Third-party TeslaOS in-car apps?

[Orcaton]

Has Tesla/Elon ever mentioned anything about opening up for external developers to create apps for the in-car display? I think that would be another game-changer and further cement Teslas lead in BEVs. A Tesla-controlled software eco-system (like iOS) would make it even harder for other car manufacturers to keep up.

Would third-party developers be interested in creating apps for Teslas? I think so. Especially since anyone buying a relatively expensive car would probably not have a big issue paying $10s to $100s for the right app. (As opposed to iOS/Android, where paying just a few $ seems to be big barrier for many.)

What are your thoughts? What third-party apps would you like to see?

 

[Runebane]

I don't think they would open it up to 3rd party. I think the best we could hope for is that they would work with other music offerings (Pandora, Google, etc...) and Tesla would build the apps and let us choose which to use. But even that I don't think would come anytime soon if ever.

 

[BenR]

Elon was asked exactly this question very recently (can't remember where) and his response was basically "not practical to develop until we reach 10m cars on the road - might think about it at that point."

 

[run-the-joules]

What are your thoughts? What third-party apps would you like to see?

I'd rather they just implement CarPlay and Android Auto, but if they roll their own "app store", I'd want Audible.com player, a podcast player that doesn't suck like TuneIn does, and Waze.

and a better app for interfacing with the music on my phone, of some sort.

 

[srs5694]

Context and disclaimer: I work in the computer field, but not specifically in computer security. I'm also not familiar with details of how the computers are connected within a Tesla. That said....

One concern I'd have with a Tesla "app store" is that there would be security breaches. That's a Big Deal on a car, even if the infotainment computer is well-isolated from more critical systems. A breach that affects the display could give you the wrong speedometer readings, for instance. If a Tesla app is to be more useful than something you could run on a cell phone now, especially via Apple CarPlay or Android Auto, then it would need access to Tesla-specific data -- say, to improve routing to SuperChargers or tweak the way Autopilot functions. Such access creates the possibility of more pernicious exploits. If Tesla were to implement this idea, they'd be wise to severely limit what third-party apps can do, which would reduce their appeal.

A CarPlay/Android Auto app, OTOH, could be limited to displaying data where the Tesla's map is located, providing better isolation from both critical display data and (probably) the rest of the car's computers. Of course, that'd mean you'd get no nifty-cool Tesla-specific features, either; but at least the risk of the car driving itself into a lake because of a third-party bug or exploit will be near zero. This approach would also mean that developers wouldn't need to develop for yet another platform, which from a developer's perspective, looks a lot like this....

 

[Knightshade]

Tesla has thusfar been pretty explicitly uninterested in letting anything 3rd party touch their cars (or really in offering anything approaching infotainment at all competitive with anything else from this century)

They don't even support playing media from a plugged in phone.

Hell, they don't support nav waypoints- a nut that Garmin had cracked in like 2003.

 

[run-the-joules]

Tesla has thusfar been pretty explicitly uninterested in letting anything 3rd party touch their cars (or really in offering anything approaching infotainment at all competitive with anything else from this century)

They don't even support playing media from a plugged in phone.

Hell, they don't support nav waypoints- a nut that Garmin had cracked in like 2003.

Nor do they do any text message handling, unlike a Corolla from several years ago.

Also no playlist navigation for phone media.

BUT AT LEAST WE GET VIDEO GAMES AND A FART APP

 

[Knightshade]

BUT AT LEAST WE GET VIDEO GAMES AND A FART APP

Yeah, like Pole Posit...oh, wait...

 

[eXntrc]

Elon talked about going to an app mirroring model back in 2016, but I've not seen any updates since. If anyone has anything newer than this article, please share:

Tesla is moving away from an SDK and instead planning to allow app mirroring from iPhones to center consoles

 

[chris irwin]

Context and disclaimer: I work in the computer field, but not specifically in computer security. I'm also not familiar with details of how the computers are connected within a Tesla. That said....

One concern I'd have with a Tesla "app store" is that there would be security breaches. That's a Big Deal on a car, even if the infotainment computer is well-isolated from more critical systems. A breach that affects the display could give you the wrong speedometer readings, for instance. If a Tesla app is to be more useful than something you could run on a cell phone now, especially via Apple CarPlay or Android Auto, then it would need access to Tesla-specific data -- say, to improve routing to SuperChargers or tweak the way Autopilot functions. Such access creates the possibility of more pernicious exploits. If Tesla were to implement this idea, they'd be wise to severely limit what third-party apps can do, which would reduce their appeal.

