CarPlay mostly works, and although it has issues it's definitely user-focused.
CarPlay doesn't prevent car makers from tracking vehicle activity.
In the end, consumers don't give a shit about the in-vehicle infotainment. It sucks, or it's AA/CarPlay. The first generation of iDrive showed that rich people people will buy cars in spite of the in-car stuff. In fact, most car infotainment sucks, yet people still buy cars.
Let's turn the question around: why would car makers want to spend millions of dollars a year rolling their own infotainment system? So they can make incremental revenue selling ads and user data? So they have control? Control over what, exactly?
> Let's turn the question around: why would car makers want to spend millions of dollars a year rolling their own infotainment system? So they can make incremental revenue selling ads and user data? So they have control? Control over what, exactly?
There's an interesting (and apparently often misunderstood) article called "IT Doesn't Matter" [0]. In it, Carr is largely arguing that IT, as a business differentiator, was over for many of the things people thought were differentiators. That is, things that helped a company (say American Airlines) get a lead on their competitors in the 1960s had become commoditized. Now every airline was offering flight search and booking online (directly and through aggregators). The IT edge had become table stakes, you didn't do it to beat out a competitor but just to stay in the game. And, even more importantly, many of the things that used to be IT differentiators became commoditized.
Car infotainment was once a differentiator for car manufacturers or for classes of vehicles within the same manufacturer. Today, it's table stakes. Not all the manufacturers have figured that out (have any?).
It's a bit aggravating that a slice of money I pay when buying a car goes to paying for development of the infotainment, which can be shoddy. It's like paying the Windows tax when buying a laptop with it pre-installed.
Imagine if car manufacturers started offering cars with no infotainment OS, and as users having a choice of open source OS distros to install on them..
Personally, the existence of a car infotainment system is a disincentive for me to buy the car. I certainly won't use it, and would prefer not to pay for it.
If that system is required in order to control a car's functionality (windshield wipers, etc.), then its existence is not just a disincentive, but a dealbreaker.
I wonder how much of those frustrations has to do with Japanese market share of car industry; I think the touchscreen infotainment is not there because car manufacturers value it as integral and central part of car experience but simply because otherwise their product loses against one of Japanese brands.
Japanese road network is a disorganized weighted node graph and absolutely not a grid, and a bulletproof navigation unit has been a must for a car in Japan since its inception around 1990. It is also preferred that they are 2DIN compatible so it can be later upgraded. AFAIK, those are not high priority checkbox items elsewhere, but all cars nevertheless follow the Japanese manufacturer layout because of manufacturers' collective dominance. Cars before 2DIN navigation units seem to have had 1DIN AM/FM radio units with radio buttons[1], by the way.
That dominance leaves a 4:3 8" diagonal hole in immediate view of driver for all cars globally that must be filled with something of value. That doesn't have to be a touchscreen but usually are, and it ends up being a navigation-audio combo unit, and it's outsourced to the lowest bidder. It is not the primary interaction point for cars by overwhelming global demands or principles of automotive product design. That leads to jarring subpar experience that appear to be but are perhaps not intended to be part of core UX of the whole car. I think.
I'm not quite sure what you mean by saying that the Japanese road network is not a grid. In terms of actual layout, it's quite grid-like. I would even say it's usually much closer to a nicely laid-out grid than the mess any European city is.
What Japan lacks are addresses that can be found easily without using a map. Apart from Kyoto, roads in Japanese cities don't have names (or number), so addresses within cities are not "{number} {name of street}". Cities are cut in areas smaller and smaller all the way down to a block. The last number will be the house on that block. So addresses within cities are "{name of area} {sub-area number} {block number} {house number}", with some variations from city to city.
An address might be "Nantokacho 11-16-8", which means the 8th house around the 16th block of the 11th sub-area of the Nantoka area. Good luck figuring that out without a map!
There needs to be a screen: I'm not sure what the incremental cost is of it being touch-sensitive. And I think a lot of designers (or cost cutters) may have figured that since they have to have software anyway, it may easier/cheaper to deal with software buttons than with moving parts like physical buttons and dials.
I'm pretty sure a touchscreen is cheaper than buttons, or at least awful close. You'll still get some buttons, unless the car maker really hates you, and hopefully something to turn for volume control (which may even work most of the time, if you're lucky).
Touchscreen is such a stupid low value move for driver. Physical knobs are safer in many ways, last much much more, I don't need to lose contact with whats happening in front of the car to manage these, just muscle memory. BMW has nice big physical knob (turn&press) to control all menus, much better than hunting on flat glossy screen with fingers.
Some stuff can be managed from steering wheel, but instead of putting there things like AC there are voice controls, while we all know very well this increases danger on the roads and calls shouldn't be made while driving.
That being said, I choose (used) cars based on many metrics, and this is just one of them and not the most important one. So its sort of self-inflicted move to worse.
> Physical knobs are safer in many ways, last much much more, I don't need to lose contact with whats happening in front of the car to manage these, just muscle memory.
None of which some up on a spreadsheet in the manufacturer's accounting department.
Similarly: open concept offices with hot desking lowers the amount of rent (per employee) that shows up on an expense report, but the productivity drop of open offices does not.
I used to agree, but the point I made in GP is that the touchscreen front and center could be a requirement, with what to do with one being open to local market interpretation.
IF that's the case, and that's one big if, debating touchscreen vs buttons might not be meaningful, because touchscreen MIGHT be a requirement that cannot be removed. With un-eliminable touchscreen, buttons becomes just added cost, rather than being a resilt of cost calculations between all-touch vs all-buttons.
Button-y interface can be handled on basically any IO-heavier 8-bit microcontroller with some muxes as all it needs is to handle interrupts and emit event messages over whatever-bus. The solution can even be hard real-time if there is such requirement.
Touchscreen will probably run atop whatever OS powers infotainment system. In theory touch events CAN be handled on the same 8-bit uC, but converting click coordinates to button-press messages requires knowing what is visible on the screen, which either hinders expressiveness of touch interface (fixed islands for controls) or complicates implementation (constant updating of element-coordinate maps). The moment safety-critical control runs along with non-safety software huge validation challenges arise - the performance characteristics of the whole system, including UI, must become at the very least bounded.
Hard button solution can be internally commoditized, whereas soft button solution requires constant babysitting throughout development lifecycle. One is much more expensive than the other, however, once the move is made incremental cost drops dramatically.
Don't Japanese people have smartphones or tablets with navigation on them that they can use? I'd rather cars just have a place I can mount my own device, rather than include any kind of screen whatsoever with crappy un-updated, un-maintained software.
This is sort of the goal of Android Auto and CarPlay, but not as a mount for your phone. Rather, it turns the screen into a dumb terminal for your phone, bypassing all the shitty built-in software and providing a UX designed specifically for use while driving.
And critically lets you directly access CarPlay interfaces to apps for streaming, podcasts, navigation, etc, all of which is much richer than a dumb audio only passthru with skip buttons.
Cars can have better GPS reception, which is important when you're trying to drive in an urban area with a lot of buildings. Phones are bad at accuracy there.
(CarPlay can provide car GPS to the phone though.)
It just pretends to be more accurate by using AI. A car tends to be more raw, which is what you want when navigating at speed.
As someone who used to help out with Waze maps, raw GPS tracks from phones can be WAAAY off from actual roads and other GPS tracks.
Your phone looks for the most likely road for you to be on and snaps you to that road. Your car does a similar thing but has much higher GPS resolution than your phone so it is more likely to be right.
Gosh, so the moment you use an if-statement, or an aggregate function (like sum or average), or weighting coefficients to mark some inputs as more significant than others, it's "AI" already? And I naïvely thought it was just, well, "computing". You could do "AI" on a 8086, ffs.
This is true of literally any software and if you really wanna oversimplify it, our brains are just a bunch of if statements (condition based logic) too...
no it IS more accurate. When i am off-roading i always record my track and have to unplug my phone from the trucks system to get an accurate un-smoothed track.
google maps does weird things but gaia just uses the raw GPS data from the phone and doesn't snap, however when you are plugged into carplay the phone seems to always consider the car/trucks data more accurate even when its clearly not.
this could very well my specific truck (dealer has refused to look and i don't have another to compare to) or a general tacoma thing but it both smooths the GPS data to make "nicer" looking lines and often introduces drift over time as likely some processing it does compounds an error
(i also realize i said the wrong phone, iphone 14 not 12 but even with the 12 it was notably better then the truck but the 14 pro has a much improved GPS and it shows)
AFAIK, and I could be wrong (though I did google this first), plugging your phone into CarPlay does not use the car’s gps unit. That’s just your phone trying to find a road because it’s connected to a car. Even if you are getting the car’s gps, it’s not going to give you raw values, but display values to show in an app, which should be smooth, IMHO.
But the in-car unit has both GPS (with the antenna being on the roof) as well as wheel speed. Noticeable in longer tunnels with traffic jams that make the speed vary. Then it's clear Google maps takes a random guess at where you are while the in-car system knows how far along you are.
the only time i am paying attention is when off-roading and recording a track, and then its very obvious when the GPS is doing a good or bad job and unplugged from carplay i get a far more accurate track.
Phones also use WiFi access points and some Bluetooth beacons to get a better location. AFAIK only Google and Apple have up-to-date maps of these that are useful for navigation, and no one else does (Mozilla threw the towel a year or two ago).
Cars can theoretically have better GPS antennae, and can do better dead reckoning using wheel position and wheel rotation speed. But do they? I haven’t seen any evidence that any car actually does better than a modern phone.
Especially inside a dense city, WiFi location can be much more precise than GPS.
i only really care about accurate when off-roading in the middle of nowhere with no reception and recording my track - my iphone performs far better when unplugged from the truck then when plugged into it.
That's possible too, but the issue I was thinking of wasn't exactly reception - it's that GPS signals bounce off buildings and so it looks like you're somewhere else. Stronger reception might actually make that worse.
Japan has local satellites to help with this (QZSS) and local cars may handle that better than global phones.
The Japanese navigation units in cars have been really good for at least 15 years. The signs and lane markings usually match exactly what’s ahead of you in real-life, and there’s a radio system on the highway gives traffic and road closure information even if you’re not connected to the internet. It’s only recently that smartphone apps are as good.
Phone navigation apps are surprisingly bad in Japan (Google ones in particular). Not unusable, but the in-car GPS is more competitive than in other markets IMHO.
Also with an aging population a phone screen is just too small for many.
BTW, to parent's point VW has been trying different approaches with a top mounted GPS/infotainment unit that can be omitted on cheaper trims.
That's not true. Google Maps, in particular, has by far the best route guidance. It used to be different - but not anymore. Most in-car systems don't even attempt to route the last 300m here because the detailed one-way street narrow layout of tiny streets isn't in their mapping data, or their algorithms are too weak. Google has mostly no problem with that.
Entering Japanese addresses IS tricky, though; here, custom-built Japanese solutions outshine. There are (mainly) no street names; instead, you specify your location by filtering down from Prefecture (Tokyo), City (Ota-ku), Commune (Kugahara), District (1-Chome), Block (26), House Number (1).
Japanese systems allow you to enter it this way - with Google (or, even worse, Apple Maps), it's a bit hacky. You would specify it as Kugahara 1-26-1 and hope for the best.
Google Maps fails in subtle but weird and sometimes costly ways.
On one side, it can't route through a bunch of valid paths. I first assumed it could be because of residents asking Google to cut traffic, but sometimes it's not even through residential areas. I see routes on the map that are avoided in favor of bigger loops, and when trying the shorter routes they're perfectly fine. Or perhaps it's the vehicle size and they optimize for SUVs ?
On the other side it will happily route you through paths that are restricted to specific categories of cars. It's up to the driver to carefully avoid them, but it really wants you to go through and reroutes you there when you deviate, so it's a huge PITA in areas you don't know and try to navigate while ignoring the navigation instructions. Cops seem to have noticed it, We've got fined the first time we fucked up, and now that I'm aware of the issue I see the cops in many of these spots basically waiting for the jackpot.
I notice that Google Maps does very poorly on my iPhone 13 Pro with the multi-level roads, like riding around in the tunnels under Tokyo or on the roads with highways above the local roads, whereas builtin navigation units usually have no problems with this. I also find their directions to be much harder to follow than the built-in navigation units when riding around the major roads where turning right requires you to exit to the left for instance. Also, Google Maps fails to provide the variety of route options with fine-grained toll-booth costs, that all navigation units I've used in the last 5 years have gotten spot-on.
The weird part is Maps will accept them for search, but not display the plus code except when looking at POI (randomly selecting a point on the map doesn't show it for me, I only could get that from the /pluscodes/ map)
This seems genuinely useful, but as usual we're having the chicken and egg problem to get it adopted ?
In mobile at least, you can touch and hold to make a red temporary POI that will show you the plus code.
But yeah, nobody uses it in the real World. If I need to share a point with someone else I'll just send them the Google Maps share URL instead, which has its own shortcode
> The Quasi-Zenith Satellite System (QZSS), also known as Michibiki (みちびき), is a four-satellite regional satellite navigation system and a satellite-based augmentation system developed by the Japanese government to enhance the United States-operated Global Positioning System (GPS) in the Asia-Oceania regions, with a focus on Japan.
...
> The primary purpose of QZSS is to increase the availability of GPS in Japan's numerous urban canyons, where only satellites at very high elevation can be seen. A secondary function is performance enhancement, increasing the accuracy and reliability of GPS derived navigation solutions. The Quasi-Zenith Satellites transmit signals compatible with the GPS L1C/A signal, as well as the modernized GPS L1C, L2C signal and L5 signals. This minimizes changes to existing GPS receivers.
That definitely does not match my experience. I was in Japan in December of 2022 and used my phone to get everywhere, the directions were extremely accurate, more-so than in my home town in the US.
That’s one reason I use CarPlay - the interface hasn’t changed for years. Apart from one really annoying change to Apple Maps when they replaced the “silence directions” toggle from a big square to a touch, wait, touch tiny icon interface.
Ironically, the nav unit I used on a rental car circa 2007-08 (birth of touchscreen smartphones) was exactly like the GP describes, a separate unit that had a mount like the modern smartphone variants.
I don't own my own car here, but whenever I ride taxis or rent cars in Japan I noticed that they almost all appear to have aftermarket infotainment units. This makes sense now!
Honestly I would prefer to have a 2DIN hole where I can put whatever infotainment I want, instead of being stuck with whatever the car maker included, and that I can change without changing my car.
Yeah, it can't get pro-consumer and pro-right-to-repair than that. Disappointment with lackluster stock infotainment followed by disappointment in the move away from repairability felt a bit hypocritical to me knowing that world. 2DIN holes are fine!
> why would car makers want to spend millions of dollars a year rolling their own infotainment system?
This is addressed in the article:
> There used to be a big difference in driving characteristics and technology between premium and budget brands. Compared to a Volkswagen, a BMW used to have a more powerful engine, better handling, and comfort features like seat-heating and cruise control. However, a Volkswagen Golf now has similar tech as a BMW and with the transition to EVs, drivetrains and handling won't be the same differentiator as before.
> Now, the focus has moved to the interior. Infotainment systems have become a central part of that, so carmakers are coming up with unusual concepts to set them apart, both in hardware and software. CarPlay 2 is going exactly in the opposite direction, more towards standardizing the in-car software.
But I think a great differentiator for upmarket brands would be to offer as few screens as possible and as many buttons as possible. Electronics are a plague. Some features are very useful: mainly, route planning and the ability to play music in the car. But the rest is a nightmare.
I don't want to talk to my car; certainly not when passengers are sleeping during a trip, and even not when I'm alone. There's a subtle humiliation associated with talking to a machine. More importantly, I don't want to randomly target zones on screens to set up things, and I don't want to look at the screens because I don't want to take my eyes off the road. Give me buttons that have a fixed location, and that I can feel without looking.
I’m happy with the way Mazda has integrated CarPlay. No touch while moving, the navigation wheel moves a highlight around the screen to select buttons/links. I find it much less distracting.
The audi wheel with buttons around it is terrible. I have to look down to find the correct button, and then I need to look at the screen to wheel to the correct option.
Touchscreen is not better than physical buttons but at least I can still look forward while operating it.
This audi is the first time I actually started using Siri - and not because I like it.
Don’t you think it’s possible that car makers are moving in this direction in anticipation of a day when you no longer have to be paying attention to that extent while driving? Or do people still think that day is very very far off because based on my experience with FSD we’re not as far as everyone thinks.
That's the danger of FSD. It's just good enough to trick poorly trained, unskilled drivers into thinking that they no longer have to pay attention. This works fine ... until they encounter an edge case that FSD isn't programmed (or trained) to handle.
GM's plan is to sell a subscription service that covers all the things you already pay for on your phone (maps, music streaming, etc). It's why they're killing CarPlay, because they know that even if their service is good, nobody will pay $20/mo for shit they already get for free on their phones.
Basically, their goal in life is to be a worthless middleman who takes peoples money while providing no real value to society.
How is GM a middleman when they are producing the product I am buying which is a car? If GM bundled OnStar with a top tier infotainment / navigation / ChatGpt / I’m hungry where do I eat app, I’ll pay for that as long as it’s reasonable and lets me know I’ll be taken care of if I crash and need assistance.
They are trying to be a middleman between you and the internet.
It’s analogous to a TV and monitor manufacturer saying they are no longer providing non-wifi inputs, app side loading, or supporting screen casting from local devices.
Then the manufacturer would charge a monthly fee for the monitor to keep working.
Hopefully, next, the CEO acts surprised when reporting the resulting massive drop in shipments.
Maybe for you. If it doesn't exactly mirror the stuff on my phone, though, it's distinctly worse than using my phone. If it doesn't start playing music from my phone when I turn the car on, it's distinctly worse than using my phone. If it doesn't act like my phone does today, it's distinctly worse than using my phone.
For me, there is no room for GM to become a player here. I have friends' addresses, etc. in my iPhone's contacts and they show up on Maps. GM isn't going to do that, so GM isn't going to get my business. I'm starting to look for a new car right now, and this is a hard no. If it doesn't support CarPlay and Android Auto (both--I've used both in the last five years, and might switch back again to Android in the lifetime of my next car), it's outside my observable universe.
No, that’s a half ton class truck (f-150 comparable). A 250 class truck (f-250, 2500, etc) is a 3/4 ton class and is what you need if you want to tow >10k pounds while having basically any load in the truck.
While Tesla does have a $10/m plan, you still get basic Internet without paying for it... Tesla's free is better than any other company's paid subscription.
That said, $10 is far too much for what it is, should be $5.
A Tesla costs $80k and lasts 10 years. Why are you saying $1200 for internet is too much and $600 isn’t?
Why do they even bother charging - just bundle it in with the headline price. It’s a tiny portion of the revenue, nobody is going to think “oh that $87,345 Tesla is too much I’ll go for the $86,145 one”
Either you pay with your data (Google) or upfront for the phone (Apple).
I use Organic Maps on GrapheneOS and add businesses I want which aren't there (hence I had to look them up in Google Maps in Firefox Focus, which is painful enough that I don't want to have to repeat it).
You pay for cell service on your phone, which lets you access things like Google/Apple maps, spotify streaming, etc.
GM wants to charge for a separate internet service tied to your car, that still goes over the cellular networks, in order to access these exact same features in their head unit. It is completely redundant. CarPlay already offers a near perfect experience in this regard, without charging an extra monthly fee.
Planned obsolesence. Without proper CarPlay/AA integration, car manufacturers get to decide when those whiz-bang infotainment features stop working. You'd have to replace the whole car to get those features back instead of just buying a new phone.
Heck simple bluetooth audio playback has degraded year over year in my car. After an android update a few years back I don't get to see the track name any more, pause/play sort of works, and thankfully audio still comes through and I can go to the previous/next track.
Without constant updates, software that is part a a larger ecosystem will eventually breakdown.
>Without constant updates, software that is part a a larger ecosystem will eventually breakdown.
This is why we should be so skeptical of tight software integration with durable hardware (e.g., cars and appliances with operational lifespans 10+ years easily). Software has a pretty short half-life, especially software that integrates with internet services; vulnerabilities get discovered in third-party components and remote APIs shift out from under you.
Durable goods manufacturers have little skill or interest in long-term software upkeep (maybe they like the profits and the rent-seeking, but not the actual maintenance), so the most sustainable design is one where the software is easily seperable and replaceable from the core durable item. Manufacturer-specific internet-connected infotainment in cars ain't it.
>so the most sustainable design is one where the software is easily seperable and replaceable from the core durable item.
No, the most realistically sustainable design is one where the software is built into the durable item and not connected to the internet, so it never changes and doesn't need to worry about security issues. You can keep using the thing until it fails from mechanical issues and can't be easily repaired.
