+ margin on a car all things considered. Tooling, patent royalty payments, on-going component support for several years after last sale, maintenance facilities, warranty issues, distinct (and IMO almost certain at some point) possibilities of high-profile, headline-grabbing, legally-mandated recalls that damage Apple's brand, and toss in liabilities and lawsuits over injuries and possibly deaths. Other than "hey me too" why would they continue to pursue a product in a market where no profits at all are possible and with so many downsides?
Personally I think Apple's management has realized that building a car in the foreseeable future doesn't make sense for Apple as a company. Nice thought and well worth investigating since even if a car is never built the learning Apple gained from putting some effort into it can pay dividends in other areas like AR, mapping, robotics, display technologies, etc., but I believe Apple producing a car of their own is more problematic than practical as things stand. Even the traditional automakers who have been working towards this for years and years are struggling with where the profits will come from with future transportation, and how autonomy will fit in actual practice. At some point everyone will have to be if not on the exact same page at least in the same chapter for all of this to work. It is not a space where Apple will be able to dictate terms.
IMHO Apple will not be building their own car from the ground up.
IMO, if Apple acquired McLaren, it would be like the PA Semi acquisition - the PA Semi expertise vaulted Apple's iDevice performance to the top.
Is today April 1st and I missed it?
"Some sought to build the software off Apple's Swift programming language, while others wanted to employ C++."
No freaking way was THAT a debate. Was Fortran the 3rd contender?
You need to stick to watching Google, my friend. You clearly haven't been watching Apple.
Why are you ignoring the first sentence of his reply. "That's not the point." Soli's not arguing that Apple is/should be taking a software-only approach. He's just saying "wait and see; time after time people have looked foolish when they pooh-poohed Apple's strategies based on initial glimpses."
https://www.wired.com/2017/01/probing-teslas-deadly-crash-feds-say-yay-self-driving/
https://www.wired.com/2017/03/uber-self-driving-crash-tempe-arizona/
Bus/taxi/train drivers can be prosecuted for accidents or fatalities so if the vehicle software is found to be to blame then I'd expect accidents to result in lawsuits for the manufacturers. There are a list of payouts here from a law firm due to vehicle accidents, they are usually in the hundreds of thousands of dollars but go up to a few million:
http://www.farrin.com/settlements-cases/north-carolina-attorney-legal-case-settlements
The vehicles will have data to say exactly what happened. They will have all manner of safety features, even fully automated vehicles can have an emergency system that causes the vehicle to move quickly to a safe state. The AI software can be tested in virtual environments.
Basic software results in this:
https://www.twitch.tv/videos/166060401
but more advanced software would result in this:
Software simulations can run billions/trillions of scenarios through the control software before it's deployed. They can simulate the data the vehicle sensors would get in different weather conditions, they can put human drivers in the simulation. Apple might have been scanning the roads to build this kind of simulation. Even in the event of accidents, the software can evolve to never allow them to happen again, human drivers will keep making the same mistakes because humans have to keep learning how to drive every generation that comes along so there's always a new supply of inexperienced and aggressive drivers, a new supply of elderly drivers, tired drivers etc.
The number of vehicle fatalities is fairly low worldwide (1.25m) with 90% happening in developing countries:
http://www.who.int/mediacentre/factsheets/fs358/en/
The number of injuries is higher at 20-50m out of 1b vehicles. Replacing 1b vehicles is a pretty big task. Tesla is only going to manage ~100k/year just now. Using ride-sharing is easier to make an impact because people only have to pay a journey fare to benefit rather than the full cost of a car and deal with charging. Ride sharing and taxis can transition millions of people to driverless overnight.
That's where providing software alone, like CarPlay ( https://www.apple.com/ios/carplay/available-models/ ), can work out better. They can do both though like how Google makes phones and Android for other phones. Apple's vehicles would be among the best ones made and set a standard while their software can be used with partners. They'd have to be involved in the manufacturing even with 3rd parties to integrate the sensors properly and control the different vehicle classes.
Large cities should really be more proactive with this and closing down routes to be only used by automated vehicles. For example, have a road that cuts right through the city and only allow automated vehicles on it and people use the vehicles to get from one side of the city to the other. Then they can use taxis or public transport beyond that. The more that it reduces traffic and accidents, the more they can expand it until entire cities are only automated. This will also reduce infrastructure costs because automated cities can operate with much fewer vehicles and it cleans up the cities from having huge parking facilities and cars littering the sides of the roads.
2) As @randominternetperson astutely comments, my point is to just wait and see instead of asserting Apple doesn't know what they're doing and/or won't do something based on a rumour. Many here said Apple would never release a digital personal assistant for the home and argued that if that was a viable market they would've beaten Amazon, despite history showing that Apple is rarely first. History also shows they're willing to change up how they approach a product when the dynamics are different.
One thing that was often referenced with Blackberry buying QNX is that it was a "true" realtime OS that was in nearly all automobiles. It looks like it built using C, C++, Embedded C++, and/or Java. Does anyone know what Tesla uses?
PS: I've alway been curious just how much extra effort auto makers had to put in to make Android Auto work after Apple created a foundation for CarPlay to work.
I agree that it makes "sense"...because I want it. I want to be stunned and amazed at what a clean slate and years of R&D and Apple finesse add to a car.
It also makes no "sense"...because it's a waste of engineering and software for the few AI choices that would have to be made to drive between one Apple facility and another. No one has educated me on the world-winnning approach Apple will bring to what amounts to utility software, especially in the face of a hundred other competitors.
You've not explained why you think they're definitely building their own complete car from the ground up. I explained why I think they're not and the reasoning behind it. Don't take the easy way out with some vague comment that anyone could make without thought. So again, to what end?