A CarPlay/Android Auto app, OTOH, could be limited to displaying data where the Tesla's map is located, providing better isolation from both critical display data and (probably) the rest of the car's computers. Of course, that'd mean you'd get no nifty-cool Tesla-specific features, either; but at least the risk of the car driving itself into a lake because of a third-party bug or exploit will be near zero. This approach would also mean that developers wouldn't need to develop for yet another platform, which from a developer's perspective, looks a lot like this....

I love the cartoon, where did you find it? Do you mind if I use it elsewhere?

 

[srs5694]

I love the cartoon, where did you find it? Do you mind if I use it elsewhere?

It's from XKCD. Here's the link to that specific cartoon:

https://imgs.xkcd.com/comics/standards.png

I have no relation to the site, other than as a reader. Each cartoon provides an explicit link for embedding images, which I take to mean that the author encourages this type of sharing.

 

[afadeev]

Has Tesla/Elon ever mentioned anything about opening up for external developers to create apps for the in-car display?

Yes, and NO, he wont do it.

There are two excellent app-store ecosystem already in place: Google's and Apple's.
If you want cool apps in your Tesla, all you would need is to convince Elon to support AndroidAuto and CarPlay.
But he wont do it.

a

 

[Knightshade]

Yes, and NO, he wont do it.

He already said he would do an app store type deal once they had a much larger install base.

During the Ride the Lightning interview a couple months ago he said-

In order for it to be worthwhile for somebody to write an app, there has to be enough of an install base to warrant the effort. Even if you are going to port something, it’s still got to be worth the effort. As our number of vehicles grow, it starts to potentially makes more sense to develop games and other applications for Tesla. We just need a lot of cars.

 

[afadeev]

He already said he would do an app store type deal once they had a much larger install base.

Reading comprehension test - the LAST thing any of us need, and the opposite of what I commented on, is another walled-garden apps ecosystem from Elon, just for Tesla. Two much bigger and vastly superior application platforms already exist, and those cover 98+% market share of the smart phones in the drivers hands.
Mobile Operating System Market Share Worldwide | StatCounter Global Stats

Integration with CarPlay and AndroidAuto is what every other automaker, reluctantly, has conceded is the best for the customers.
Tesla might figure this out some day, but likely not any time soon.

 

[Knightshade]

Reading comprehension test

Yes, you need one.

You said he was asked about doing it and would not do it.

I cited an interview showing your claim was not what he said- and in fact what he said was yes, once the install base with a lot larger.

Where specifically are you getting lost?

 

[eXntrc]

Normally I try to avoid getting in the middle of an argument at all costs, but this is a subject that's very important to me so I'd like to try and weigh in respectfully.

@Knightshade, the issue with the quote you posted from Elon is that it includes this important phrase:

"Even if you are going to port something, it’s still got to be worth the effort."

Applications written for Android Auto or CarPlay do not require porting. Any application written for those platforms will run, unmodified, on any vehicle that supports those platforms. This is the sticking point that @afadeev is referring to when he refers to a "walled garden".

If Tesla ends up creating a new application platform it will require developers to port their applications. And there has to be a shockingly high bar to incentivize developers to do it (just ask Windows Phone). Again, this is what Elon appears to be referring to when he says:

"As our number of vehicles grow, it starts to potentially makes more sense to develop games and other applications for Tesla."

The problem with this approach is that any application platform with a small user base will always lag behind in updates and features. Developers barely have time to support 2 operating systems as it is. Even if they agreed to support a 3rd, that 3rd OS just won't get the attention it deserves. Not so long as it's user base is dwarfed by the other two.

Instead of a custom platform, if Tesla were to move forward with supporting Android Auto or CarPlay, hundreds (thousands?) of already existing applications would just work. Tesla would be opting into an existing ecosystem, which means developers wouldn't have to port anything or support anything different than what they already are. It would also mean that Tesla wouldn't need a massive amount of users to convince developers to support their platform.

Unfortunately, supporting Android Auto or CarPlay would mean giving up control over the app platform. Which in turn means giving up any potential revenue from the applications and subscriptions sold on that platform. This is what Tesla appears reluctant to do, even though it's probably in the best interest for its consumers.

I'd love to be wrong here. I'd love to see Tesla roll out both Android Auto and CarPlay support. But I'm afraid this is just not something I'm going to hold my breath for.

 

[JasonR67]

Apple CarPlay is not the same as iOS. You need special permission to develop CarPlay apps. It’s not the solution to yay! I get all my apps!

 

[eXntrc]

Apple CarPlay is not the same as iOS. You need special permission to develop CarPlay apps. It’s not the solution to yay! I get all my apps!