If it's internet-connected for some reason, then the software should be open-source so interested parties can take over maintenance after the OEM decides they don't want to bother any more. However, I don't consider this "realistic": how many manufacturers are going to do this in reality? Almost none: they want you to buy a replacement device.
> No, the most realistically sustainable design is one where the software is built into the durable item and not connected to the internet, so it never changes and doesn't need to worry about security issues. You can keep using the thing until it fails from mechanical issues and can't be easily repaired.
My car isn't connected to the internet, it has not received any updates. However Google enjoys messing with the Android BT stack for whatever reason (there was a period in the mid 2010s where BT on Android was dramatically ~broken~ rewritten every couple of years), so accordingly my car has lost functionality.
> No, the most realistically sustainable design is one where the software is built into the durable item and not connected to the internet, so it never changes and doesn't need to worry about security issues.
How would that work in a car, where people expect map updates and streaming music or radio over the internet?
Just like that. Robust, simple software, open standards a r/w partition for map data. Having well-defined, narrow access to the outside world is much easier to secure and keep working.
I don't think it would. I was really thinking more about things like home appliances. An embedded computer is handy for a washing machine, for instance, but it doesn't need to be connected to the internet (no, notifying you it's finished isn't necessary: the machine should tell you when you press "start" approximately how long it will take).
For a car, I really don't know. I don't see any way of putting up-to-date navigation software into a dashboard without a significant security risk, which means needing the ability to keep software updated. Even CarPlay/AA are non-trivial and would need security updates.
Do people do that? I just use my phone beside the old useless infotainment. Honestly, I wouldn't buy a new car that used an in-house infotainment specifically because they go obsolete quickly (<5-10 years) and cannot be economically (or at all) to the latest tech. Carplay seems to be long lived.
You don’t like paying $300 for a map update? On my car, Nav + traffic was $3k. CarPlay was $300 and I get multiple map choices like Waze or Google maps. Only infotainment that comes close is Tesla yet map data is always out of date with construction road closures.
My vehicle (a mid teens Volvo) doesn’t support CarPlay or AA. Its infotainment system gets its data from 3G, so it doesn’t get data anymore. Music is through Bluetooth, which works well enough. Navigating means using the clunky in-car nav with no traffic data or a phone with the volume turned way up.
It’s far from ideal but at least I have zero expectations for my next car.
I replaced the head unit on my car from the same era with an aftermarket Sony XAV-9000ES unit that has wireless carplay support. It takes about 10 seconds to connect after powering up the electronics but is otherwise seamless. Full steering wheel controls too (volume, next/prev track, pause/resume, mute). No car info, but the car is old enough to have a proper gauge cluster.
There are also dash-mounted "carplay screens" that are both much cheaper and much easier to install but lack integrations/require hacky solutions to get, say, audio out.
Auto makers want what everyone else wants: recurring revenue. They want to find a way to sell subscriptions to something. The infotainment system is a potential angle for that. CarPlay makes that irrelevant.
> "So they have control? Control over what, exactly?"
I think a key thing to consider is that there are in fact three separate questions at play here:
1 - Does infotainment/software UI differentiation matter in the car market? Is there a significant enough market advantage for having better UI that anyone should care?
2 - If there is an advantage for better UI, is it enough of an edge that would compel you to build your own? Or is it the case where it simply has to be good enough?
3 - If there is enough differentiation to be worth building your own, is your company good enough at software to pull it off?
Personally I think the answer to #1 is YES. I think cars with better UIs - while not sufficient in and of itself - have a market advantage.
Where car makers start veering off from each other is the answer to #2. If you believe that you just need a "good enough" experience to not be actively awful, then you buy off-the-shelf. You see this with Volvo/Polestar and Google Automotive. The "skin" around the stock experience is minimal at best, with only minor customizations.
If you believe that being excellent at it confers some advantage, you'd try to roll your own. This would include folks like BMW and Mercedes-Benz.
Now, where the latter strategy really goes off the rails is question #3. That said, if you believe the answer to the first two questions compels you to roll your own - would you easily surrender to a third-party? Or would you at least try to level up your software orgs to make a serious play?
For what it’s worth I agree on #1. I really like RR/Jaguar’s current UI. When I’m in a rental or someone else’s car and I’m forced to use CarPlay I hate it. Feels like I’ve been pushed to kindergarten and given crayons… any car manufacturer that just expects me to use CarPlay is probably not on my potential buy list.
I might be unusual in my preference but I really expect people to have preferences as strong.
Android Automotive is already way beyond what manufacturers are releasing as "their own".
So unless it costs a lot of money to integrate, not picking AA isn't about building your own that's better, but it has to be about control. Or at least independance from a big tech, which I can understand.
I thought about that as well. At the moment, CarPlay can't take over every function so carmakers still have to make their own. But in the long term, there is an opportunity for Apple to make the whole thing and monetize it as a cost-saver to carmakers
When (if?) auto-driven cars become a thing I'd expect some people will want a giant monitor in their car if nothing else, so they can watch movies, play video games, tele-conference, etc....
Disney's "The Magic Highway" might be dystopian to some but the relaxing vehicle seems pretty cool to me
The auto industry is really at a crucial point with how it integrates tech. As cars get more autonomous, infotainment is becoming a key part of our driving experience. I'm wondering, are carmakers at risk of falling behind if they don’t embrace platforms like CarPlay? Or do they have solid reasons to keep developing their systems in-house to keep control over their tech narrative?
Yeah it's CarPlay or whatever Android uses, or I want nothing to do with it.
Every time I rent a car it's a HELLSCAPE of figuring out whatever crappy UI this brand of car created for that year ... until I get my phone hooked up. Man I just want to get to my hotel not futz with some garbage UI in the garage forever.
> Let's turn the question around: why would car makers want to...
The tech giants are not component suppliers you can symbiotically partner with to add value to your product. They are predatory and parasitic goliaths. They are wolves more powerful than governments with designs on your hen-house. It is frankly insane to let them own the primary interface to your customer and auto will likely regret letting them get this far.
Google/Apple infotainment ventures cannot even be called trojan horses given how open they are in their investments and desires to seize the auto industry the moment tech/profit makes it feasible. Unattractive low-margin manufacturing keeps them at bay but they are gambling on "software eats the world" long-term. For now, they will drain every available high margin service for themselves.
>In the end, consumers don't give a shit about the in-vehicle infotainment. It sucks, or it's AA/CarPlay.
My ideal "infotainment" is a button that lets me pair bluetooth and volume/skip controls on my steering wheel. I have a phone mount, like almost everyone does. I don't need a display on my dashboard.
AA/CarPlay (and everything else) are genuinely just distracting annoyances that take up space. Especially in newer cars where the screens are for some reason, no longer matte and often angled upwards so that they just blast reflected sun into your eyes.
Likewise. I vastly prefer the physical buttons, as I can hit them without taking my eyes off the road. The only two things I use infotainment for is mapping and music control.
Bluetooth should let you select phone audio sources from your car buttons.
It’s usually terribly implemented. iOS doesn’t support it for third party apps or first party podcasts, and apple music’s support is a shit show. Android is better, but it is hit-or-miss depending on the car.
I wish the government would declare this a safety feature and mandate recalls for incorrect implementations (including phones and cars).
If find the Bluetooth interface too limited because it requires interacting with the touchscreen on the phone. I want to use the large screen on the car with an interface that is optimized to work while in the car not small controls on a small phone.
No touch screen i
should be accessable by the driner unless the car is in park. I don't care if it is the phone or infotainment, no touchscreen is allowed.
The person I was replying to was arguing for a Bluetooth connection to their phone instead of CarPlay. CarPlay is much better from a safety standpoint than a phone screen.
You just choose a superlong playlist before starting the car and then play it randomly. You don't have to interact with the phone.
Last time I had a car I had an old mp3 portable mp3 player constantly hooked up to the car with 4GB of music inside. I would just hit play and drive and would only adjust volume from the steering wheel if I needed to. If I had a car again I would probably do the same, that is the lowest distracting way to have music in your car.
Top of my head, most car makers aren't rolling their own. It's either off the shelf with some white labeling or they buy it from another care manufacturer. i.e. Mazda default infotainment can be found in some Toyotas
Not sure if there's any more to it than this, but the Toyota Yaris is a rebadged Mazda 2. I suspect you won't find Mazda infotainment in other Toyota models, though.
I mean, theoretically they could make money by selling additional services. Tesla sells "Premium Connectivity" for ~$10/mo, though it might be against the license agreements traditional manufacturers have with dealers to sell enhancements directly to the consumer.
IMO it's more about control over the user experience. You don't want your customers' UX to be dependent on the whims of Apple or Google, because now you're implicitly building a long-term dependency with a third party that may not be acting in your interests in the future. You're moving closer toward a future where the vehicle becomes commoditized, and now you have more trouble differentiating from competitors. And keep in mind: it's only very recently that the "Apple car" project was cancelled.
That said, traditional automakers are also famous (or infamous?) for sourcing tons of components (including infotainment systems) from the same parts manufacturers. But I guess at least that retains the ability to pivot and use it as a point of differentiation in the future.
Historically, Hyundai/Genesis charged $300/yr for BlueLink connected care. Think my 2023 Ioniq 5 is one of the last (EV) models with that yearly charge, and that year's Ioniq 6 and subsequent models get it free.
Unfortunately, I don't think they're planning on offering it free to my model or earlier, and I don't know if the same deal exists for ICE cars.
If it was user-focused, they wouldn't make unnecessary changes to the UI every release. If it was user-focused, they would put more effort into refinement and fixing bugs.
Regarding "it sucks", if we're talking mapping/directions I disagree. Google maps is really a pretty piss poor application. It hasn't changed in 15 years and it's obvious its maintainers haven't ever driven anywhere using the application, even around the Google campus in MT View, or downtown SF. It's deeply bad. In my experience some car manufacturer mapping applications are quite a bit better. Since they obviously suck at software, who knows how good it could get with the combination of (not Google) AND (competent team)?
I use Apple rather than Google as I try to avoid Google as much as possible but it’s the same principle. I want long term stability in software, not new changes put in to earn a PM a promotion.
>> In the end, consumers don't give a shit about the in-vehicle infotainment.
Drivers don't car, but I think car buyers actually do. Remember that many people are not buying cars for themselves but for other people, usually family. They fall into the trap of thinking that those other people might want such features, if not now then in the future. Look at automatic transmissions. I know many people who much prefer manuals, but they always end up buying an automatic because they believe that other people will want the automatic. And a few years later, all the cars are automatics. The same is happening with in-car entertainment systems. We buy them not because we ant them but because we think other people do.
Unless you are single, and even then you carry occasional passengers, many/most cars are used by multiple people. So people who buy cars are thinking not just about the primary driver but about all the other people who will drive/ride in the car too. Nobody buying a car actually wants in-dash 4k movies, but they think that their partner/kids will.
This doesn't really ring true for me. I agree that when people buy cars, they are often not just buying for themselves. But the purchase decision is not usually made without input from the other drivers. Well, a purchaser will probably discuss things with their spouse, and ask if they care about particular features. But in the case of someone buying a car for their kid, it's usually "they'll get what they get, and they'll like it".
Kids are in the back and thus a screen in the back is what’s needed. An iPad or similar will do that, so a place to mount would be good.
By the time they are old enough to be in the front they have their own phone and headphones and care less about a mounted display than they do about a TV.
It's true for some features (the wife (and I, to be fair) were pretty strongly on the "heated seats" side of things) but things like infotainment were not dealbreakers if everything else lined up.
But CarPlay is darn close; I'd not say I'd never buy a car without it but having it means I don't need to worry if the infotainment setup is crap or not, because I won't be using it.
There's an interesting gap when discussing carplay/aa.
Cars from the 90s up until about 2013 can be easily fitted with a $500 head unit upgrade, and support carplay quite well. With the right tools, it can be done in about two hours right in your driveway.
Cars from 2018 and up pretty universally support carplay, and it's generally quite well integrated into the car's infotainment system.
But, between 2013 and 2017, things were a complete mess. In-car systems were too integrated to be replaceable with a third party 2-DIN unit, but too primitive to run carplay/AA. People who have cars from this era either sell them (for less than they're worth, since only 21% of people will buy a car without carplay!) or put up with it for another 8 years or so until the car's wound out.
For example, my rustbucket '06 Toyota has a great sounding stereo with carplay but my sibling's 2017 Nissan is stuck with flaky and poorly integrated bluetooth.
Or, if you do want to upgrade your 2013-2017 car, you end up replacing half of the in-dash components with ones from a couple model years up, tapping into the car's CAN bus to recognize the new controller, and then running some sketchy scripts to patch the firmware to remove component protection since the VIN's don't match up anymore. Not for the faint of heart.
> But, between 2013 and 2017, things were a complete mess. In-car systems were too integrated to be replaceable with a third party 2-DIN unit, but too primitive to run carplay/AA
Mazda sells a retrofit kit for their 2014 (Mazda 3) / 2016 (other models) and newer cars that didn't come with native CP/AA. It's about $200 and DIY if you're the least bit handy:
I have installed that unit but I wouldn't characterize it as for the "least bit handy". It involves removing all the panels from the screen, center console, and shifter, and fishing some wiring behind the climate controls. Definitely do-able, but not for the faint of heart. The wiring is easy to connect, to be sure, Japanese are outstanding at that. Totally worth the money. I just wish I could put in one of the factory backup cameras, neither the dealer nor my trusted local stereo shop will do that for me.
You also have to update the software first, which is a 45 minute process that requires touching the brake pedal every 20 minutes to prevent the head unit from going to sleep.
I've installed it twice: a few years ago in a CX-9 and two-weeks ago in a Mazda 3. There's YouTube videos for it so you know what you're getting yourself into. I found the procedure straight-forward. Mostly you need to know which parts to pull off and where the bolts, clips and fasteners are. If you're someone who'd install their own dash cam, you can probably do it.
Having done this recently, the pro tip is that you can pop out the old USB/SD panel by inserting a hook pick into the holes on the front and yanking. It will pop the plastic faceplate off, and then you can access the clips to remove the entire unit. Mine was actually fine and I was able to pop the faceplate back on if I wanted to reuse it (not sure why I would), but no guarantee.
Turned it into a 15 minute job vs probably an hour+ of messing with the center console. Babysitting the software update takes the most time.
Very much depends on the model, trim level, and generation. Even on the Mazda 3, the location of the USB hub depends on whether the car has a CD player.
I did the "drill holes in the front of the USB hub" trick so I didn't have to remove the console. But then I dropped a socket and ended up having to remove the console to retrieve it and it only took maybe 15 minutes. I could do it again in 5 knowing all the steps.
For other brands, there are third party retrofits that replace the integrated stereo with a DIN mount, or with a carplay compatible head unit. The software on those is usually clunky, but it's not like OEM car head units are known for their UX.
I replaced my 2014 CX-5 head unit with a 3rd party unit which supports CarPlay/AA, integrates with the steering controls, the factory Bose amplifier and retains the ability to change the car settings through the head unit. Took about 2 1/2 hours to install it.
Which head unit is that? I looked into it years ago but I definitely couldn't find one that still allowed you to change the car settings and I felt that was a bit of non-starter.
It is a Kenwood. The integration is done through an iDatalink Maesteo RR unit. They work with a wide range of cars and head units not just Kenwood. When you get the iDatalink module you connect it to your computer via usb, tell it which model car, head unit and wiring harness you have. It programs it for that combination. If you go to Crutchfield and tell it which car you have it will display a list of head units which will work with your car. Once you select a head unit it will present you with an installation package needed. The packages include the 2-din adapter, trim pieces, wiring harnesses and the iDatalink unit. It will also tell you what functionality the kit will allow. You can add it all to your order with a single click. I also upgraded the head unit in my 2017 Outback this way. You will need to run a new microphone because the one which comes with the car is powered. It took me 10 minutes to replace the factory one and run the cable through the headliner and down the pillar. Super easy. You will also need to replace the factory gps antenna. The factory one is located with the middle front speaker which is directly above where the head unit goes. Your new head unit will come with a microphone and gps antenna.
I believe other car audio sites will similarly present you with installation packages. I’ve been using Crutchfield since I replaced the stereo in my 1st car in 1985 so that is who I went with.
Thank you. This is helpful information on how it works; I did a bunch of googling after your last comment (and went to Crutchfield) but not much explanation how it could work.
Sorry if I got your hopes up. I edited my comment to be more accurate. I didn't realize 2014 only applied to the Mazda 3 and it's 2016 for the other models.
Porsche has a reasonably priced upgrade path for their head units even in quite old cars which add Carplay. The value is obvious to them - it maintains a high resale value on older vehicles which maintains their higher current sale value.
Something I found interesting when I was hacking my VW to add Carplay (without paying £300) was that a ton of manufacturers in the ~2016 era use the exact same head unit and OS but with a different front fascia, different button layout and a relatively advanced UI skinning system. If you can be bothered you can add e.g. Audi track G-sensors to your VW or use a different skin.
I think you’re finding “different” car manufacturers using that system because they’re all VW owned brands. You’ll find the rest of the parts on the car are mostly the same also. If you want to annoy an Audi owner, point out which VW model is 95 percent identical… in some cases like the Q7 vs VW Touareg, the VW model is even higher end and nicer.
People like to complain about touch screens and lack of buttons, but the loss of the double DIN slot is a much worse loss. It more or less forces you to buy a new car to get satellite navigation updates or properly functioning Bluetooth.
Agreed. These days you can get a double DIN carplay/aa deck with a slot load DVD player for under $80 USD. That is absolutely shocking. I mean, we couldn't even get a 5" slot load CD drive for our computers for that kind of money and these things have a 5" screen on them too!
We had a 2014 that suffered from this. It some touchscreen that was basically useless besides showing the backup cam and a few random, infrequently changed settings. However, replacing it was essentially impossible it also served as a critical system for a few things.
By contrast, my 98 Jeep and '08 sedan were both easy upgrades to modern head units with standard DINs.
Not having CarPlay/AA was actually a massive reason I wanted to get rid of that car.
I have a 2014 Mazda CX-5 and a 2017 Subaru Outback. I have replaced the head units in both with 3rd party units which support CarPlay/AA. Both installations were easy and took about 2-3 hours. Both support changing the car settings through the 3rd party headhunt, integrate with the steering wheel controls, factory backup camera and the factory installed amplifiers (Bose and Harmon Kardon). The key to these integrations were iDatalink Maestro units. I ordered mine through Crutchfield but most decent car audio websites will generate an installation package including the appropriate wire harnesses, iDatalink unit, 2-DIN adapter and trim pieces once you select a head unit to purchase.
> Both support changing the car settings through the 3rd party headhunt, integrate with the steering wheel controls, factory backup camera and the factory installed amplifiers (Bose and Harmon Kardon)
Wow, I'm impressed. I wouldn't have thought manufacturers nowadays would be forthcoming enough to allow that.
Do you know if it is something specific to (some?) Japanese manufacturers, or is it a form of info-display standard that would also make it possible with other brands?
I know the iDatalink Maestro works with some European and US cars also. I don't know that there is any kind of standard. I suspect the unit has reversed engineered the canbus messages for the different manufacturers. Before installing the unit you have to program it via usb. They provide software for Windows and Mac. You select your vehicle, wiring harness model and head unit model and it programs it appropriately. It also downloads pdfs with instructions for installation including wiring harness connections. For my Mazda no splicing of any of the factory wiring was necessary. For the Subaru I had to splice a single wire for the Starlink (similar to Onstar and not SpaceX Starlink related) unit.
3rd party headhunts from various manufacturers have a port for the iDatalink unit which allows it to integrate with the head unit. Both of mine are Kenwood but I know a few others have support. There are also alternatives to the iDatalink Maestro it is just what I have used.
> Cars from 2018 and up pretty universally support Carplay, and it's generally quite well integrated into the car's infotainment system.
I don't know what cars you drive, but they are not budget or Chinese. I recently rented a bottom-of-the-line Nissan '24 and just getting audio from my iPhone was a struggle - and there was absolutely no Carplay in sight. Most Chinese cars I've rented (all 23 or 24 models) have advertised Carplay, but what they really have is a half-baked phone mirroring, which maybe worked on ios11 or something, but definetely doesn't today.
I've got a cheap 2021 Nissan Versa, and it has CarPlay plus a bunch of useful electronics features (front/back proximity detection, auto-dipping headlights, etc).
I'm interested in where this data is sourced: "only 21% of people will buy a car without carplay". I find that really hard to believe even if the survey was isolated to SF.
I am in that 21%. Heck, I would MUCH prefer to buy a car with no touch screens at all. All I want is a decent car ala 2004ish but with modern crash safety and headlights that don’t suck. I do not care about Bluetooth even!
I realize I will not be able to buy a car with no screens due to the backup camera rules. But there is actual safety value in that for me so I’m ok with it.