I knew it wasn't the same thing as iOS, but I wasn't aware special permissions were required from Apple. You learn something new every day. Well, that certainly is... less than ideal. Is it at least the same dev tools and experience as traditional iOS development?

 

[afadeev]

I knew it wasn't the same thing as iOS, but I wasn't aware special permissions were required from Apple. You learn something new every day. Well, that certainly is... less than ideal.

Apple's CarPlay is a bit more restricted (API's are by invitation only, like Apple TV) than Android Auto.
But that's Apple being Apple.
Is iOS CarPlay API Public? How to Integrate CarPlay?
My Journey on developing a Custom Android Auto app

Nevertheless, either ecosystem offers a very decent suite of apps, starting with better Nav than what Tesla offers (Google Maps or Waze - pick your favorite), endless number of IP streaming services, proper voice and SMS integration with your phone, etc, etc:
https://play.google.com/store/apps/collection/promotion_3001303_android_auto_all?clp=CioKKAoicHJvbW90aW9uXzMwMDEzMDNfYW5kcm9pZF9hdXRvX2FsbBAHGAM=:S:ANO1ljJmeV8&gsr=CiwKKgooCiJwcm9tb3Rpb25fMzAwMTMwM19hbmRyb2lkX2F1dG9fYWxsEAcYAw==:S:ANO1ljLhcXk&hl=en_US

Basically, all the effort that Tesla is wasting selectively porting those apps to their platform, it's all freely available if they were to integrate with AA or CP.

In-house writing those apps adds no value to us, the car owners, and just costs Tesla time, money, and effort.
We, the car owners, would be far better off getting full benefit of using the existing Android Auto or CarPlay application, then waiting for Tesla to replicate them, one at a time, after years of delay.

The conflict is between what's good for us, the car owners, vs. what Tesla/Elon think is good for them.
So far, Tesla has chosen to prioritize in-house DIY application development.

BMW, Mercedes, Toyota, Honda, GM, FCA, Ford, Nissan - all tried DIY-ing infotainment for many years, before figuring out that per-OEM-brand application marketplace makes no sense, and will never take off. All have recently embraced CarPlay and AndroidAuto (one, or both).

Elon is way more stubborn and less customer focused, and more interested in prioritizing in-car games and fart sounds developing.
In all fairness, my kids love that stuff.

The rest of us get to wait, and wish...

 

[Knightshade]

Normally I try to avoid getting in the middle of an argument at all costs, but this is a subject that's very important to me so I'd like to try and weigh in respectfully.

@Knightshade, the issue with the quote you posted from Elon is that it includes this important phrase:

"Even if you are going to port something, it’s still got to be worth the effort."

Applications written for Android Auto or CarPlay do not require porting. Any application written for those platforms will run, unmodified, on any vehicle that supports those platforms. This is the sticking point that @afadeev is referring to when he refers to a "walled garden".

If Tesla ends up creating a new application platform it will require developers to port their applications And there has to be a shockingly high bar to incentivize developers to do it (just ask Windows Phone).

. Again, this is what Elon appears to be referring to when he says:

"As our number of vehicles grow, it starts to potentially makes more sense to develop games and other applications for Tesla."

Err... Tesla is running Linux.

The same OS Android is based on.

It's not really a "new platform"

"porting" would mostly be adapting existing stuff for the specific controls and displays available in the car- not reinventing the wheel. And even that would be fairly minimal in many cases since mobile apps are already designed for touch screens, and usually already have code to handle being run on devices of different screen sizes.

This is also why it's been so trivial for non-Tesla developers to port nearly all the existing games so far.

All Tesla would really need do is add an SDK with some additional libraries, virtually all of that would be off-the-shelf open source stuff too, no real "development" involved.

The problem with this approach is that any application platform with a small user base will always lag behind in updates and features. Developers barely have time to support 2 operating systems as it is. Even if they agreed to support a 3rd, that 3rd OS just won't get the attention it deserves. Not so long as it's user base is dwarfed by the other two.

Again this is not, at all, what they'd be doing. "porting" an android app (already running under a modified Linux kernel- and already designed to handle touch input and multiple display sizes) to a different flavor of Linux is a MUCH smaller lift than porting it (and supporting it) on a "new OS"

Hell there's already at least a few existing Linux projects explicitly for the purpose of just straight up running native android apps within a Linux OS

If there was suddenly a market of millions of Teslas to run those apps on the relatively trivial amount of work involved would be pretty attractive to quite a few developers.

(Tesla will cross over a million vehicles sometime early next year most likely, and probably double that by end of 2020, and maybe doubling again the year after with the Y coming online)