My phone works great for driving directions. I plug it in and turn up the volume when I'm driving by myself. I plug it in and turn off the volume and give it to another human to tell me what to do when I have passengers I can trust to navigate for me. I only really need to use the phone for navigation like 2 or 3 times per month, at most. My daily commute to the office never has traffic and in nicer weather times I often ride my bike.
I wonder if that assertion is based on new car sales.
For example, I wouldn't buy a new car without android auto, because cars with it are readily available and hardly cost anything extra.
But if I had say $6k to spend on a used car instead of $20k+ on a new car? Your options are much more limited and the tradeoffs unpalatable. I know I can still put a phone holder in any vehicle and I'm off to the races.
> People buying used cars will take Bluetooth as a nice perk, but to think they make major decisions based on carplay in a 2015 Chevy Volt is funny.
That was my position until we bought our last car. We got it second hand, so we did not really have a definite set of required features and CarPlay was just nice to have. After having used it over the last 4 years, it completely changed though, and I can guarantee I won’t consider a car without it for my next one. It helps that around here all non-garbage car have had it for quite a few years now.
Nowadays there are low cost Chinese kits that directly add AA/Carplay to most of those early integrated systems without even replacing the unit. They’ve reverse engineered the lcd and touch screen interfaces so they can hijack the screen.
>(for less than they're worth, since only 21% of people will buy a car without carplay!)
...is there a handy list of which models these are? I'm in the market for a used car, and I'm perfectly happy to pay less for something without CarPlay.
Article is really about the unofficially called "Carplay 2", a deeper integration that Apple announced in 2022 but hasn't been heard about since.
As the article itself says, CarPlay Original Flavor is a massive success, I'm in the "79% of drivers only consider a car if it has CarPlay" (bye-bye GM!)
Apparently this isn't true. If you have a car where the 2nd screen behind the steering wheel also shows CarPlay-related stuff, you have CarPlay 2 (or so I've been told). It's just not as invasive and all-or-nothing as it was billed to be.
No. There are three things, which makes it confusing.
1: “original” CarPlay. Shows up on your head unit
2: enhanced “original” CarPlay. Some cars can show the map or turn by turn instructions behind the steering wheel on a screen there
3: “new” CarPlay. This is where CarPlay fully takes over every screen. Nothing has been released, I think only one very expensive model has even said they’re going it.
Today e dry one has #1 or #2. The article is about a mix of #3 and maybe a version of #3 run on the car not the phone, like Android Automotive.
The lack of distinct names for things on both the Apple side and Google side (auto vs automotive) just makes this all very hard to discuss at times.
Any idea when #2 was released? I have a 2019 Honda. Android Auto can use the instrument panel as a second screen but CarPlay cannot. I miss seeing the next-turn distance countdown right in front of me.
I know it requires both sides to support it. My car originally didn’t but a later update to my car enabled it.
I don’t know when the first car got that, it’s possible it simply didn’t exist when yours was made. Or maybe it did and they just didn’t bother? I don’t know.
My car doesn’t have the version where it just shows Apple Maps, which I would love. Instead the next turn/etc. is clearly being sent to the cars operating system because it gets displayed the same way the cars internal navigation stuff does on the screen behind the wheel.
To be fair, at this point only the Ultium EV-based models won't have CarPlay, and they are struggling to build them anyway. Even if they do figure out the production issues and achieve the targets of their wildest dreams, they will still only account for around 6% of all the vehicles GM produces. They're not exactly betting the farm on it.
>Continuing with the above example, let's imagine I just downloaded a new podcast app on my phone (one that is also available in the automotive Play Store). The next day I have to go on a long drive and I want to listen to a podcast episode I downloaded on my phone. When I enter my car, I have to go to the Play Store, find the app, download it, log in, and then download the episode. When I use CarPlay, I only have to connect my phone.
Does the author have no experience with Android Auto? The same happens there, if the app is on my phone and it supports Android Auto then it will automatically be available in the car, along with all media on my phone. This isn't a CarPlay only functionalitiy, it's just how phone mirroring works on both platforms. Author seems to think there is an extra step involved on Android. Perhaps by "Play Store" they mean the car manufacturer's own app store?
Source: I use Android Auto constantly in my own vehicle and in the 15+ rental cars I have every year.
Alright, but then comparing CarPlay 1 to Android Automotive is comparing apples and oranges. The author should be comparing CarPlay 1 to Android Auto and CarPlay 2 to Android Automotive.
Sure but I’m just providing more info to the other person? I’m not defending the author. I think many of their criticisms on both sides have flawed arguments and misconceptions .
Ummm, no? Sort of. First we don’t really know much about CarPlay 2. It was announced in 2022 but hasn’t been detailed yet. However, all the assumptions are that it runs of your phone just like CarPlay 1, but with more control and integration in the car. Nothing I have seen from Apple or others suggests that Apple is building an Android Automotive-like product. I.e: Apple software that runs off the car itself. Also it will be very unlike Apple to sell software that car companies run, which is what Android Automotive is. Unless the plan is to sell an iOS box that car manufacturers integrate into the car. But I haven’t seen any thing that suggests that.
> Android Automotive which is different than Android Auto
I know this isn't the highest value comment in the world, but god Google's product naming/differentiation is terrible. I can't imagine a more guaranteed-to-confuse (or worse, more guaranteed to lead to misleading second-party marketing) nomenclature.
Android Automotive (AA) is obviously different from Android Auto (AA). You see, AA runs on the car and runs the infotainment system whereas AA is just screen mirroring for Android phones. AA supports CarPlay just like it supports AA. Apple doesn’t have anything like AA, CarPlay is equivalent to AA.
Lastly, it’s not like “Auto” is short for another word such that people would mistake the two. It’s clear as day.
I’m sorry, I’m not clear on what you are talking about. Apple TV, the device? Apple TV, the app? Or Apple TV+, the streaming service? :)
It makes for great sentences like: The other day I was using Apple TV on my Apple TV to watch a show on Apple TV+. Which just rolls of the tongue and is sure to cause no confusion. /s
>> Continuing with the above example, let's imagine I just downloaded a new podcast app on my phone (one that is also available in the [Android A]utomotive Play Store).
They lost a capitalization which makes it a bit ambiguous but their scenario is this:
There are two app stores being discussed: iOS (for CarPlay), and Android Automotive (the infotainment system's play store).
If you have Android Automotive on an infotainment system and an iPhone and can't connect your iPhone to the infotainment system, you have to download the app twice: iOS App Store and Android Automotive Play Store.
I don't know the numbers, but Android Automotive infotainment systems don't universally support CarPlay. Some only got it via updates over the last couple years (that's also a selling point of them, though, that they could do it via software updates and not a whole hardware refresh).
There's a difference between Android Auto (Google's CarPlay equivalent for phone projection to any infotainment system) and Android Automotive (Google's Infotainment OS).
It's pretty funny to me to see an Apple user complaining about lack of interop with other ecosystems. You've chosen your walled garden; now it's time to realize you're stuck in it.
The author seems to blame everyone but Apple for this, though. Apparently it's Google's fault that his preferred podcast app isn't available on Android. What about the app developer, who has decided not to support Android? And it's the auto manufacturer's fault for deciding to go with Android Automotive as the infotainment OS. What about Apple, for refusing to license any of their OS offerings to third parties? (Let alone not having an automotive-suitable version of any of their OSes anyway.)
Instead, Apple wants auto manufacturers to implement a complex, proprietary Apple protocol to give iPhones the ability to control most aspects of the car's infotainment functions. Why would auto manufacturers want to do this? Not only would the auto manufacturer have to develop their own infotainment system (perhaps even using Android Automotive, but perhaps something else), but then they have to do all this extra work to essentially allow drivers to not use their own system. (And while I agree that most auto manufacturer software is weak or outright garbage, that point of fact isn't really relevant here.)
I don't agree with the author that some car makers are not excited about CarPlay 2 primarily because they want to own the automotive experience. It's because of all the extra work they need to do to support CarPlay 2 (probably an order of magnitude what it takes to support CarPlay 1 or Android Auto), and doing so isn't really in their interests, outside of Apple users who might refuse to consider their cars without CarPlay 2 support. But this is a chicken-and-egg problem: CP2 doesn't exist yet in any car, so customers don't know what they're potentially missing.
> It's pretty funny to me to see an Apple user complaining about lack of interop with other ecosystems. You've chosen your walled garden; now it's time to realize you're stuck in it.
I am still amazed that some people see it as a brain-free non-decision to buy into a status symbol.
I have an iPhone. It has things that I like and things that I hate. I have a brain that can process nuance, and it’s entirely possible to say nice things about some features and rant against others.
I bought my current phone knowing what I can do and cannot do with it. It is an informed decision and a compromise by necessity. I bought it expecting it to integrate with my car via CarPlay, and I would definitely complain if it either stopped working or if car manufacturers dropped it.
> and doing so isn't really in their interests, outside of Apple users who might refuse to consider their cars without CarPlay 2 support.
I would be interested in CarPlay 2 if I were buying a new car when it’s out. Infotainment is still a huge weakness in many cars, and is definitely a plus to many users. Carmakers are free to ignore it of course, but unless they are maybe Tesla, they have no good reputation to rely on in selling their own custom infotainment experiences.
> You've chosen your walled garden; now it's time to realize you're stuck in it.
Yes and no. Android Automotive supports android apps, but GM's implementation (for example) still locks out android phones from using Android Auto. Yes, you can install the same apps, but the nice thing about hooking up your phone is that you're maintaining all of your apps, data, and state in one place instead of two and it's inherently different for as many different drivers as there are phones.
> And it's the auto manufacturer's fault for deciding to go with Android Automotive as the infotainment OS.
Yes, as far as I know, any automaker can choose to make the screen in the dash compatible with Carplay, at least the non fancy Carplay that will be able to show all the apps on the phone.
Zero reason, in my opinion, to not offer a Carplay compatible screen, other than to inconvenience car users. Not the CarPlay 2 thing, but just original Carplay.
CarPlay 2 needs a deeper integration to the car, as part of the vehicle's SDV stack. From what I understand, Apple will license software that runs on the vehicles SDV stack (RTOS, etc) with support for various Auto specific SoCs (Nvidia, Qualcomm, Renesas, NXP, etc).
There is an existing vibrant ecosystem of Tier 1 partners who can do the CarPlay (or AA) integration, lessening the burden on the Auto OEMs for such specialised integration work (Luxoft, etc).
CP2 is an eventual pathway (pure conjecture here) for Apple to offer other upgrades in the future, like an L2+ ADAS package, via a subscription model, with a potential profit split between Apple and the OEM.
Somewhat like what MobilEye currently offers and their current business model. Apple gets valuable realtime traffic info from such an L2+ offering, which can feed into making Apple Maps better with up to the minute traffic updates. And the OEM gets to leverage a valuable "sticky" offering that is bound to attract the Apple crowed (usually well heeled folks with deeper pockets). And they both get to monetize CP2 via a paid subscription model.
I also think the author misses the mark on this statement
> Apple’s long-term goal is likely to find ways to directly monetize CarPlay. It could license CarPlay to carmakers at some point. But with the entire global car market amounting to only around 70 million cars per year and the car industry’s slim profit margins (ranging from 5% to 15%), it must reach a high market share to get significant revenue from it. However, carmakers may be hesitant to invest in what they see as a competing product.
I don’t think Apple sees this as a promising revenue stream. It’s just another way to encourage people to stay within the walled garden. That seems many times more valuable than any possible licensing fees they would get from manufacturers
> I don’t think Apple sees this as a promising revenue stream.
Isn't the value of CarPlay/AndroidAuto indirect? As in that the value is coming from the data being generated to enable better mapping? Because the truth is that some of these mapping services' main utilities depend on a large network of users (e.g. information about congestion, road notices, etc[0]). In that way it seems beneficial to get your app onto as many vehicles as possible and even gives motivation for making CarPlay cross-platform (even if the main benefits only go to Apple users).
[0] Also, wtf AndroidAuto. You pull data from Waze users for things like speed traps, road hazzards, etc, but you don't let me contribute when using Auto? Why is there no button (or voice command) "report road hazard"? Hell, this'll even help you classify events for your ML models.
The information you are referencing could be retrieved without CarPlay though. I am sure car specific data has some value to Apple, but given they aren’t really in the car business it might not be the primary driver/motivation here. It’s probably more likely that the data was way more valuable when they were building their own car but now that has been scrapped perhaps not as much?
Wait I gotta be missing something. Are seat warmers a subscription now?
Damn, indeed I missed that section
> With their own software platforms, they are desperately trying to find recurring revenue streams, whether it's a subscription for seat heating or driver assistance systems
That’s called “leasing a car”, and it comes with a limited number of miles per year, beyond which you pay $X/per 100 miles iirc. I wouldn’t know the details, because I bought my car instead of leasing, but I remember looking at lease terms that all had a mileage limit +extra pay per N number of miles.
> I don’t think Apple sees this as a promising revenue stream. It’s just another way to encourage people to stay within the walled garden. That seems many times more valuable than any possible licensing fees they would get from manufacturers
I agree this is the optimal way of seeing it. Apple did not do the same with App Store though. The initial vision was to just apply enough fees for hosting and vetting the apps. Now with the way they hold on to the 15/17/30% fee, they are treating it like a revenue stream instead of a "courtesy service".
Pretty much all new car reviews are complaining about the built in infotainment, and it only gotten worse after manufacturers decided to jump on the all out touch screen bandwagon (to save a few bucks) but without Tesla’s generally alright UI and responsiveness.
In that respect a UI made by Apple will certainly appeal to an audience already demanding current CarPlay. On the other hand, it could be that the current level of CarPlay integration is already enough.
> give iPhones the ability to control most aspects of the car's infotainment functions
Because the solutions the auto makers came up with _sucked_ and as it’s not core to the operation of the car, I’d prefer it be done by someone that even approximately knows what they’re doing with software?
I am not buying the car for their infotainment features, I’m buying it for the car. I’d like it the entertainment bits were just delegated to the setup I already have…
That's why I'm happy with CarPlay/AA for infotainment... ONLY. Maps, music, podcasts, that's all I need.
I don't need my whole car UI to be taken over by my phone. My car maker solution to display and control speed, HVAC, gas tank/battery status is perfectly fine.
personally, i am hoping for (and working slowly toward) modernized AACS so that eg. a raspberry pi can be used to control most modern car head units unmodified: https://github.com/tomasz-grobelny/AACS
Not if it's anything like the video standardization efforts. The result is royalty free and open source. Every provider and consumer benefits when the infra is open
That’s a ridiculous statement - he was responding to substantive points raised by the article. I’m a huge Apple fan, but let’s be fair here and not flame others.
The only reason I use CarPlay is because I have an iPhone. The experience with Android auto is way ahead. Did you ever try to call a business that’s not in your contacts? Did you ever try to ask a simple question? Siri integration is horrible
I feel like we should separate Voice Assistants from infotainment in general here.
Yes, Siri is terrible, but the EUs new DMA law requires that I be able to select a different voice assistant on iOS too, and that should theoretically, once implemented, also apply to CarPlay.
In my opinion, and while I haven't used "Android Auto" I have used "Android Automotive" (the OS version, without a phone driving the integration), CarPlay the infotainment integration (besides Siri) offers a better user experience.
Siri, in general, is horrible. Just ask Siri: "set an alarm for 12:00 UTC" or "set an alarm for 12:00 Eastern Standard Time" and she will create an alarm named UTC or Eastern Standard Time ... but it's actually in your current timezone.
Super annoying when you want to set an alarm in a timezone you are going to be traveling to.
I much rather have car manufacturers creating ‘exclusive to their brand’ themes and skins for the CarPlay 2 dials than not having car play (2) at all.
Car manufacturers have had decades to develop their software but almost all are difficult to work with, slow, give bad user feedback (full screen pop-up with entire paragraphs while navigating) and are cumbersome to get accustomed to when switching cars.
I’m one of the 79% that demands CarPlay, if it’s not part of the car I don’t want it.
Having had to do with car executives, especially in Europe, let me tell you... they're a peculiar lot.
The older ones confuse their old school pride for business acumen more often than not, because of survival bias from an era when the two things actually coincided.
The younger ones are frustrated by the fact that older upper management is extremely conservative and they're often put down or given toy projects to tinker with.
On the other hand, incumbent carmakers might have also seen how Apple was able to turn the carriers' business around by convincing Cingular/AT&T about giving up their control and they don't like the idea of ending up like the carriers, with overarching licensing fees for keeping on using the tech...
A bit of a tangent, but those "distinctive" interior mockups made me instantly cringe. I'm fine with designers having a bit fun with every part of the car other than the top of the dashboard, which should be flat, dull, and dark (ideally covered in something like Vantablack), because any bit of visual clutter there gets superimposed over the road in front of you thanks to the reflection of the windshield in sunny weather.
> But it won't be possible to sell a subscription for a feature when all is controlled by Apple CarPlay
Good I bought the car, now get out of my life. I’m not paying hundreds of dollars for GPS map updates when I can plug in my phone and use any other mapping app’s evergreen maps. Also, subscriptions for heated seats are patently absurd.
Ugh, CarPlay 1 is exactly the division of responsibilities I want. I don’t want Apple anywhere near the actual driving functions and instruments. It takes a long time for my car to connect to CarPlay and it’d be super distracting to have the instruments suddenly change to Apple’s as I’m driving.
If you're in the market for a new-to-you car, then I recommend looking at the model years where they switch from wired car-play to wireless car-play and buying the previous model-year.
Typically this difference of one model-year can add thousands to the cost of the vehicle, especially because wireless car-play is so coveted. The experience of wireless is fantastic, but is it worth several thousand dollars? Maybe, but herein lies the trick.
Buy a dongle. They're about $100 for a good one. They can be tucked away in the vehicle. They work almost* as good as integrated wireless car-play.
*Maybe add 5 seconds to auto-connect when you get in your car.
I have one wired and one wireless vehicle. I will absolutely pay more for integrated wireless. It just works flawlessly.
We've upgraded the wired one with one of those dongles. It mostly works, but has some quirks. The three most annoying:
* Phone calls result in a feedback loop for the other end. Essentially, they break the in-car noise cancellation and playback the caller audio to the caller.
* When my wife pulls in the garage, my phone will connect - even though it should have been connected to her phone.
* The USB port that connects to the head unit remains power for a period of time after the vehicle is off. Annoying when I'm in the kitchen (next to the garage) and my phone keeps trying to CarPlay.
EDIT: I'm also realizing that I believe the car with Wireless has associated each of our keys with our phones. Despite them both being paired, it will prioritize the phone last used with the key. That's pretty handy for not having to fight with pairing.
I actually prefer wired. Otherwise when I get in the car and turn it on, it pairs with my wife's phone half the time instead of mine. I also almost always want to charge the phone anyway.
I have an actual wireless carplay head unit from Pioneer in my 2011 RAV 4, but I still use it in wired mode with one of these dongles because Pioneer's implementation is so buggy as to be practically useless and they refuse to issue any firmware updates for the device.
Moral of the story: Don't buy a head unit from Pioneer. They suck ass. This is quite possibly the shittiest tech product I have ever spent money on.
I haven't used wireless yet, so maybe I just don't know what I'm missing, but I kinda just don't see why I'd ever need it. I'm going to plug in my phone anyway to charge, so what's the draw?
It’s kinda convenient for short trips (thus no plugging required) because it just works: it shows up on screen immediately when you get into the car, no extra movements. Also, I think cars with wireless CarPlay are equipped with wireless charging for a convenient combo.
But boy it’s enraging to use with multiple phones when you share a car with someone. It will stay connected to other person phone, while in range, and, if you to call that person, car will ring. It’s chaos… or maybe it’s just me.
Yeah my RAV4 has a nice wireless charging feature which I expected to make lots of use of. Just place the phone on a little nonslip pad and it charges as I drive! Perfect and so convenient.
Except I misunderstood that my year didn’t have wireless CarPlay, so I have to choose between CarPlay and wireless charging :( Hopefully my next car will have both.
I haven't seen a car USB port that does fast charging, only slow 5V @ 1 or 2 amps. So if I need a charge, I'd rather use a fast 18W "cigarette lighter port" charger and wireless connection.
Otherwise my trips are usually short enough. Also my phone's dual screen addon case (LG V60) blocks the USB port.
Being able to simply leave it in my pocket is really nice. Particularly nice when running errands around town (in and out frequently)
Prior to switching to Apple, I was running into major issues with USB-C ports being hyper sensitive to physical positioning. Charge would always work, but Android Auto would simply drop out seemingly randomly. Didn't matter on cable. It was just something that would happen.
I don’t find I need to charge - my phone lasts all day even with relatively heavy CarPlay use. I suspect it uses less power than you think it might because it doesn’t need to power the screen.
I hate it. In my car (admittedly the only one with wireless I’ve ever tried) there is an extremely noticeable delay between input and action.
Tapping on the screen for the next song or pressing the steering wheel button to do the same can take a half second or more. Doing it on the phone is nearly instantaneous, just a very tiny audio delay.
Using a cable? Never any delay at all. Unnoticeable.
Between that and the fact that it can really burn battery I’m happy to plug it in every time.
Or just get a car that you can retrofit a head unit to. Could be a third party one or a better one from another model. In my VW I retrofitted a head unit with navigation and cruise control, amongst other things. Be sensible and get things like parking sensors that are hard to retrofit, but don't pay for things you can easily fit yourself.
Android Auto: I use the Motorola wireless dongle in a VW. It's great when only 1 phone is paired, but my wife and I share the vehicle in question. It allows multiple pairings (despite the instruction manual hardly mentioning that ability, if at all) but it's finicky as hell. Half of all attempts at swapping phones end in unplugging the thing and using a cable, to the point where I'm thinking I keep wireless AA for myself (driver 90% of the time) and have her use a vent mount with only BT audio (driver 10% of the time). Connection handshake delay before visible feedback is about 30 seconds, which makes troubleshooting an extremely latency-riddled nightmare. But the other half of the time, it's as simple as selecting the desired phone in the Bluetooth menu of the car. Once connected, the experience is identical to wired AA.
This problem didn't exist at all before going wireless.
Some of the dongles function as Android tablets as well, allowing you to run pretty much any Android app (like Youtube, Netflix, etc without any restriction, though not all apps are safe and may even be illegal while driving)
I am not sure if this would be a problem with wireless as well, but I've had very inconsistent experiences having Carplay start at all using a dongle. I'd say right now it works about 70% of the time, and about 25% of the time when it doesn't, I can get it working by unplugging and re-plugging after the car is on. The rest of the time I have to stop the car and start it for it to work. This all started about a year ago though. (2022 RAV4 Prime, multiple iPhone and iOS versions, genuine Apple lightning cable, both with and without an extra Lightning->USBC dongle.)
In my experience it's over cable CarPlay is reliable if the head unit is already booted and the phone is unlocked when plugging in. If either of those things aren't true then no guarantee the phone will connect after the booting / unlocking finishes (I use and recommend the setting to disallow accessories when the phone is locked).
The wifi dongle audio quality was reminiscent of the old iPod FM Transmitters that would plug into the cigarette lighter. Not to mention the annoying delay.
Agreed, I bought a 2017 VW Golf recently, it has wired Apple CarPlay.
Bought a dongle for £55, works fine.
Love having a car with a steering wheel with real buttons, and climate controls with real knobs.
It feels like around 2018 is the zenith of Human Machine Interface in cars and it's all been downhill since, as they cram everything in a fucking touchscreen.
It seems to be on the upswing, at least for some manufacturers. I went from a 2017 CRV to a new one recently and they replaced the touch-based steering wheel volume adjuster with a physical wheel and moved the only thing that I ever wanted to use the touchscreen for while driving (HVAC mode control) to a physical button. The only controls that are stuck on the touchscreen are ones that a reasonable person would never want to try to change while driving.
In my experience wireless car play is a nothingburger -- you end up plugging in your phone to charge it anyway. If it's not plugged in, the wireless interaction draws so much power that you need to plug it in. Stable configuration is therefore: plugged in. (yes you can sometimes have the combination of wireless CP and wireless charging, but wireless charging also sucks).
I've had a Carlinkit 3.0 for two years connected to a standard $400 or so Pioneer head unit with CarPlay and it works fine.
Initial connection is a bit slower than wired, maybe 20 seconds or so, but it's up and running by the time I'm moving my car.
There's very small amount of input lag for stuff like skipping songs or pause/play. I'd say that lag is almost exactly the same as when I used to only use bog-standard Bluetooth to connect to a head unit with my phone so I think that's just the downside of a wireless connection-- wired doesn't have this lag.
I don’t know man. Never underestimate Apple’s marketing. If it’s a good implementation, all they would need is 1 automaker to do it the others will have to. I wouldn’t consider a new car without AA/CarPlay and many others are in the same boat. Automakers had to start offering upgrades to their old cars to support AA/CarPlay. There is no way automakers could invest as much into software development as Apple or Google. I hear people like their Tesla infotainment systems, but I don’t have a Tesla nor plan on buying one. Also don’t see how they could match the AppStore/Play Store.
But if Android Automotive runs CarPlay, I’m not sure how much I care. And I’m a huge fanboy.
I’d like to be able to use Siri to control the temperature and a few other things. And taking over the whole screen could be nice.
But Apple is such a control freak about things I have a hard time seeing most car makers agree to it. And just about every industry Apple has touched has learned the lesson that Apple takes things over. Car makers clearly already dislike the fact that CarPlay is so popular, I don’t think they want to give Apple MORE leverage.
I find the idea of Android taking over the industry a problem (monoculture = bad, manufacturers are clearly not up to it themselves). Apple could be a hedge. But they’re not a play nice company even when they should be.
They haven’t learned humility again after losing it during iPod/iPhone rise.
> Google jumped on this opportunity by releasing Android Automotive. Unlike Android Auto, Google's equivalent to Apple CarPlay, Android Automotive runs natively inside the car and any carmaker is free to use it.
Tangential I know, but I giggled at yet another example of Google's incompetent branding and product naming. How a company filled with so many brilliant people can be so bad at brand strategy is baffling to me.
The actual name is Google Automotive Service [0], which shortens as GAS. That's not bad I think, at least it's easy to remember.
The fact that it's still confusing with android auto being another thing is I think partly on purpose, to make it look like Android is doing more than what it's actually responsible for.
Few things are as big a win as ensuring your customers are never quite sure they really understand your product line up. Because they actually never really understand your product line up.
I can understand this kind of thinking in a here-and-gone used car lot.
> Google isn't a plucky startup with just a few dedicated engineers.
They made great long term corporate value and customer relationship decisions then, for their context.
When big corporations full of talent get sloppy, it is usually either a very competent emphasis on short term financial growth or value multiple growth. Or the result of multiple competent actors infighting or working at cross purposes, due to a lack of strong internal alignment. The latter jungle case saves a CEO from having to deep dive, they can just see what happens and select winners.
Either way, by competent I mean someone is accomplishing exactly what they intend. Optimizing their net worth given the company's incentive and power structures, in the absence of a CEO who cares more about the long than the short term.
It only takes that one difference at the very top to trickle down. The greater the talent, the greater the effect.
> There used to be a big difference in driving characteristics and technology between premium and budget brands. Compared to a Volkswagen, a BMW used to have a more powerful engine, better handling, and comfort features like seat-heating and cruise control. However, a Volkswagen Golf now has similar tech as a BMW and with the transition to EVs, drivetrains and handling won't be the same differentiator as before.
The thing is: that didn't used to be that way. You can blame it on the transition to EVs, but part of that transition seems to be that a bunch of manufacturers decided not to build their own platforms, motors, etc and are just licensing from other manufacturers*. The article's correct to note that flattens a lot of the value proposition of any given manufacturer, and if that's forcing them to lean in a lot on the software, that's a weird position for an automotive manufacturer, because that's never been anyone in the industry's strong suite - there's a reason 80% of drivers won't buy a car without CarPlay.
It's also notable that the brands who do seem to be going in on CarPlay are those that still make a point of building their own engines and platforms - Aston, Porsche, and even Polestar tries to differentiate itself there.
* to be clear, this was happening before EVs, too - BMW put out a car that shared a platform with a Toyota, in a move that should've caused a plague of locust to descend on Munich if God existed and had a driver's license, and Stellantis put a Lancia badge on a Chrysler a couple years back.
What happened is there are no real "budget" cars anymore. Mandatory safety, engine, emission, backup cameras (yes, they're mandatory) has steadily driven the base price of a car so high that the added cost of creature comforts doesn't actually increase the price of a car that much relative to the total price. Why would someone spend spend $40k on a bare bones car when they could get a well-equipped car for $45k? That's basically where we're at. So, basically all cars, "budget" or luxury, need to have all the major creature comforts.
Cars require hundreds to thousands of microchips these days. Airbags require sensors and circuits. Emission systems require expensive sensors, and engines require complex electronic and mechanical real-time tuning systems to constantly monitor and adjust fuel usage to limit emissions and meet minimum mpg requirements. Automatic transmissions are now extremely complex, including 8 to 10 gears or using CVT to achieve the best fuel efficiency. Backup cameras need a monitor, and it costs more than $200. And now that you must put a monitor in the car, you can't just have it be useless except for a backup camera, so now you have to add a bunch of software features. Meeting emissions requirements means making cars much lighter. This means more expensive composites and materials. This means increased manufacturing and engineering costs to meet structural requirements while designing cars with the least amount of material.
I don't think you realize how simple cars used to be. Engines were extremely simple and purely mechanical, essentially 50 year old technology. Transmissions were simpler, either manual transmissions or ATS with only a handful of gears. Engines and cars were made entirely of cheap steel. You had no expensive screens, no expensive sensors or electronics. You had no expensive research on structural design and materials science. You had no expensive emissions systems. You could fix nearly everything in the car with hand tools.
As far as airbags, my assumption was that you're going to have at least 2 no matter what, and I don't think more than 2 is mandatory.
The big differences to engine and transmission are interesting, I believe you that it's a big difference but I'm not sure how to evaluate specific numbers for it.
Building the entire car with more expensive structure and paneling sounds like a pretty bad thing to effectively require, especially when most losses are aerodynamic.
For backup cameras I was including the monitor in that price. You can buy kits for under $50, so I don't think $200 is unrealistic for something built in.
I will say that having a thousand microchips rather than a dozen is probably because it's cheaper that way.
Re: "global manufacturing" - yeah, but there's also a choice being made there. I own an older BMW, and there's been a very clear shift away from what was once a differentiated product to a kind of blah middle-object, which is almost certainly in pursuit of a larger market share, but again, that's a choice. I could tell you what the value prop of a BMW made between 1975 and 2005 was, I can't really tell you what it is now, and that's a choice it seems like a lot of manufacturers are making.
For a very long time, BMW's cars were intentionally driver-centric - if you got a BMW, you were getting a car that was luxurious and well built, sure, but more than that, was just sublime to drive in a way the spec sheets don't capture. The handling, steering, suspension feel, engine and transmission, everything combined for something that was engaging and genuinely fun to drive, but didn't sacrifice actual daily usability. That was their sales pitch: Want luxury? Grab a Benz. Care about the actual driving experience? Here's your car.
I really don't know what happened after 2005, and it wasn't overnight, but the general driving experience of the cars has declined every generation since then - steering's gotten a bit more numb, suspension's not quite as communicative, engine noises are more muffled, and they just seemed to lose the thread over the years. I'm not sure if some of it was leaning into technology at the expense of the driving experience, but aside from the one or two models per generation that still have some of the old soul, they really don't seem to know what their sales pitch is anymore.
> BMW put out a car that shared a platform with a Toyota, in a move that should've caused a plague of locust to descend on Munich if God existed and had a driver's license
Why is this bad on BMW? Doesn't Toyota make the most reliable cars?
It’s nothing to do with the reliability. It’d be like finding out that Apple had shipped a laptop based on Dell internals - whatever the results, it’d indicate something tectonic had happened in Cupertino. It’s just not something the company would have even considered a decade or so ago.
The BMW Z4 2018 and later models shares platform with the Toyota Supra 2019 and later. The platform, at least according to both brands marketing, was co-developed, and it is manufactured in Austria by a 3rd party that also built cars for other brands.
And no, Toyota doesn't make the most reliable cars. It might have been true for some period of time, but isn't true for every model and every year (even Corolla, often suggested as a very reliable sedan, has some versions that are in fact infamous for their tendency for oil sludge buildup that result in engine failures).
I don't understand why they never formed a committee and decide core basic features an OS should provide, set up an open spec API (a good one), and have that as a basis for android, ios, and mobile phones. Something like that could easily hold up for 20 years. Obviously we could put in climate controls, but all the cars should have the same basic "position" and "heirarchy tree" of where basic controls/settings are. Whenever I rent a car it's always a fkn hastle to find settings. That's all I ask :)
After spending a very long time trying to find the perfect non-Tesla EV with both carplay, good range, cargo space etc… I gave up and “settled” for a Model Y LR.
The software in a Tesla is just as good as Elon is bad. Not to turn this into an anti-Elon thing… But the software in the Tesla is really second-to-none, and more folks would find out if it weren’t for Elon.
I was in the same spot, decided to do a 3-year lease on a gas Mercedes while waiting for the US EV market to deliver the mythical "Tesla Y with Carplay and manual air conditioning grates". Hyundai Genesis is really close, needs a bit more baking.
Elon was, for a long time, the genius billionaire who saves us from climate change and gets us to Mars, so even in liberal cycles he was liked until he bought Twitter (or until he became a sh1tposter on Twitter).
Some people just like the software, some don't. Me personally, find it way too much, too complicated and I am not going to control my AC over touch screen, sorry.
You don't have to download a new podcast app because you could just play it through Bluetooth. CarPlay and AA probably wouldn't even exist if it weren't for car vent phone holders and 2.5mm/Bluetooth upending automakers' grip on navigation and entertainment.
It's amusing how some new cars seemingly deliberately have no place to put your phone, while some older ones were specifically designed to hold your iPod.
Two things scare me about Android / Apple car integration - I’d be interested in experiences others have had:
Obsolescence - ‘your phone is too old to use with this car’ or your car being too old for your phone. Double points for software updates to either which disable functions or features.
Vendor lock-in - are the players mutually exclusive? Does having a phone from one manufacturer limit my car choices?
I love CarPlay, but has anyone noticed that Apple just isn’t testing on non-touch interfaces? Some cars use a wheel to navigate the UI, and there are some awful CarPlay UI bugs where certain buttons become unreachable.
Thankfully it seems to be mainly just Mazda and Alfa Romeo in currently on sale cars doing the disable CarPlay touch screen while moving thing? There are probably other outliers, but those two are the primary culprits who made it a brand policy.
In some 2024 Mazda models they are finally letting you choose to keep the touch screen working while moving, letting you use either the wheel or the touch screen. I don't think either company will still be disabling touch screens/wheel only inputs for CarPlay for very much longer. Similarly, the new infotainment platform for Lexus cars has done away with weird input devices ("Remote Touch") and you can just poke the screen with a finger.
I’ve a touch screen but I never use it and almost always use the wheel (bmw), and I’ve rarely had bugs or difficulty using it. What vehicle dyou have, out of curiosity?
> A lot of the concerns around branding focus on the instrument cluster as it's one of the most recognizable parts of an interior. If you look at the CarPlay concept for Porsche and remove the steering wheel, there is no way you can tell it's a Porsche. I'm sure many brands took notice of this... With many traditional differentiators being democratized, design is a good way to stand out.
This is a really good case study in how difficult it is to find a balance between co-branding and maintaining a consistently high-quality design system across co-branding partners. There's a massive amount of work across UI/UX design and implementation done at Apple that assumes that widgets are not only using a serif font, but a specific serif font with specific kerning; that color-primary-60 and color-primary-50 and color-for-text-on-top-of-primary-60 are distinguished in a very specific way.
(Light/dark mode and localization efforts force a degree of flexibility here, but there are still a finite set of QA targets if you focus on primary language markets.)
But what happens if multiple partners want their own primary color and font? This suddenly has far-reaching, costly ramifications across multiple organizations. Even having planned your APIs from day one around color and style customizability doesn't guarantee that this can be done successfully. Taken to an extreme, frontend engineers (not just their embedded designers) are practically required to hold the context of all future potential customization needs in mind when implementing a component - a nigh impossible ask.
Which is to say that there are few companies that could pull off what the OP posits that car manufacturers are requesting, having a world-class interface that is customized to their brand. That's a tall order even for Apple's depth of talent.
If it mattered that much, I think Porsche could still just use their classic gauges with 1 being a screen (as they currently do on most of their cars).
Probably unpopular opinion, but I preferred mechanical gauges anyway.
As a former iPhone and current Android user, I would be very hesitant to buy any car that has Apple software in it. I don't trust Apple's software to work well with whatever phone I end up picking next.
My understanding is that most cars ship with both Android Auto and Apple CarPlay, so you (or your passengers) don't need to commit to one mobile OS only because of the car you bought.
Oh yes, current cars do. I'm concerned about the situation where Apple provides the onboard infotainment system, which is what's being talked about in the post.
No it’s not what’s being talked about. I’m not sure where you’re getting this. There aren’t a lot of details about CarPlay 2, but even this article says that CarPlay 2 is a way for the car and the phone to communicate more and allow the phone to control more of the car. What you’re thinking about is Android Automotive where there is an Android box in the car that is THE infotainment system. I have yet to see anywhere that Apple is building an iOS box that sits in the car itself that is the only infotainment system.
Yeah, I believe so! But the post talks about using your iPhone to do things like turn on climate control remotely. I'm incredibly dubious of the idea that Apple would build an Android app with the full feature set that they'd ship for iPhones. And I'm never going to accept a $$$$$ purchase locking me into any particular ecosystem.
We are in the golden age of phone<->car integration, so enjoy it while it lasts. In a few more years 100% of auto manufacturers are going to start charging a monthly subscription for this, split between them and Apple/Google.
People would just buy a head unit for a fixed price, and use the built in screen only for adjusting aircon and whatever is vehicle-specific. Also carmakers would never give up to have a total control over their car screen like this. They are struggling for a decade now with more and more horrid iterations, with no end in sight.
>Competition aside, from Apple's perspective, CarPlay is also not ideal in its current form. For a company that values control over the user experience, it's an eyesore to see their product embedded inside another product.
Why can't we let the car deal with car-related functions and Apple deal with playing podcasts and showing maps? It seems like the best of both worlds to me. Apple is solving a problem that doesn't need solving.
Don't get me wrong, I'm sure Apple could design a better interface than car manufacturers, but is it really worth the hassle?
This! Cars should be dumb and use physical buttons for all functionality I might use while driving. The entertainment system should just be a larger screen and better speakers for my phone. I have zero interest in dealing with a infotainment system developed by the car maker that's built-in. This way I can upgrade by buying a new phone, it's synced to my phone because it's my phone and everything works the same in rental cars and I don't need to sync any of my data to rental cars. I'm pretty happy with my car, but I'm sometimes annoyed when it thinks it's ok to use the display or speakers for things that aren't coming from my phone. It's the same issue as stupid "smart" tvs!
I went from CarPlay (Kia e-Niro) to Tesla (Model 3) to CarPlay (Volvo XC40 Recharge) and I vastly prefer CarPlay. I've tried wireless CarPlay (BMW M2) and while convenient when entering and exiting the vehicle I found it to be quite laggy compared to wired. While Tesla does have the Apple Podcasts app it's a really bad experience. It won't sync playback state properly so I have to memorize what I'm listening to and how far I've gotten when I switch from car to phone or back. Similar story with Apple Music and even Spotify (which is usually great at keeping state between devices but not with Tesla for some reason). Yes, I paid for "premium connectivity". Tesla did have better route planning though.
I also really don't like wireless charging. It spares the port from constant wear but my device tends to get really hot and the charging is very inefficient. I plan to go back to slow wired charging everywhere to put less strain on the battery. I already have in my car and by my PC. Unfortunately my XC40 doesn't support dual screen CarPlay so I use the Google Maps in Android Automotive for navigation and use CarPlay for music, podcasts, etc. Switching from AA Google Maps and CarPlay and back is not a high friction interaction though so it's an okay solution. I'm also able to deactivate wireless charging so I'm pretty happy with this solution.
If you don't mean "make more money"/"squeeze out every penny possible", I doubt the reasons are the same. These are 3 totally different classes of products, with different development/support costs, and pricing orders of magnitude apart.
One thing that is ridiculous with CarPlay is that I can't read messages: it only allows to "speak" it and while I use my phone in English I message with my friends in my native language and speaking works terrible. Even my car's shitty infotainment system can display messages.
This results in the worst possible outcome: I need to take my phone and read the messages on my phone screen, creating a safety risk.
That's the problem. I don't know if the message is important or not unless I see it. There wouldn't be a safety risk if at least the beginning of the message was displayed on screen. That would be enough to figure out if it's something worth stopping for or something that can wait.
Well it could be important. CarPlay has no problem displaying a notification of a message with a red dot or a popup from below. If it didn't, then I get it.
But it does show that there's a message but doesn't show the message itself: worst of both worlds.
A huge benefit to CarPlay for me is I can travel someplace like Japan, get a car with a CarPlay-supported head unit, and immediately I have Google or Apple Maps available straight off the bat in English with all my saved locations like hotels or places that are bookmarked. Nothing to set up or anything.
The only downside I've experienced with it in Japan is the GPS can be wonky in tunnels whereas the car's built-in GPS seemingly doesn't.
The car probably has odometry and inertial sensors to do position updates while GPS is unavailable. That's actually how some of the original car nav systems worked even pre-GPS in the 80s.
Cars simply have bigger, more power antennas because they don't have to be packaged in a phone.
One of our cars has a wireless modem that can be used for a hotspot. I've driven to plenty of areas where my phone has no service (think remote and hilly terrain), but the car still has two or three bars.
When I worked for an Auto maker (left 2 years ago), the cluster was an independent system from infotainment. They could follow messages from each other on the bus (it wasn’t CAN bus, we’d moved on from that) but basically there wasn’t a way for CarPlay to talk to the Cluster.
I’m willing to bet that the soonest any large OEM could offer CarPlay 2 would be next year. The development process is just too slow to turn around something.
I still drive a 2008 model car with nothing more than a radio and a basic Bluetooth handsfree system, and it's all I've ever needed. When I bought the car, I consciously opted out of a 3500€ built-in GPS, predicting that smartphone navigation would soon surpass it. Over a decade later, that decision still holds up—my iPhone handles all my navigation needs more effectively than any built-in system could.
Now, as I face the inevitability of replacing my car due to dubious environmental regulations (current car emits less than modern but twice heavier ones), I'm genuinely concerned about the current state of car infotainment systems. They feel overly complex, bloated, and fragile, prioritizing features over functionality and user experience. The shift towards these systems seems more like a downgrade, sacrificing reliability and usability for unnecessary add-ons that don't improve the driving experience.
Anyone in the same position? What car would you recommend that meets environmental standards but keeps the infotainment system super simple, or even non-existent?
The AA/Carplay experience is incredibly superior to a cell phone clipped into some type of holder. With wireless Carplay, I couldn't even imagine going back to plugging my phone in, or even taking it out of my pocket. There is no downside because you aren't forced into using it.
I guess YMMV. I've recently had the chance to test drive some of the latest models from BMW and Audi, and honestly, I found their infotainment software to be quite laggy and sometimes unresponsive. It felt inexact and bloated, reminiscent of an old Windows XP laptop with a Samsung UI skin. While the CarPlay integration was ok, it still didn’t surpass the simplicity and efficiency of just using my iPhone mounted on the dashboard as a makeshift heads-up display.
I'd agree with the comments about the built-in system for any car but I disagree that mounting your phone on the dashboard is more simple and efficient then just leaving it in your pocket for the same functionality.
All I want from CarPlay is for it not to start blaring the last thing I was playing on the car speakers as soon as I plug it in, especially given there’s about a 10 second window between the sound playing and the head unit being booted enough to actually show the screen so I can pause it.
I have no idea if it’s the car or CarPlay but could I please just have a toggle somewhere?
It's the car I'm pretty sure. If anyone has another take I'd love to hear what they think, but I've used many implementations of CarPlay and have never encountered what you're describing.
A similar thing happens to me. I don't use Apple Music regularly. This means the last song I played is from something like 2-3 years ago. I constantly listen to podcasts and other music streaming services on my phone. Constantly. In the shower, running, in the gym, walking around town...constantly.
Yet when I enter my car and CarPlay starts for some reason maybe 50% of the time (not all of the time even) it randomly decides to play the last song I streamed in Apple Music 2-3 years ago. WTF Apple?
> This is a little unfair because Castro is a small, iOS-only app. [...] But it does mean that if I want to listen to podcasts in these cars, I have to change my podcast app to one available in the Play Store.
That's the downside of choosing apps that are not cross platform. Don't do that if you want to have the freedom to use different systems.
I really hope the lock in with have with car/phone a connection is something the EU addresses next.
I don't like that car manufacturer has to use Apple / Google software, why can they not make an app that you install and your phone connects to the car in a useful way.
The reason is that only Apple apps have the required permissions and system access to do it.
> I don't like that car manufacturer has to use Apple / Google software, why can they not make an app that you install and your phone connects to the car in a useful way.
I've had some rental cars that wanted me to install a manufacturer app. Super annoying! Just let me plug my phone in and project my screen and speakers. I don't want to get adjusted to a new UI with each car I drive, or worse sync stuff. I also have zero interest to find out what software a car company will develop. Car companies should make cars and leave infotainment to my phone and companies that are experts in it. Built-in infotainment in cars is even worse than "smart" tvs.
I do agree, I think mostly car manufacturers are gonna make something worse, but it would be nice to have some more competition on the market as carplay and android something are also not good.
100% agreed on more competition. In addition to not having much face in cart manufacturers ability, I also worry about how a successful interface by a car manufacturer would impact availability of carplay and android auto. It's similar to net neutrality and ISPs offering streaming services themselves.
What permissions do iOS or Android withhold that prevent you from doing this? The reason car manufacturers don't do this is that it makes no sense to put the software in the phone if you're not trying to make the software portable between different cars.
I'd be perfectly happy with carplay for media/navigation only. I want my analog dials, knobs and switches for all other car control and climate control functions.
The screenshots there make me feel pretty good about keeping my car from the 90's if that's the current state of new cars.
Carplay2 seems cool, but if it takes away my ability to just jump into whatever new car with my iphone or android phone and use aa/carplay I don't want to lose carplay. I have 3 cars with carplay and I don't want to have to set all of them up and lose portability.
How significant is the demand for advanced infotainment systems like CarPlay 2 among consumers? Are there market studies or consumer surveys indicating that buyers would prioritize "CarPlay 2" capabilities when purchasing new vehicles?
Tbh CarPlay runs really badly in my 2018 VW GTI and Castro in particular is extremely buggy. It’s pretty much random if the turn by turn audio navigation works at all.
The fact that everyone seems to think CarPlay is great though I guess suggests that the problem lies with the implementation details with the VW.
> there is no match for seamlessly transferring the phone's state to the car.
I’ll point out it’s not at all seamless. In fact it’s extremely cumbersome to fish out a lightning cable and plug the phone in. A good car infotainment system where I didn’t have to connect my phone at all would be seamless.
I use CarPlay but it is very clunky - I need to plug in the device etc. I've been working on using and old iPhone and keep it connected in the car all the time. I plan on using wifi hotspot for connectivity and since the device can sign-in using my ApplID, it can have access to my calendar or Messages as needed. Sharing the routing address from another Apple device can be done easily from Maps.
This seems like a missed opportunity for Apple - they can release a screen less device bundled with a cheap 4g connection and smooth out other integration issues for sharing / setup etc.
That’s specific to your car’s head unit. When I get into my vehicle, the iPhone in my pocket connects to its head unit via Bluetooth with zero actions from me. Seriously, I get into the car and the CarPlay display comes on showing my phone’s apps. I don’t have to plug it in, tap a button, or do anything else at all.
Almost bought a ford mustang but for some reason that year ford did not have CarPlay that worked via Bluetooth , you had to have the phone hooked up via usb and maps was glitchy as hell during the test drive. Bought a challenger instead which is a bigger nicer car anyway and CarPlay is flawless. I have a 2020 Yukon XL as well which is amazingly flawless for CarPlay , too bad GM dropped CarPlay. I wouldn’t buy another GM without it.
Shouldn’t this be exactly like the bluetooth integration? It’s just a defined protocol for basic functions like start, stop, play etc. so similarly stream any Apps from your phone with basic commands? I am happy to limit myself to a large google maps screen and the rest is control of my entertainment like I do today (open the app and then control via bluetooth).
By this we have no vendor lock in and just making sure there is a protocol downwards compatibility…
I recently rented a car with Android Automotive. I was not impressed. It was laggy, clunky, and confusing. Fortunately it supported CarPlay, so I mostly used that.
The one cool feature was that Google maps driving directions showed up on the front instrument cluster in addition to the center console. That’s a feature I’d like to have more broadly.
The author is talking about that they only use Google Maps, Spotify and some podcast app and that those are not available in the car. Interestingly for me only the Google Maps thing is a problem, actually Google Maps doesn't work here in South Korea so I have to use Kakao Maps or T-Map. Anyway music and especially podcast work wonderfully over bluetooth.
Recently had two rental cars with CarPlay. Oddly, the experience with a Chevy (who is abandoning CarPlay support) was much better than the Hyundai, though its main problem was connecting and constantly asking permission.
Since my older cars do not have it, having a great mount (Peak Design) makes a huge difference, but damn I want it integrated if only for the bigger screen.
> Thanks to a deep integration with the software stack of the vehicle, CarPlay 2 can control most infotainment functions. It can therefore take over the entire infotainment display, the instrument cluster, and any passenger displays.
The instrument cluster?
Any other safety-critical functions that the iPhone isn't isolated from?
Normally the instrument cluster runs a different software stack, but can overlay some graphics from the infotainment in certain areas to show navigation or what music is playing right now.
All I look for is a USB connection to the simple stereo, plug my phone in and mount it on the dash. I fail to see why I have to deal with the additional cognitive load of an other interface that WILL go wrong when I already have a perfectly good one in my pocket.
I also prefer this, but I find that lots of cars for some reason end up with weird volume levels for turn-by-turn compared to the music. Have you found a bullet proof way of adjusting this?
Interesting, no, I don't have this, luckily. I do have a weird thing that Siri speaks through the phone's speaker not the car's speakers, even when I'm playing music through the car's speakers over USB. This would probably be fixed if I used Bluetooth for the audio, but that's a whole level of pain I'd rather avoid.
I disagree with this love of hooking my $1000 phone to my $30000 car.
Everywhere else in my house, I try to make gadgets independent of one another. I remember when a printer worked only when it was hooked to a PC to be its server. That was bad. It’s much better for my printer to hook to the network directly.
Some thermostats require a phone to program them. That’s bad. The thermostat should have full functionality as a stand-alone device. My dishwasher requires a smartphone app to work some of its features. Bad. My old dishwasher was fully controllable from the panel on front.
A car is no different. I don’t want to have to plug my phone into my car like a dongle, even wirelessly. The car should connect to the network itself and have needed applications like nav and music. Maybe car-makers suck at making this software, so maybe Apple or Google should make it for them. Or maybe carmakers should get better at it.
But no, I’m never going to think that plugging my phone into my car is any better of a kludge than is forcing me to get out my phone to work my dishwasher.
I so much prefer hooking up my phone, rather than having to authenticate every device I own to all the services I want to use. And given that most of the services that I want to use in the car are also services on my phone, e.g., GPS maps, music, podcasts -- why would I want to have to use a whole new app for this purpose?
I love that when I search for something with Google Maps it knows, typically in a letter or two, what place I'm trying to go to. I don't want to use a different music service in my car -- it has all my playlists and my "likes" already.
Do I want Ford to write its own implementation of Google Maps or Pandora? Why? Seems stupid to have 10 different implementations of the same service.
You make a great case for why your car should have Google Maps built in. That wouldn’t require you to hook your phone to the car.
Having to hook the phone to the car is s giant leap backward. We’ve got Gmail: email from any device. We don’t say: My phone has all my email, I hook it up anytime I want to read mail, if the battery is dead or I forgot my phone, I’m screwed. No, we use Gmail from a tablet, phone, PC, etc.
But for some reason people believe it’s an awesome advance that we can hook the phone to a car, as if it’s impossible to build Google Maps into a $30000 car.
That’s like arguing for a smart TV instead of a dumb TV plus a smart set top box. I’d vastly rather my car’s system be a dumb display for my upgradable phone. If Google stopped making Maps, your idea would mean my car could no longer navigate. With CarPlay or Android Auto, you can just switch to a different map app.
The lifecycle of a car and those gadgets are quite different. Without upgradable hardware in your car you'd be stuck with a very old version of Gmail on your car.
This is the same logic as why some people would rather buy the dumbest TV they can and let the smarts live in an external replaceable box.
In addition to the points other comments are making, please consider rental cars. So much easier there if I can just plug in my phone and get the same Infotainment UI everywhere.
If I were to place a bet, I would bet on car manufacturers abandoning updates to the vehicle software far earlier than I desire, rather than maintaining it appropriately. I’d also guess ads or other unwanted intrusions will be coupled with it.
I’d rather connect my phone, for which I’ll update the hardware 4+ times over the time I own my car, and get the corresponding OS and software updates yearly or more frequently. Send that data to the car’s reliable and dumb screen(s) from my chosen phone and I’m in control and happier for it.
This article is very surprising to me, people actually use and want this software? I always avoid using car infotainment because they suck and just link my phone via bluetooth because it isn't the worst possible thing.
As CarPlay aims to dominate even more screens in our cars, isn't Apple simply replacing buttons and dials with potentially distracting and fingerprint-smeared touch screens just to extend its software reach?
I and most people I know are part of that 70% of people
who would never buy a car without CarPlay. Even my elderly relatives who are not GREAT with tech in general seem to love it.
As a shopper of a vehicle. I want my car to control car operation/cockpit information. I have enjoyed the continuous improvement of navigation and entertainment functionality of Android Auto/Car Play.
For the love of god, keep them separate. One is critical vehicle functionality. There other can crash/reboot/have connectivity issues, without me being concerned about knowing the engine is overheating/battery pack is dead, a tire is blowing out via TPS or I'm speeding.
I don't want a car that is CarPlay only, guess what, my car is not an accessory to my phone. The genius of current Android Auto/CarPlay is that the car head unit can act as a mostly "dumb" head unit for my external mobile processor.
It is bad enough that Mercedes decided to emulate Tesla. Now, fuck this, I don't want my merc dashboard to look like an IPhone. It cheapens the experience.
I am not even talking about the software, because I think it is a matter of habit and expectations, but this horrible trend of refusing to have a proper dashboard and instead just slapping what looks like a bunch of tablets in front of you, and of course, the death of physical buttons. I absolutely abhor this.
In my view, Apple does a better job designing high-end digital surfaces than Mercedes. Last week Mercedes decided to add a green polo shirt to all the user avatars in my vehicle, which to me seems like a clown move, and something that Apple would never do. You can always just not connect your phone to your car if you prefer the stuff the car does on its own.
A lot of this is a parable about what is wrong with the tech industry, which rewards dominance and market seizure rather than utility, longevity, and other consumer benefits.
Car makers what "differentiation". Uh huh. That is literally, exactly, what consumers do NOT want or benefit from.
I want consistency so that I can get in a rental car, or my friend's car, or my new car,
and have sh-t just work.
The LAST thing I want is a curated subset of familiar apps and controls, plus crappily designed but "cool looking" replacements all about convenying Brand(tm) Values.
> I use only three apps when driving: Google Maps for navigation, Spotify for music, and Castro for podcasts. The number of cars on sale today with these three apps available is zero.
My car (a Volvo) has two of these three apps built-in.
This article could really use a tl;dr. I tapped out about a third of the way through after I got bored with slogging through so much windup in search of a “risky bet.”
It’s amusing when the author says “when the phone crashes.”
My 2022 Subaru Outback head unit frequently crashes or is really unresponsive. My iPhone never crashes.
Car infotainment systems are garbage. I’d never buy a car without CarPlay. It’s maybe the #1 measure I have for vehicle quality when I rent on vacation.
I don't know about that - my Subaru head unit crashes a lot more when I'm clicking around in the Subaru programs, and it very rarely crashes when I'm clicking in CarPlay. My feeling is there are bugs in the Subaru provided programs, and very few bugs in the program that launches CarPlay and gives it control of the screen (probably because that is a very simple program).
So it seems that the more features I use via carplay the less it crashes. It's not perfect, but it certainly helps.
I had/have a 2020 Subaru Forester and 2024 Subaru Ascent and both head units have completely frozen to the point removing the battery was the only fix (even the radio fuses didn’t reset the underlying computer).
It’s really quite a shame as they’re fantastic cars otherwise. But yes, it’s never CarPlay that does it.
I had the same problem with my Volvo XC-90 (2020 Model). It's been years since any of my iPhones have crashed. Never had a problem with carplay in my BMW or Chrysler minivan either. I hope Volvo has found the problem and released a fix by now.
In my case, I own a VW SUV and to my surprise when I go nearby a particular antenna, if I have my Android phone connected it crashes the whole infotainment system, and I can't do anything with it unless I turn the car off for several minutes, which is obviously impossible if I'm going somewhere.
However my wife has an iPhone and although going next to the same antenna does crash the system as well, it does so very briefly and just resets the connection or something, and as soon as we go out of range it starts working again.
So it does look like there's some difference between CarPlay and Android Auto even when using the same exact car with both.
No idea why this happens and I can only assume the antenna is transmitting in a frequency that interferes with the wifi connection (it does not require a cable like older VW cars) which triggers a bug in the Android's stack, but Apple's stack seems more resilient somehow.
I would love to know if anyone else has had the same experience because I find it quite baffling and very annoying when I forget to disconnect my phone when going through that antenna.
that’s less true in reality than you think, because I am in a similar situation as OP.
Infotainment systems are prone to runtime errors and crashes, if you interact less with the internal workings of the infotainment system itself, it does a somewhat good job of just passing audio through and you have fewer sessions with a crash.
That said; i have noticed a disparity in what the infotainment gps navigation can do and what carplay can do- carplay is often limited to the center console, but the built-in navigation can go into the HUD of the car. An annoying difference.
Once my Ford's infotainment system (Microsoft CE, I believe) crashed hard, and I had to open and close the door while I was at a stoplight to get it to reset.
On the other hand, the voice command system works as badly as it did when I got the car, while Siri has gotten worse.
Agreed! Seems like these systems are about 10 years behind in general, so I was pleasantly surprised that the voice system worked after a fashion. Just requires too specific phrasing to make it do stuff, and then I didn't pay to keep the maps up to date...
Also appreciate that the main UI follows Fitt's Law and I can jam my finger to the corner to go to audio (though if it would just stay on that screen between drives I wouldn't have to do anything).
Second this, had VW, GM and Ford cars, complete unusable shit without carplay, doubt the other car makers are any better.
The insane lag to enter anything on google maps on my 2024 Yukon is ridiculous. Worse, the fact it makes me TYPE in the touchscreen instead of letting me do this on the phone is yet another reason to completely ditch it. Google could easily let me control it through the phone (as they do with youtube on my tv) but I doubt they care enough to do that.
It should be straightforward, pick a direction on my phone, beam to car, car opens it automatically in google maps. That is not difficult.
Drove a rental GLA last September, and the system was quite nice to use. No noticeable lag.
Also, the integration with the iphone was top-notch. It understood that the iphone provided navigation, so if I pressed the nav button while on the radio screen, it would activate google maps from the iphone. The only thing which wasn't great was typing in some long address, which is why I connected the iphone.
Rented some kind of Audi a few weeks later, and it had the usual crappy integration. "nav" is the car's laggy navigation, etc.
Depends on lots of variables, overall not something I want to rely on just to play music. Like I got a Chevy rental and CarPlay took forever to pair then proceeded to play distorted music for some reason, while the FM radio worked fine. Worked better on another car but still took forever to pair.
I installed a new stereo for my old car today, intentionally avoided CarPlay and got a simple non-touchscreen JVC. $60, 15 minute DIY job, BT paired instantly, has aux too, and sounds great.
CarPlay is alright on rentals, but I don’t care for it as a daily driver. Gimme a USB port I can plug a 32GB flash drive full of MP3s into. I don’t mind using a phone holder in the vent when I need to map, but I only map on <10% of trips. This is probably the most luddite thing about me: I loathe using a phone while driving. I just find it too distracting.
It shouldn't matter, it's just music. And one car that failed the test was a $50K Suburban. If it's this flaky, maybe some other car works fine but one day it stops.
I think we should collectively start saying things like "wow, look at how horrible this system is. I wonder how bad the computer control for steering or brakes is!"
The point being, further hurt the sales of any cars like this, because really, it could be true.
Eh, that's something a lot of people would notice and the government would get involved in, and it's been played out already. I generally trust GM's steering and brakes to work. Infotainment is not their expertise, and if they were releasing a self-driving car, I'd also be wary given their track record.
My iPhone crashes... intentionally. Because the Spotify app caches everything rather than doing refreshes when loading "new episodes" of podcasts. I had to abandon them for podcasts because of it.
That said, a car without CarPlay is simply out of the pool of possible cars I would buy. It's a requirement.
My infotainment unit is pretty unreliable. Sometimes I have to play music or use my map from the phone speakers because the cars unit just goes black and won’t reboot
Somewhat agreed (I prefer the Tesla infotainment system over CarPlay, but they both have their own advantages, despite being imo equally good overall), but I would say it’s not super relevant to the argument. And only because Tesla infotainment is the only car manufacturer’s first-party infotainment system I’ve seen that can hold a candle to CarPlay. Not a single other one I’ve seen was even close. CarPlay was the best I could hope for, because anything that was first-party (excluding Tesla) was absolutely awful.
I don’t remember which specific brands had CarPlay and which ran their own infotainment systems, but I remember a couple years ago visiting dealerships of BMW/Mercedes/Honda/Toyota/Kia/Subaru/Ford and walking away either relieved that the car had CarPlay or internally cursing at their first-party infotainment system and promising myself to not even bother trying next time.
Doesn't matter because CarPlay and AA allow you to bring whatever apps you want to with you. If you listen to podcasts or books on your phone, you can continue them from the CarPlay/AA interface next time you drive.
CarPlay and Android Auto give you the choice of your own app and full integration with the rest of your listening life.
That's a statement about audio playing only though. To first approximation the core job of a car infotainment product is map and navigation display. Everything else is noise.
I think CarPlay is kind of stupid because the ONLY apps that make any sense are audio and maps. Both of which don’t need much variety… Google Maps, plus a handful of audio apps, and the infotainment system is complete.
Hmm waze is hugely popular where I come from, because of all the police speed traps it crowdsources. And I’ve not seen that as part of an official infotainment system like ever.
And to be fair I prefer waze, not just because it makes driving simpler, but the community feeling you get by marking stopped cars or other obstructions on the road for others to pay attention to.
Unless automakers organize and come together to build a common os/ui layer, I don’t see apps like that ever being present in official infotainment systems.
And audio book apps change all the time too - local variants with local libraries, language learning apps etc. There is quite the variety of stuff you could do in a car.
I imagine a lot of companies would like to integrate their apps to cars. Think the police, ambulances, taxis/Uber, etc. Currently they just solve the problem by mounting more devices to the car.
And even if you want to use only audio and maps, then there are still a selection of audio streaming providers and map providers around.
Tesla's Spotify app had a long standing bug where it couldn't recognize playlists with more than 100 songs. To be dependent on a car company to give bugs for every individual apps rather than to just mirror from your phone, which will have patched apps right away..
You don't need premium data for online maps. Premium data is for sat maps, music and live camera views. Also at $10 per month it's actually quite cheap.
Which is $10 per customer per month more than they’d get if everyone was using CarPlay / Android Auto.
It’s been proven time and time again that vertical integration is a massive money earner and you can bet that if Tesla were to release a smartphone then support for that in their cars would be a software update that follows immediately after. But since there’s no plans for a phone, the next best thing is a subscription-based infotainment system.
I assume every touchscreen monitor has its own processor, but that does not mean it is seeing any data other than the video stream. And I imagine that video stream data is not going elsewhere.
Source? I cannot find a single reference to Apple collecting fees, much less onerous fees, for implementing CarPlay functionality.
Automakers don’t like CarPlay because it makes cars fungible. I like CarPlay because it makes cars fungible. Apple seems like it would give away CarPlay (at least for now) to broaden the reach of its platform.
It has a far more limited selection of apps and Some I use daily are not available so it really doesn’t matter if it works better because for me it doesn’t work at all.
I used to be fascinated with Apple but I no longer want anything to do with them.
I could care less if my car has Apple or android auto at this point. I would rip it out immediately if it had one. Any always on connectivity would be removed.
Car manufacturers are increasingly selling off your private data and leveraging all of these technology upgrades you paid for to do it. As soon as the car is connected to the internet, it’s shipping off your private data and selling it to data brokers. Manufacturers are hiding behind their wall of text called “terms of service” to do so [1]
In some cases the manufacturers are reporting your driving history to insurance companies so they can get any reason to bump your rates or deny you coverage . [2]
My dream car is now a “dumb” car.
Give me a car with a simple backup camera, manual transmission, and regular sized vehicle (no trucks or suvs, fuck that).
CarPlay and Android Auto run on your phone and outright prevent the car manufacturer access to data (they only expose the video and audio streams). Whatever data selling is going on it utterly disconnected from those technologie.
So what exactly are you ranting about here? Why this rant when you don't seem to know what those technologies are?
--- very clear opt-in method for having your data sold
--- rejecting cannot prevent regular use of services
--- heavy penalties for breaking these rules
Problem is that no politician wants to touch this because
--- manufacturers sell data to subsidize the product
--- if they can't sell data, costs are going to shoot up
--- if they do this in response to a law, they get to raise
costs even more because it affects the whole industry at the same time and there's a clear scapegoat
Consumers care a LOT more about their cheap, connected devices than their privacy. Because getting by your data like [2] happens to individuals, but costs affect the group.
EDIT: To clarify, the MP only suggested that costs would go up and people don't care. The rest is my personal speculation.
It's a shame, because a lot of this connected stuff does provide useful functionality to a lot of people.
I'd much rather see stronger privacy legislation that bars companies from selling your data to third parties -- period, no exceptions, no opt-in terms that offer customers more features or services if they'll agree to privacy violations.
And then actual enforcement, with fines high enough that they can't just be shrugged off as the cost of doing business.
While I don't think we'll ever get quite there, I do think we're going to be moving in that direction. I do have a car with connected features, and I live in California, and I've already instructed them that they may not sell my data to third parties. I don't know if they actually follow the law on that, but it's a start.
Meanwhile I rented a car last week and experienced Android Auto for the first time (vs just having my phone precariously balanced somewhere in the car, and using it for navigation), and now I never want to go back or buy a car without Android Auto.
This isn't meant to be a dig of Apple vs Android - I think I would feel the same about CarPlay if i was on iOS.
I think we just value different things. As someone who wants to be able to drive, navigate with google maps, listen to a podcast, and handle incoming messages, but do it safely, these integrations are incredible.
I don't disagree with your points about connected cars, but at least for CarPlay (I don't know about Android Auto), it's entirely an orthogonal concern. CarPlay is a standard for, basically, connecting an external display to your iPhone - it doesn't need to connect to anything other than the iPhone itself, and doesn't grant the car any access to the internet or anything, and IIRC the phone can't read anything meaningful from the car either.
I really wanted to like that Mozilla report, but their methods only looked at the privacy policies of cars, not actual software/hardware capabilities in the wild. Let's say I own a Honda C-HR, but don't use the Honda Link app or connect the onboard wifi to a phone hotspot. Does Honda still get remote telematic information from me? Is there a live always-on 4G/5G connection in the car itself? It's unclear.
> In some cases the manufacturers are reporting your driving history to insurance companies so they can get any reason to bump your rates or deny you coverage . [2]
Obviously good. Driving is incredibly dangerous to other people and insurance companies are the most motivated to make it safer.
Not sure how one could reclaim driving privacy in an age where we have license plate scanners and E-ZPass scanners at places not on toll roads all while you and everyone else drives around with multiple Bluetooth devices turned on. Nothing stopping your car from recording everything anyway and having the dealer just download it each time you have an oil change. This mirrors privacy everywhere else. Think this ship already sailed.
"Give me a car with a simple backup camera, manual transmission, and regular sized vehicle (no trucks or suvs, fuck that)."
Mazda, cx5 <= 2016. Small. Has usb sd, simple knobs, backup camera. It reminds me of when apple removed the escape key, I had to learn how to buy a used mac.
> used to be fascinated with Apple but I no longer want anything to do with them.
> I could care less if my car has Apple or android auto at this point. I would rip it out immediately if it had one. Any always on connectivity would be removed.
Every partnership Apple enters must result in Apple winning and the other party simply being a subservient cog
So, no one wants to do business with them in new markets where Apple doesn't have leverage
This is why they couldn't get a manufacturing partner for the now-dead Apple car...no one wanted to be the Foxconn of cars (they do all the work, Apple gets all the credit)
How would Apple not have leverage in the car market? There are 18 companies and only 1 Apple Logo.
I think Apple failed on cars because they are significantly more complex then hobbling off-the-shelf electrical parts together and slapping an OS on it. Think of how many companies can assemble electronics and make OS. Its easy, literal sub 18 year olds make prototypes like this.
CarPlay doesn't prevent car makers from tracking vehicle activity.
In the end, consumers don't give a shit about the in-vehicle infotainment. It sucks, or it's AA/CarPlay. The first generation of iDrive showed that rich people people will buy cars in spite of the in-car stuff. In fact, most car infotainment sucks, yet people still buy cars.
Let's turn the question around: why would car makers want to spend millions of dollars a year rolling their own infotainment system? So they can make incremental revenue selling ads and user data? So they have control? Control over what, exactly?
There's an interesting (and apparently often misunderstood) article called "IT Doesn't Matter" [0]. In it, Carr is largely arguing that IT, as a business differentiator, was over for many of the things people thought were differentiators. That is, things that helped a company (say American Airlines) get a lead on their competitors in the 1960s had become commoditized. Now every airline was offering flight search and booking online (directly and through aggregators). The IT edge had become table stakes, you didn't do it to beat out a competitor but just to stay in the game. And, even more importantly, many of the things that used to be IT differentiators became commoditized.
Car infotainment was once a differentiator for car manufacturers or for classes of vehicles within the same manufacturer. Today, it's table stakes. Not all the manufacturers have figured that out (have any?).
[0] https://hbr.org/2003/05/it-doesnt-matter and https://www.nicholascarr.com/?page_id=99
Imagine if car manufacturers started offering cars with no infotainment OS, and as users having a choice of open source OS distros to install on them..
If that system is required in order to control a car's functionality (windshield wipers, etc.), then its existence is not just a disincentive, but a dealbreaker.
Japanese road network is a disorganized weighted node graph and absolutely not a grid, and a bulletproof navigation unit has been a must for a car in Japan since its inception around 1990. It is also preferred that they are 2DIN compatible so it can be later upgraded. AFAIK, those are not high priority checkbox items elsewhere, but all cars nevertheless follow the Japanese manufacturer layout because of manufacturers' collective dominance. Cars before 2DIN navigation units seem to have had 1DIN AM/FM radio units with radio buttons[1], by the way.
That dominance leaves a 4:3 8" diagonal hole in immediate view of driver for all cars globally that must be filled with something of value. That doesn't have to be a touchscreen but usually are, and it ends up being a navigation-audio combo unit, and it's outsourced to the lowest bidder. It is not the primary interaction point for cars by overwhelming global demands or principles of automotive product design. That leads to jarring subpar experience that appear to be but are perhaps not intended to be part of core UX of the whole car. I think.
1: https://www.alamy.com/1956-mercedes-benz-190-sl-steering-whe...
What Japan lacks are addresses that can be found easily without using a map. Apart from Kyoto, roads in Japanese cities don't have names (or number), so addresses within cities are not "{number} {name of street}". Cities are cut in areas smaller and smaller all the way down to a block. The last number will be the house on that block. So addresses within cities are "{name of area} {sub-area number} {block number} {house number}", with some variations from city to city.
An address might be "Nantokacho 11-16-8", which means the 8th house around the 16th block of the 11th sub-area of the Nantoka area. Good luck figuring that out without a map!
Reminds me the Township & Range system in the US, which is still used in public land management: https://web.gccaz.edu/~lynrw95071/Township%20Range%20Explana...
With back up cameras being mandated in the US (and other jurisdictions?):
* https://www.cnbc.com/2018/05/02/backup-cameras-now-required-...
There needs to be a screen: I'm not sure what the incremental cost is of it being touch-sensitive. And I think a lot of designers (or cost cutters) may have figured that since they have to have software anyway, it may easier/cheaper to deal with software buttons than with moving parts like physical buttons and dials.
Some stuff can be managed from steering wheel, but instead of putting there things like AC there are voice controls, while we all know very well this increases danger on the roads and calls shouldn't be made while driving.
That being said, I choose (used) cars based on many metrics, and this is just one of them and not the most important one. So its sort of self-inflicted move to worse.
None of which some up on a spreadsheet in the manufacturer's accounting department.
Similarly: open concept offices with hot desking lowers the amount of rent (per employee) that shows up on an expense report, but the productivity drop of open offices does not.
IF that's the case, and that's one big if, debating touchscreen vs buttons might not be meaningful, because touchscreen MIGHT be a requirement that cannot be removed. With un-eliminable touchscreen, buttons becomes just added cost, rather than being a resilt of cost calculations between all-touch vs all-buttons.
Touchscreen will probably run atop whatever OS powers infotainment system. In theory touch events CAN be handled on the same 8-bit uC, but converting click coordinates to button-press messages requires knowing what is visible on the screen, which either hinders expressiveness of touch interface (fixed islands for controls) or complicates implementation (constant updating of element-coordinate maps). The moment safety-critical control runs along with non-safety software huge validation challenges arise - the performance characteristics of the whole system, including UI, must become at the very least bounded.
Hard button solution can be internally commoditized, whereas soft button solution requires constant babysitting throughout development lifecycle. One is much more expensive than the other, however, once the move is made incremental cost drops dramatically.
(CarPlay can provide car GPS to the phone though.)
Like waaaaaay better and more accurate
As someone who used to help out with Waze maps, raw GPS tracks from phones can be WAAAY off from actual roads and other GPS tracks.
Your phone looks for the most likely road for you to be on and snaps you to that road. Your car does a similar thing but has much higher GPS resolution than your phone so it is more likely to be right.
It’s a play on the quote “Life is a series of choices and reactions” from “Before The Fall”
google maps does weird things but gaia just uses the raw GPS data from the phone and doesn't snap, however when you are plugged into carplay the phone seems to always consider the car/trucks data more accurate even when its clearly not.
this could very well my specific truck (dealer has refused to look and i don't have another to compare to) or a general tacoma thing but it both smooths the GPS data to make "nicer" looking lines and often introduces drift over time as likely some processing it does compounds an error
(i also realize i said the wrong phone, iphone 14 not 12 but even with the 12 it was notably better then the truck but the 14 pro has a much improved GPS and it shows)
> If your vehicle has wireless CarPlay, the iPhone's GPS chip is completely ignored.
> If your vehicle has wired CarPlay, the iPhone's GPS chip may be ignored, depending upon what the head unit tells the phone to do.
this happens when near roads or not. the truck is doing something weird and wrong with the GPS data and forcing the phone to use it
Cars can theoretically have better GPS antennae, and can do better dead reckoning using wheel position and wheel rotation speed. But do they? I haven’t seen any evidence that any car actually does better than a modern phone.
Especially inside a dense city, WiFi location can be much more precise than GPS.
Japan has local satellites to help with this (QZSS) and local cars may handle that better than global phones.
Also with an aging population a phone screen is just too small for many.
BTW, to parent's point VW has been trying different approaches with a top mounted GPS/infotainment unit that can be omitted on cheaper trims.
Entering Japanese addresses IS tricky, though; here, custom-built Japanese solutions outshine. There are (mainly) no street names; instead, you specify your location by filtering down from Prefecture (Tokyo), City (Ota-ku), Commune (Kugahara), District (1-Chome), Block (26), House Number (1).
Japanese systems allow you to enter it this way - with Google (or, even worse, Apple Maps), it's a bit hacky. You would specify it as Kugahara 1-26-1 and hope for the best.
On one side, it can't route through a bunch of valid paths. I first assumed it could be because of residents asking Google to cut traffic, but sometimes it's not even through residential areas. I see routes on the map that are avoided in favor of bigger loops, and when trying the shorter routes they're perfectly fine. Or perhaps it's the vehicle size and they optimize for SUVs ?
On the other side it will happily route you through paths that are restricted to specific categories of cars. It's up to the driver to carefully avoid them, but it really wants you to go through and reroutes you there when you deviate, so it's a huge PITA in areas you don't know and try to navigate while ignoring the navigation instructions. Cops seem to have noticed it, We've got fined the first time we fucked up, and now that I'm aware of the issue I see the cops in many of these spots basically waiting for the jackpot.
At this point at google, I wouldn't be surprised if they were selling your route to advertizers to "optimize your driving experience."
https://web.archive.org/web/20230129231747/https://www.waze....
https://www.waze.com/ads/
The weird part is Maps will accept them for search, but not display the plus code except when looking at POI (randomly selecting a point on the map doesn't show it for me, I only could get that from the /pluscodes/ map)
This seems genuinely useful, but as usual we're having the chicken and egg problem to get it adopted ?
But yeah, nobody uses it in the real World. If I need to share a point with someone else I'll just send them the Google Maps share URL instead, which has its own shortcode
https://en.wikipedia.org/wiki/Quasi-Zenith_Satellite_System
> The Quasi-Zenith Satellite System (QZSS), also known as Michibiki (みちびき), is a four-satellite regional satellite navigation system and a satellite-based augmentation system developed by the Japanese government to enhance the United States-operated Global Positioning System (GPS) in the Asia-Oceania regions, with a focus on Japan.
...
> The primary purpose of QZSS is to increase the availability of GPS in Japan's numerous urban canyons, where only satellites at very high elevation can be seen. A secondary function is performance enhancement, increasing the accuracy and reliability of GPS derived navigation solutions. The Quasi-Zenith Satellites transmit signals compatible with the GPS L1C/A signal, as well as the modernized GPS L1C, L2C signal and L5 signals. This minimizes changes to existing GPS receivers.
Its got a neat orbit that does a figure 8 with the apogee over Japan (see also https://en.wikipedia.org/wiki/Tundra_orbit )
https://www.daihatsu.co.jp/lineup/move_canbus/03_exterior_in...
This is addressed in the article:
> There used to be a big difference in driving characteristics and technology between premium and budget brands. Compared to a Volkswagen, a BMW used to have a more powerful engine, better handling, and comfort features like seat-heating and cruise control. However, a Volkswagen Golf now has similar tech as a BMW and with the transition to EVs, drivetrains and handling won't be the same differentiator as before.
> Now, the focus has moved to the interior. Infotainment systems have become a central part of that, so carmakers are coming up with unusual concepts to set them apart, both in hardware and software. CarPlay 2 is going exactly in the opposite direction, more towards standardizing the in-car software.
But I think a great differentiator for upmarket brands would be to offer as few screens as possible and as many buttons as possible. Electronics are a plague. Some features are very useful: mainly, route planning and the ability to play music in the car. But the rest is a nightmare.
I don't want to talk to my car; certainly not when passengers are sleeping during a trip, and even not when I'm alone. There's a subtle humiliation associated with talking to a machine. More importantly, I don't want to randomly target zones on screens to set up things, and I don't want to look at the screens because I don't want to take my eyes off the road. Give me buttons that have a fixed location, and that I can feel without looking.
Touchscreen is not better than physical buttons but at least I can still look forward while operating it.
This audi is the first time I actually started using Siri - and not because I like it.
Basically, their goal in life is to be a worthless middleman who takes peoples money while providing no real value to society.
It’s analogous to a TV and monitor manufacturer saying they are no longer providing non-wifi inputs, app side loading, or supporting screen casting from local devices.
Then the manufacturer would charge a monthly fee for the monitor to keep working.
Hopefully, next, the CEO acts surprised when reporting the resulting massive drop in shipments.
For me, there is no room for GM to become a player here. I have friends' addresses, etc. in my iPhone's contacts and they show up on Maps. GM isn't going to do that, so GM isn't going to get my business. I'm starting to look for a new car right now, and this is a hard no. If it doesn't support CarPlay and Android Auto (both--I've used both in the last five years, and might switch back again to Android in the lifetime of my next car), it's outside my observable universe.
Toyota should also plan to expand their manufacturing capacity to pickup GMs lost sales to anyone under the age of 60.
Lol
I wish any manufacturer offered a 250-class electric truck with a basic trim line. They probably want to sell more cars per battery for now.
That said, $10 is far too much for what it is, should be $5.
Why do they even bother charging - just bundle it in with the headline price. It’s a tiny portion of the revenue, nobody is going to think “oh that $87,345 Tesla is too much I’ll go for the $86,145 one”
I use Organic Maps on GrapheneOS and add businesses I want which aren't there (hence I had to look them up in Google Maps in Firefox Focus, which is painful enough that I don't want to have to repeat it).
GM wants to charge for a separate internet service tied to your car, that still goes over the cellular networks, in order to access these exact same features in their head unit. It is completely redundant. CarPlay already offers a near perfect experience in this regard, without charging an extra monthly fee.
Planned obsolesence. Without proper CarPlay/AA integration, car manufacturers get to decide when those whiz-bang infotainment features stop working. You'd have to replace the whole car to get those features back instead of just buying a new phone.
Without constant updates, software that is part a a larger ecosystem will eventually breakdown.
This is why we should be so skeptical of tight software integration with durable hardware (e.g., cars and appliances with operational lifespans 10+ years easily). Software has a pretty short half-life, especially software that integrates with internet services; vulnerabilities get discovered in third-party components and remote APIs shift out from under you.
Durable goods manufacturers have little skill or interest in long-term software upkeep (maybe they like the profits and the rent-seeking, but not the actual maintenance), so the most sustainable design is one where the software is easily seperable and replaceable from the core durable item. Manufacturer-specific internet-connected infotainment in cars ain't it.
No, the most realistically sustainable design is one where the software is built into the durable item and not connected to the internet, so it never changes and doesn't need to worry about security issues. You can keep using the thing until it fails from mechanical issues and can't be easily repaired.
If it's internet-connected for some reason, then the software should be open-source so interested parties can take over maintenance after the OEM decides they don't want to bother any more. However, I don't consider this "realistic": how many manufacturers are going to do this in reality? Almost none: they want you to buy a replacement device.
My car isn't connected to the internet, it has not received any updates. However Google enjoys messing with the Android BT stack for whatever reason (there was a period in the mid 2010s where BT on Android was dramatically ~broken~ rewritten every couple of years), so accordingly my car has lost functionality.
How would that work in a car, where people expect map updates and streaming music or radio over the internet?
For a car, I really don't know. I don't see any way of putting up-to-date navigation software into a dashboard without a significant security risk, which means needing the ability to keep software updated. Even CarPlay/AA are non-trivial and would need security updates.
It’s far from ideal but at least I have zero expectations for my next car.
There are also dash-mounted "carplay screens" that are both much cheaper and much easier to install but lack integrations/require hacky solutions to get, say, audio out.
This isn't "we're going to deprecate your car, buy a new one". People will buy them anyway.
It is "You're going to pay for AOL, even though we have the internet"
And when you sell your car, some other dumb schmuck will buy it used and sign up for AOL-of-cars.
I think a key thing to consider is that there are in fact three separate questions at play here:
1 - Does infotainment/software UI differentiation matter in the car market? Is there a significant enough market advantage for having better UI that anyone should care?
2 - If there is an advantage for better UI, is it enough of an edge that would compel you to build your own? Or is it the case where it simply has to be good enough?
3 - If there is enough differentiation to be worth building your own, is your company good enough at software to pull it off?
Personally I think the answer to #1 is YES. I think cars with better UIs - while not sufficient in and of itself - have a market advantage.
Where car makers start veering off from each other is the answer to #2. If you believe that you just need a "good enough" experience to not be actively awful, then you buy off-the-shelf. You see this with Volvo/Polestar and Google Automotive. The "skin" around the stock experience is minimal at best, with only minor customizations.
If you believe that being excellent at it confers some advantage, you'd try to roll your own. This would include folks like BMW and Mercedes-Benz.
Now, where the latter strategy really goes off the rails is question #3. That said, if you believe the answer to the first two questions compels you to roll your own - would you easily surrender to a third-party? Or would you at least try to level up your software orgs to make a serious play?
I might be unusual in my preference but I really expect people to have preferences as strong.
So unless it costs a lot of money to integrate, not picking AA isn't about building your own that's better, but it has to be about control. Or at least independance from a big tech, which I can understand.
Disney's "The Magic Highway" might be dystopian to some but the relaxing vehicle seems pretty cool to me
https://youtu.be/Vo4-rYNGEwE?t=199
But if you can call a reliable self-driving one-seater within 5 minutes guaranteed?
Every time I rent a car it's a HELLSCAPE of figuring out whatever crappy UI this brand of car created for that year ... until I get my phone hooked up. Man I just want to get to my hotel not futz with some garbage UI in the garage forever.
obviously i rather have carplay if i could but i am worry about security implication, plugging to a car i don't know
The tech giants are not component suppliers you can symbiotically partner with to add value to your product. They are predatory and parasitic goliaths. They are wolves more powerful than governments with designs on your hen-house. It is frankly insane to let them own the primary interface to your customer and auto will likely regret letting them get this far.
Google/Apple infotainment ventures cannot even be called trojan horses given how open they are in their investments and desires to seize the auto industry the moment tech/profit makes it feasible. Unattractive low-margin manufacturing keeps them at bay but they are gambling on "software eats the world" long-term. For now, they will drain every available high margin service for themselves.
My ideal "infotainment" is a button that lets me pair bluetooth and volume/skip controls on my steering wheel. I have a phone mount, like almost everyone does. I don't need a display on my dashboard.
AA/CarPlay (and everything else) are genuinely just distracting annoyances that take up space. Especially in newer cars where the screens are for some reason, no longer matte and often angled upwards so that they just blast reflected sun into your eyes.
It’s usually terribly implemented. iOS doesn’t support it for third party apps or first party podcasts, and apple music’s support is a shit show. Android is better, but it is hit-or-miss depending on the car.
I wish the government would declare this a safety feature and mandate recalls for incorrect implementations (including phones and cars).
Last time I had a car I had an old mp3 portable mp3 player constantly hooked up to the car with 4GB of music inside. I would just hit play and drive and would only adjust volume from the steering wheel if I needed to. If I had a car again I would probably do the same, that is the lowest distracting way to have music in your car.
Perhaps you should reconsider your driving needs vs safety?
Also, the Mazda infotainment in the mid-late 2010s was made by Johnson Controls. The newer systems might be too, I don't know.
IMO it's more about control over the user experience. You don't want your customers' UX to be dependent on the whims of Apple or Google, because now you're implicitly building a long-term dependency with a third party that may not be acting in your interests in the future. You're moving closer toward a future where the vehicle becomes commoditized, and now you have more trouble differentiating from competitors. And keep in mind: it's only very recently that the "Apple car" project was cancelled.
That said, traditional automakers are also famous (or infamous?) for sourcing tons of components (including infotainment systems) from the same parts manufacturers. But I guess at least that retains the ability to pivot and use it as a point of differentiation in the future.
Unfortunately, I don't think they're planning on offering it free to my model or earlier, and I don't know if the same deal exists for ICE cars.
If it was user-focused, they wouldn't make unnecessary changes to the UI every release. If it was user-focused, they would put more effort into refinement and fixing bugs.
No, CarPlay is Apple-focused.
Good. It works.
I use Apple rather than Google as I try to avoid Google as much as possible but it’s the same principle. I want long term stability in software, not new changes put in to earn a PM a promotion.
Drivers don't car, but I think car buyers actually do. Remember that many people are not buying cars for themselves but for other people, usually family. They fall into the trap of thinking that those other people might want such features, if not now then in the future. Look at automatic transmissions. I know many people who much prefer manuals, but they always end up buying an automatic because they believe that other people will want the automatic. And a few years later, all the cars are automatics. The same is happening with in-car entertainment systems. We buy them not because we ant them but because we think other people do.
Can you explain this? I guess maybe the devil is in "many"?
By the time they are old enough to be in the front they have their own phone and headphones and care less about a mounted display than they do about a TV.
But CarPlay is darn close; I'd not say I'd never buy a car without it but having it means I don't need to worry if the infotainment setup is crap or not, because I won't be using it.
Cars from the 90s up until about 2013 can be easily fitted with a $500 head unit upgrade, and support carplay quite well. With the right tools, it can be done in about two hours right in your driveway.
Cars from 2018 and up pretty universally support carplay, and it's generally quite well integrated into the car's infotainment system.
But, between 2013 and 2017, things were a complete mess. In-car systems were too integrated to be replaceable with a third party 2-DIN unit, but too primitive to run carplay/AA. People who have cars from this era either sell them (for less than they're worth, since only 21% of people will buy a car without carplay!) or put up with it for another 8 years or so until the car's wound out.
For example, my rustbucket '06 Toyota has a great sounding stereo with carplay but my sibling's 2017 Nissan is stuck with flaky and poorly integrated bluetooth.
Or, if you do want to upgrade your 2013-2017 car, you end up replacing half of the in-dash components with ones from a couple model years up, tapping into the car's CAN bus to recognize the new controller, and then running some sketchy scripts to patch the firmware to remove component protection since the VIN's don't match up anymore. Not for the faint of heart.
Mazda sells a retrofit kit for their 2014 (Mazda 3) / 2016 (other models) and newer cars that didn't come with native CP/AA. It's about $200 and DIY if you're the least bit handy:
https://mazdaparts.org/mazda-3-smartphone-mirroring-kit.html
You also have to update the software first, which is a 45 minute process that requires touching the brake pedal every 20 minutes to prevent the head unit from going to sleep.
I've installed it twice: a few years ago in a CX-9 and two-weeks ago in a Mazda 3. There's YouTube videos for it so you know what you're getting yourself into. I found the procedure straight-forward. Mostly you need to know which parts to pull off and where the bolts, clips and fasteners are. If you're someone who'd install their own dash cam, you can probably do it.
Turned it into a 15 minute job vs probably an hour+ of messing with the center console. Babysitting the software update takes the most time.
I did the "drill holes in the front of the USB hub" trick so I didn't have to remove the console. But then I dropped a socket and ended up having to remove the console to retrieve it and it only took maybe 15 minutes. I could do it again in 5 knowing all the steps.
I believe other car audio sites will similarly present you with installation packages. I’ve been using Crutchfield since I replaced the stereo in my 1st car in 1985 so that is who I went with.
Something I found interesting when I was hacking my VW to add Carplay (without paying £300) was that a ton of manufacturers in the ~2016 era use the exact same head unit and OS but with a different front fascia, different button layout and a relatively advanced UI skinning system. If you can be bothered you can add e.g. Audi track G-sensors to your VW or use a different skin.
https://www.amazon.com/Portable-Receivers-Display-Stereo-Blu...
By contrast, my 98 Jeep and '08 sedan were both easy upgrades to modern head units with standard DINs.
Not having CarPlay/AA was actually a massive reason I wanted to get rid of that car.
Wow, I'm impressed. I wouldn't have thought manufacturers nowadays would be forthcoming enough to allow that.
Do you know if it is something specific to (some?) Japanese manufacturers, or is it a form of info-display standard that would also make it possible with other brands?
3rd party headhunts from various manufacturers have a port for the iDatalink unit which allows it to integrate with the head unit. Both of mine are Kenwood but I know a few others have support. There are also alternatives to the iDatalink Maestro it is just what I have used.
I don't know what cars you drive, but they are not budget or Chinese. I recently rented a bottom-of-the-line Nissan '24 and just getting audio from my iPhone was a struggle - and there was absolutely no Carplay in sight. Most Chinese cars I've rented (all 23 or 24 models) have advertised Carplay, but what they really have is a half-baked phone mirroring, which maybe worked on ios11 or something, but definetely doesn't today.
Occasionally I’ve had something like a D or so, or a van, but even the last van I had had CarPlay.
I realize I will not be able to buy a car with no screens due to the backup camera rules. But there is actual safety value in that for me so I’m ok with it.
For example, I wouldn't buy a new car without android auto, because cars with it are readily available and hardly cost anything extra.
But if I had say $6k to spend on a used car instead of $20k+ on a new car? Your options are much more limited and the tradeoffs unpalatable. I know I can still put a phone holder in any vehicle and I'm off to the races.
I sold cars up til a few years ago. Countless cars got sold where Bluetooth itself barely works, and people are basically fine with it.
People buying used cars will take Bluetooth as a nice perk, but to think they make major decisions based on carplay in a 2015 Chevy Volt is funny.
That was my position until we bought our last car. We got it second hand, so we did not really have a definite set of required features and CarPlay was just nice to have. After having used it over the last 4 years, it completely changed though, and I can guarantee I won’t consider a car without it for my next one. It helps that around here all non-garbage car have had it for quite a few years now.
That statistic is complete bullshit.
Anyway it's pretty much moot because besides Tesla, all new cars support both Carplay and AA.
https://9to5mac.com/2023/12/05/carplay-popularity-iphone-tes...
A non statistically relevant random poll of Apple fanboys
...is there a handy list of which models these are? I'm in the market for a used car, and I'm perfectly happy to pay less for something without CarPlay.
Once you're in used territory, all bets are off and you'd have to poke around.
As the article itself says, CarPlay Original Flavor is a massive success, I'm in the "79% of drivers only consider a car if it has CarPlay" (bye-bye GM!)
1: “original” CarPlay. Shows up on your head unit
2: enhanced “original” CarPlay. Some cars can show the map or turn by turn instructions behind the steering wheel on a screen there
3: “new” CarPlay. This is where CarPlay fully takes over every screen. Nothing has been released, I think only one very expensive model has even said they’re going it.
Today e dry one has #1 or #2. The article is about a mix of #3 and maybe a version of #3 run on the car not the phone, like Android Automotive.
The lack of distinct names for things on both the Apple side and Google side (auto vs automotive) just makes this all very hard to discuss at times.
I don’t know when the first car got that, it’s possible it simply didn’t exist when yours was made. Or maybe it did and they just didn’t bother? I don’t know.
My car doesn’t have the version where it just shows Apple Maps, which I would love. Instead the next turn/etc. is clearly being sent to the cars operating system because it gets displayed the same way the cars internal navigation stuff does on the screen behind the wheel.
Dual screen CarPlay was released in 2019 https://www.theverge.com/2019/9/20/20875604/apple-carplay-io...
To be fair, at this point only the Ultium EV-based models won't have CarPlay, and they are struggling to build them anyway. Even if they do figure out the production issues and achieve the targets of their wildest dreams, they will still only account for around 6% of all the vehicles GM produces. They're not exactly betting the farm on it.
Does the author have no experience with Android Auto? The same happens there, if the app is on my phone and it supports Android Auto then it will automatically be available in the car, along with all media on my phone. This isn't a CarPlay only functionalitiy, it's just how phone mirroring works on both platforms. Author seems to think there is an extra step involved on Android. Perhaps by "Play Store" they mean the car manufacturer's own app store?
Source: I use Android Auto constantly in my own vehicle and in the 15+ rental cars I have every year.
Auto is driven from your phone whereas Automotive is running on your car directly.
I know this isn't the highest value comment in the world, but god Google's product naming/differentiation is terrible. I can't imagine a more guaranteed-to-confuse (or worse, more guaranteed to lead to misleading second-party marketing) nomenclature.
Android Automotive (AA) is obviously different from Android Auto (AA). You see, AA runs on the car and runs the infotainment system whereas AA is just screen mirroring for Android phones. AA supports CarPlay just like it supports AA. Apple doesn’t have anything like AA, CarPlay is equivalent to AA.
Lastly, it’s not like “Auto” is short for another word such that people would mistake the two. It’s clear as day.
/s
And don’t get me started on the Apple TV naming mess.
It makes for great sentences like: The other day I was using Apple TV on my Apple TV to watch a show on Apple TV+. Which just rolls of the tongue and is sure to cause no confusion. /s
It’s a mess for sure.
They lost a capitalization which makes it a bit ambiguous but their scenario is this:
There are two app stores being discussed: iOS (for CarPlay), and Android Automotive (the infotainment system's play store).
If you have Android Automotive on an infotainment system and an iPhone and can't connect your iPhone to the infotainment system, you have to download the app twice: iOS App Store and Android Automotive Play Store.
I don't know the numbers, but Android Automotive infotainment systems don't universally support CarPlay. Some only got it via updates over the last couple years (that's also a selling point of them, though, that they could do it via software updates and not a whole hardware refresh).
If it doesn’t, that’s a choice the automaker made. Either to disable it or to not write that part of the software.
I know Android Automotive already has all the code to support it. So for an android automotive car to not support CarPlay is 100% a decision.
The author seems to blame everyone but Apple for this, though. Apparently it's Google's fault that his preferred podcast app isn't available on Android. What about the app developer, who has decided not to support Android? And it's the auto manufacturer's fault for deciding to go with Android Automotive as the infotainment OS. What about Apple, for refusing to license any of their OS offerings to third parties? (Let alone not having an automotive-suitable version of any of their OSes anyway.)
Instead, Apple wants auto manufacturers to implement a complex, proprietary Apple protocol to give iPhones the ability to control most aspects of the car's infotainment functions. Why would auto manufacturers want to do this? Not only would the auto manufacturer have to develop their own infotainment system (perhaps even using Android Automotive, but perhaps something else), but then they have to do all this extra work to essentially allow drivers to not use their own system. (And while I agree that most auto manufacturer software is weak or outright garbage, that point of fact isn't really relevant here.)
I don't agree with the author that some car makers are not excited about CarPlay 2 primarily because they want to own the automotive experience. It's because of all the extra work they need to do to support CarPlay 2 (probably an order of magnitude what it takes to support CarPlay 1 or Android Auto), and doing so isn't really in their interests, outside of Apple users who might refuse to consider their cars without CarPlay 2 support. But this is a chicken-and-egg problem: CP2 doesn't exist yet in any car, so customers don't know what they're potentially missing.
I am still amazed that some people see it as a brain-free non-decision to buy into a status symbol.
I have an iPhone. It has things that I like and things that I hate. I have a brain that can process nuance, and it’s entirely possible to say nice things about some features and rant against others.
I bought my current phone knowing what I can do and cannot do with it. It is an informed decision and a compromise by necessity. I bought it expecting it to integrate with my car via CarPlay, and I would definitely complain if it either stopped working or if car manufacturers dropped it.
(I don’t care about CarPlay 2 at all)
I would be interested in CarPlay 2 if I were buying a new car when it’s out. Infotainment is still a huge weakness in many cars, and is definitely a plus to many users. Carmakers are free to ignore it of course, but unless they are maybe Tesla, they have no good reputation to rely on in selling their own custom infotainment experiences.
Yes and no. Android Automotive supports android apps, but GM's implementation (for example) still locks out android phones from using Android Auto. Yes, you can install the same apps, but the nice thing about hooking up your phone is that you're maintaining all of your apps, data, and state in one place instead of two and it's inherently different for as many different drivers as there are phones.
Yes, as far as I know, any automaker can choose to make the screen in the dash compatible with Carplay, at least the non fancy Carplay that will be able to show all the apps on the phone.
Zero reason, in my opinion, to not offer a Carplay compatible screen, other than to inconvenience car users. Not the CarPlay 2 thing, but just original Carplay.
There is an existing vibrant ecosystem of Tier 1 partners who can do the CarPlay (or AA) integration, lessening the burden on the Auto OEMs for such specialised integration work (Luxoft, etc).
CP2 is an eventual pathway (pure conjecture here) for Apple to offer other upgrades in the future, like an L2+ ADAS package, via a subscription model, with a potential profit split between Apple and the OEM.
Somewhat like what MobilEye currently offers and their current business model. Apple gets valuable realtime traffic info from such an L2+ offering, which can feed into making Apple Maps better with up to the minute traffic updates. And the OEM gets to leverage a valuable "sticky" offering that is bound to attract the Apple crowed (usually well heeled folks with deeper pockets). And they both get to monetize CP2 via a paid subscription model.
> Apple’s long-term goal is likely to find ways to directly monetize CarPlay. It could license CarPlay to carmakers at some point. But with the entire global car market amounting to only around 70 million cars per year and the car industry’s slim profit margins (ranging from 5% to 15%), it must reach a high market share to get significant revenue from it. However, carmakers may be hesitant to invest in what they see as a competing product.
I don’t think Apple sees this as a promising revenue stream. It’s just another way to encourage people to stay within the walled garden. That seems many times more valuable than any possible licensing fees they would get from manufacturers
Isn't the value of CarPlay/AndroidAuto indirect? As in that the value is coming from the data being generated to enable better mapping? Because the truth is that some of these mapping services' main utilities depend on a large network of users (e.g. information about congestion, road notices, etc[0]). In that way it seems beneficial to get your app onto as many vehicles as possible and even gives motivation for making CarPlay cross-platform (even if the main benefits only go to Apple users).
[0] Also, wtf AndroidAuto. You pull data from Waze users for things like speed traps, road hazzards, etc, but you don't let me contribute when using Auto? Why is there no button (or voice command) "report road hazard"? Hell, this'll even help you classify events for your ML models.
Damn, indeed I missed that section
> With their own software platforms, they are desperately trying to find recurring revenue streams, whether it's a subscription for seat heating or driver assistance systems
..........
They almost became a subscription, but BMW backpedaled at the last second[0].
0. https://www.theverge.com/2023/9/7/23863258/bmw-cancel-heated...
I agree this is the optimal way of seeing it. Apple did not do the same with App Store though. The initial vision was to just apply enough fees for hosting and vetting the apps. Now with the way they hold on to the 15/17/30% fee, they are treating it like a revenue stream instead of a "courtesy service".
In that respect a UI made by Apple will certainly appeal to an audience already demanding current CarPlay. On the other hand, it could be that the current level of CarPlay integration is already enough.
Because the solutions the auto makers came up with _sucked_ and as it’s not core to the operation of the car, I’d prefer it be done by someone that even approximately knows what they’re doing with software?
I am not buying the car for their infotainment features, I’m buying it for the car. I’d like it the entertainment bits were just delegated to the setup I already have…
I don't need my whole car UI to be taken over by my phone. My car maker solution to display and control speed, HVAC, gas tank/battery status is perfectly fine.
Bluetooth but for cars.
Such a missed opportunity.
https://en.wikipedia.org/wiki/Alliance_for_Open_Media
Or DDR?
Or M2?
I'd argue those are trusts worth boosting.
Yes, Siri is terrible, but the EUs new DMA law requires that I be able to select a different voice assistant on iOS too, and that should theoretically, once implemented, also apply to CarPlay.
In my opinion, and while I haven't used "Android Auto" I have used "Android Automotive" (the OS version, without a phone driving the integration), CarPlay the infotainment integration (besides Siri) offers a better user experience.
Super annoying when you want to set an alarm in a timezone you are going to be traveling to.
I really hope Siri gets that LLM-based upgrade in iOS18 but I can cope with it for now.
Car manufacturers have had decades to develop their software but almost all are difficult to work with, slow, give bad user feedback (full screen pop-up with entire paragraphs while navigating) and are cumbersome to get accustomed to when switching cars.
I’m one of the 79% that demands CarPlay, if it’s not part of the car I don’t want it.
Good I bought the car, now get out of my life. I’m not paying hundreds of dollars for GPS map updates when I can plug in my phone and use any other mapping app’s evergreen maps. Also, subscriptions for heated seats are patently absurd.
Typically this difference of one model-year can add thousands to the cost of the vehicle, especially because wireless car-play is so coveted. The experience of wireless is fantastic, but is it worth several thousand dollars? Maybe, but herein lies the trick.
Buy a dongle. They're about $100 for a good one. They can be tucked away in the vehicle. They work almost* as good as integrated wireless car-play.
*Maybe add 5 seconds to auto-connect when you get in your car.
We've upgraded the wired one with one of those dongles. It mostly works, but has some quirks. The three most annoying:
* Phone calls result in a feedback loop for the other end. Essentially, they break the in-car noise cancellation and playback the caller audio to the caller.
* When my wife pulls in the garage, my phone will connect - even though it should have been connected to her phone.
* The USB port that connects to the head unit remains power for a period of time after the vehicle is off. Annoying when I'm in the kitchen (next to the garage) and my phone keeps trying to CarPlay.
EDIT: I'm also realizing that I believe the car with Wireless has associated each of our keys with our phones. Despite them both being paired, it will prioritize the phone last used with the key. That's pretty handy for not having to fight with pairing.
Moral of the story: Don't buy a head unit from Pioneer. They suck ass. This is quite possibly the shittiest tech product I have ever spent money on.
But boy it’s enraging to use with multiple phones when you share a car with someone. It will stay connected to other person phone, while in range, and, if you to call that person, car will ring. It’s chaos… or maybe it’s just me.
Except I misunderstood that my year didn’t have wireless CarPlay, so I have to choose between CarPlay and wireless charging :( Hopefully my next car will have both.
Otherwise my trips are usually short enough. Also my phone's dual screen addon case (LG V60) blocks the USB port.
Prior to switching to Apple, I was running into major issues with USB-C ports being hyper sensitive to physical positioning. Charge would always work, but Android Auto would simply drop out seemingly randomly. Didn't matter on cable. It was just something that would happen.
Tapping on the screen for the next song or pressing the steering wheel button to do the same can take a half second or more. Doing it on the phone is nearly instantaneous, just a very tiny audio delay.
Using a cable? Never any delay at all. Unnoticeable.
Between that and the fact that it can really burn battery I’m happy to plug it in every time.
This problem didn't exist at all before going wireless.
The delay though… ouch.
Bought a dongle for £55, works fine.
Love having a car with a steering wheel with real buttons, and climate controls with real knobs.
It feels like around 2018 is the zenith of Human Machine Interface in cars and it's all been downhill since, as they cram everything in a fucking touchscreen.
"Kinda works" for a while, with a noticeable delay when changing songs etc.
Actually pondering replacing the infotainment system itself to get wireless airplay.
Wireless charging = heat.
Both = furnace.
Initial connection is a bit slower than wired, maybe 20 seconds or so, but it's up and running by the time I'm moving my car.
There's very small amount of input lag for stuff like skipping songs or pause/play. I'd say that lag is almost exactly the same as when I used to only use bog-standard Bluetooth to connect to a head unit with my phone so I think that's just the downside of a wireless connection-- wired doesn't have this lag.
CarPlay v1, however, is an absolute requirement. It works great and gives me pretty much everything I want.
But if Android Automotive runs CarPlay, I’m not sure how much I care. And I’m a huge fanboy.
I’d like to be able to use Siri to control the temperature and a few other things. And taking over the whole screen could be nice.
But Apple is such a control freak about things I have a hard time seeing most car makers agree to it. And just about every industry Apple has touched has learned the lesson that Apple takes things over. Car makers clearly already dislike the fact that CarPlay is so popular, I don’t think they want to give Apple MORE leverage.
I find the idea of Android taking over the industry a problem (monoculture = bad, manufacturers are clearly not up to it themselves). Apple could be a hedge. But they’re not a play nice company even when they should be.
They haven’t learned humility again after losing it during iPod/iPhone rise.
Tangential I know, but I giggled at yet another example of Google's incompetent branding and product naming. How a company filled with so many brilliant people can be so bad at brand strategy is baffling to me.
The fact that it's still confusing with android auto being another thing is I think partly on purpose, to make it look like Android is doing more than what it's actually responsible for.
[0] https://developers.google.com/cars
I can understand this kind of thinking in a here-and-gone used car lot.
I'm sure there's a few PR/Marketing people in there. It's not all "Good computer thinking people"
Google isn't a plucky startup with just a few dedicated engineers.
They made great long term corporate value and customer relationship decisions then, for their context.
When big corporations full of talent get sloppy, it is usually either a very competent emphasis on short term financial growth or value multiple growth. Or the result of multiple competent actors infighting or working at cross purposes, due to a lack of strong internal alignment. The latter jungle case saves a CEO from having to deep dive, they can just see what happens and select winners.
Either way, by competent I mean someone is accomplishing exactly what they intend. Optimizing their net worth given the company's incentive and power structures, in the absence of a CEO who cares more about the long than the short term.
It only takes that one difference at the very top to trickle down. The greater the talent, the greater the effect.
> There used to be a big difference in driving characteristics and technology between premium and budget brands. Compared to a Volkswagen, a BMW used to have a more powerful engine, better handling, and comfort features like seat-heating and cruise control. However, a Volkswagen Golf now has similar tech as a BMW and with the transition to EVs, drivetrains and handling won't be the same differentiator as before.
The thing is: that didn't used to be that way. You can blame it on the transition to EVs, but part of that transition seems to be that a bunch of manufacturers decided not to build their own platforms, motors, etc and are just licensing from other manufacturers*. The article's correct to note that flattens a lot of the value proposition of any given manufacturer, and if that's forcing them to lean in a lot on the software, that's a weird position for an automotive manufacturer, because that's never been anyone in the industry's strong suite - there's a reason 80% of drivers won't buy a car without CarPlay.
It's also notable that the brands who do seem to be going in on CarPlay are those that still make a point of building their own engines and platforms - Aston, Porsche, and even Polestar tries to differentiate itself there.
* to be clear, this was happening before EVs, too - BMW put out a car that shared a platform with a Toyota, in a move that should've caused a plague of locust to descend on Munich if God existed and had a driver's license, and Stellantis put a Lancia badge on a Chrysler a couple years back.
How is it that BYD and a bunch of chinese manufacturers able to produce ~10k car but nobody else?
I don't think you realize how simple cars used to be. Engines were extremely simple and purely mechanical, essentially 50 year old technology. Transmissions were simpler, either manual transmissions or ATS with only a handful of gears. Engines and cars were made entirely of cheap steel. You had no expensive screens, no expensive sensors or electronics. You had no expensive research on structural design and materials science. You had no expensive emissions systems. You could fix nearly everything in the car with hand tools.
The big differences to engine and transmission are interesting, I believe you that it's a big difference but I'm not sure how to evaluate specific numbers for it.
Building the entire car with more expensive structure and paneling sounds like a pretty bad thing to effectively require, especially when most losses are aerodynamic.
For backup cameras I was including the monitor in that price. You can buy kits for under $50, so I don't think $200 is unrealistic for something built in.
I will say that having a thousand microchips rather than a dozen is probably because it's cheaper that way.
The cheapest configurable Corolla here (Australia) is $35,857. More than 50% of the median yearly wage ($65000)
To me that's not a budget car. The closest thing would be a Kia Picanto for about $18500
It seems maybe tesla is the one fighting it. I think the original model S used steering wheel stalks from mercedes?
Now they do so many things themselves (or don't do in case of stalks), to the point of making their own chips for the machine learning stuff.
Re: "global manufacturing" - yeah, but there's also a choice being made there. I own an older BMW, and there's been a very clear shift away from what was once a differentiated product to a kind of blah middle-object, which is almost certainly in pursuit of a larger market share, but again, that's a choice. I could tell you what the value prop of a BMW made between 1975 and 2005 was, I can't really tell you what it is now, and that's a choice it seems like a lot of manufacturers are making.
I'd be curious for someone to explain what this is, and what happened in 2005 that changed this.
I really don't know what happened after 2005, and it wasn't overnight, but the general driving experience of the cars has declined every generation since then - steering's gotten a bit more numb, suspension's not quite as communicative, engine noises are more muffled, and they just seemed to lose the thread over the years. I'm not sure if some of it was leaning into technology at the expense of the driving experience, but aside from the one or two models per generation that still have some of the old soul, they really don't seem to know what their sales pitch is anymore.
Why is this bad on BMW? Doesn't Toyota make the most reliable cars?
And no, Toyota doesn't make the most reliable cars. It might have been true for some period of time, but isn't true for every model and every year (even Corolla, often suggested as a very reliable sedan, has some versions that are in fact infamous for their tendency for oil sludge buildup that result in engine failures).
Perhaps the group behind OBD II?
https://www.digitaltrends.com/cars/everything-you-need-to-kn...
Elon was, for a long time, the genius billionaire who saves us from climate change and gets us to Mars, so even in liberal cycles he was liked until he bought Twitter (or until he became a sh1tposter on Twitter).
Some people just like the software, some don't. Me personally, find it way too much, too complicated and I am not going to control my AC over touch screen, sorry.
Obsolescence - ‘your phone is too old to use with this car’ or your car being too old for your phone. Double points for software updates to either which disable functions or features.
Vendor lock-in - are the players mutually exclusive? Does having a phone from one manufacturer limit my car choices?
Realistically it's going to be the other way around
In some 2024 Mazda models they are finally letting you choose to keep the touch screen working while moving, letting you use either the wheel or the touch screen. I don't think either company will still be disabling touch screens/wheel only inputs for CarPlay for very much longer. Similarly, the new infotainment platform for Lexus cars has done away with weird input devices ("Remote Touch") and you can just poke the screen with a finger.
This is a really good case study in how difficult it is to find a balance between co-branding and maintaining a consistently high-quality design system across co-branding partners. There's a massive amount of work across UI/UX design and implementation done at Apple that assumes that widgets are not only using a serif font, but a specific serif font with specific kerning; that color-primary-60 and color-primary-50 and color-for-text-on-top-of-primary-60 are distinguished in a very specific way.
(Light/dark mode and localization efforts force a degree of flexibility here, but there are still a finite set of QA targets if you focus on primary language markets.)
But what happens if multiple partners want their own primary color and font? This suddenly has far-reaching, costly ramifications across multiple organizations. Even having planned your APIs from day one around color and style customizability doesn't guarantee that this can be done successfully. Taken to an extreme, frontend engineers (not just their embedded designers) are practically required to hold the context of all future potential customization needs in mind when implementing a component - a nigh impossible ask.
Which is to say that there are few companies that could pull off what the OP posits that car manufacturers are requesting, having a world-class interface that is customized to their brand. That's a tall order even for Apple's depth of talent.
Probably unpopular opinion, but I preferred mechanical gauges anyway.
Why can't we let the car deal with car-related functions and Apple deal with playing podcasts and showing maps? It seems like the best of both worlds to me. Apple is solving a problem that doesn't need solving.
Don't get me wrong, I'm sure Apple could design a better interface than car manufacturers, but is it really worth the hassle?
I also really don't like wireless charging. It spares the port from constant wear but my device tends to get really hot and the charging is very inefficient. I plan to go back to slow wired charging everywhere to put less strain on the battery. I already have in my car and by my PC. Unfortunately my XC40 doesn't support dual screen CarPlay so I use the Google Maps in Android Automotive for navigation and use CarPlay for music, podcasts, etc. Switching from AA Google Maps and CarPlay and back is not a high friction interaction though so it's an okay solution. I'm also able to deactivate wireless charging so I'm pretty happy with this solution.
(a) a standardized mechanism for attaching and powering your phone on its dashboard (think a "mini VESA mount" with options for USB/wireless charging)
(b) bluetooth HID-compatible physical controls on the steering wheel and/or center console so you don't have to reach for your device
(c) a minified car "radio" that only offers AM/FM radio and bluetooth audio connectivity
All other controls should be physical, copied straight out of a 90s/00s car.
- TV's that only display images without syphoning user data;
- One-time payment app purchase without subscription.
- A new car is an O($10k) investment.
- A TV is an O($100) investment.
- A monthly app subscription is typically O($1).
This results in the worst possible outcome: I need to take my phone and read the messages on my phone screen, creating a safety risk.
But it does show that there's a message but doesn't show the message itself: worst of both worlds.
The only downside I've experienced with it in Japan is the GPS can be wonky in tunnels whereas the car's built-in GPS seemingly doesn't.
Cars simply have bigger, more power antennas because they don't have to be packaged in a phone.
One of our cars has a wireless modem that can be used for a hotspot. I've driven to plenty of areas where my phone has no service (think remote and hilly terrain), but the car still has two or three bars.
I’m willing to bet that the soonest any large OEM could offer CarPlay 2 would be next year. The development process is just too slow to turn around something.
As soon as customers demand better, there will be better/easier integration between the two components.
Now, as I face the inevitability of replacing my car due to dubious environmental regulations (current car emits less than modern but twice heavier ones), I'm genuinely concerned about the current state of car infotainment systems. They feel overly complex, bloated, and fragile, prioritizing features over functionality and user experience. The shift towards these systems seems more like a downgrade, sacrificing reliability and usability for unnecessary add-ons that don't improve the driving experience.
Anyone in the same position? What car would you recommend that meets environmental standards but keeps the infotainment system super simple, or even non-existent?
I guess YMMV. I've recently had the chance to test drive some of the latest models from BMW and Audi, and honestly, I found their infotainment software to be quite laggy and sometimes unresponsive. It felt inexact and bloated, reminiscent of an old Windows XP laptop with a Samsung UI skin. While the CarPlay integration was ok, it still didn’t surpass the simplicity and efficiency of just using my iPhone mounted on the dashboard as a makeshift heads-up display.
I have no idea if it’s the car or CarPlay but could I please just have a toggle somewhere?
Yet when I enter my car and CarPlay starts for some reason maybe 50% of the time (not all of the time even) it randomly decides to play the last song I streamed in Apple Music 2-3 years ago. WTF Apple?
That's the downside of choosing apps that are not cross platform. Don't do that if you want to have the freedom to use different systems.
I don't like that car manufacturer has to use Apple / Google software, why can they not make an app that you install and your phone connects to the car in a useful way.
The reason is that only Apple apps have the required permissions and system access to do it.
I've had some rental cars that wanted me to install a manufacturer app. Super annoying! Just let me plug my phone in and project my screen and speakers. I don't want to get adjusted to a new UI with each car I drive, or worse sync stuff. I also have zero interest to find out what software a car company will develop. Car companies should make cars and leave infotainment to my phone and companies that are experts in it. Built-in infotainment in cars is even worse than "smart" tvs.
The screenshots there make me feel pretty good about keeping my car from the 90's if that's the current state of new cars.
The fact that everyone seems to think CarPlay is great though I guess suggests that the problem lies with the implementation details with the VW.
> there is no match for seamlessly transferring the phone's state to the car.
I’ll point out it’s not at all seamless. In fact it’s extremely cumbersome to fish out a lightning cable and plug the phone in. A good car infotainment system where I didn’t have to connect my phone at all would be seamless.
This seems like a missed opportunity for Apple - they can release a screen less device bundled with a cheap 4g connection and smooth out other integration issues for sharing / setup etc.
By this we have no vendor lock in and just making sure there is a protocol downwards compatibility…
What's the big deal about CarPlay 2, then? Who cares?
The one cool feature was that Google maps driving directions showed up on the front instrument cluster in addition to the center console. That’s a feature I’d like to have more broadly.
Since my older cars do not have it, having a great mount (Peak Design) makes a huge difference, but damn I want it integrated if only for the bigger screen.
The instrument cluster?
Any other safety-critical functions that the iPhone isn't isolated from?
(No, I don't even want Bluetooth, ugh).
Everywhere else in my house, I try to make gadgets independent of one another. I remember when a printer worked only when it was hooked to a PC to be its server. That was bad. It’s much better for my printer to hook to the network directly.
Some thermostats require a phone to program them. That’s bad. The thermostat should have full functionality as a stand-alone device. My dishwasher requires a smartphone app to work some of its features. Bad. My old dishwasher was fully controllable from the panel on front.
A car is no different. I don’t want to have to plug my phone into my car like a dongle, even wirelessly. The car should connect to the network itself and have needed applications like nav and music. Maybe car-makers suck at making this software, so maybe Apple or Google should make it for them. Or maybe carmakers should get better at it.
But no, I’m never going to think that plugging my phone into my car is any better of a kludge than is forcing me to get out my phone to work my dishwasher.
I love that when I search for something with Google Maps it knows, typically in a letter or two, what place I'm trying to go to. I don't want to use a different music service in my car -- it has all my playlists and my "likes" already.
Do I want Ford to write its own implementation of Google Maps or Pandora? Why? Seems stupid to have 10 different implementations of the same service.
Having to hook the phone to the car is s giant leap backward. We’ve got Gmail: email from any device. We don’t say: My phone has all my email, I hook it up anytime I want to read mail, if the battery is dead or I forgot my phone, I’m screwed. No, we use Gmail from a tablet, phone, PC, etc.
But for some reason people believe it’s an awesome advance that we can hook the phone to a car, as if it’s impossible to build Google Maps into a $30000 car.
This is the same logic as why some people would rather buy the dumbest TV they can and let the smarts live in an external replaceable box.
I’m dreading the day when they start shipping Chromium with your car and thus force you to jump through 0-day hoops.
Is Tesla basically this? Do they still update old models?
I’d rather connect my phone, for which I’ll update the hardware 4+ times over the time I own my car, and get the corresponding OS and software updates yearly or more frequently. Send that data to the car’s reliable and dumb screen(s) from my chosen phone and I’m in control and happier for it.
For the love of god, keep them separate. One is critical vehicle functionality. There other can crash/reboot/have connectivity issues, without me being concerned about knowing the engine is overheating/battery pack is dead, a tire is blowing out via TPS or I'm speeding.
I don't want a car that is CarPlay only, guess what, my car is not an accessory to my phone. The genius of current Android Auto/CarPlay is that the car head unit can act as a mostly "dumb" head unit for my external mobile processor.
Except it isn't. Can you start or continue driving without any of that info? Then it isn't critical.
Critical functionality is:
- steering
- drivetrain
- brakes
- tyres
Except Tesla is doing it right. Mercedes software is a mess. It's hard to explain why, but use it for a few days and it's just obvious.
Car makers what "differentiation". Uh huh. That is literally, exactly, what consumers do NOT want or benefit from.
I want consistency so that I can get in a rental car, or my friend's car, or my new car,
and have sh-t just work.
The LAST thing I want is a curated subset of familiar apps and controls, plus crappily designed but "cool looking" replacements all about convenying Brand(tm) Values.
My car (a Volvo) has two of these three apps built-in.
I stopped reading the rest of the article.
My 2022 Subaru Outback head unit frequently crashes or is really unresponsive. My iPhone never crashes.
Car infotainment systems are garbage. I’d never buy a car without CarPlay. It’s maybe the #1 measure I have for vehicle quality when I rent on vacation.
So it seems that the more features I use via carplay the less it crashes. It's not perfect, but it certainly helps.
It’s really quite a shame as they’re fantastic cars otherwise. But yes, it’s never CarPlay that does it.
However my wife has an iPhone and although going next to the same antenna does crash the system as well, it does so very briefly and just resets the connection or something, and as soon as we go out of range it starts working again.
So it does look like there's some difference between CarPlay and Android Auto even when using the same exact car with both.
No idea why this happens and I can only assume the antenna is transmitting in a frequency that interferes with the wifi connection (it does not require a cable like older VW cars) which triggers a bug in the Android's stack, but Apple's stack seems more resilient somehow.
I would love to know if anyone else has had the same experience because I find it quite baffling and very annoying when I forget to disconnect my phone when going through that antenna.
Infotainment systems are prone to runtime errors and crashes, if you interact less with the internal workings of the infotainment system itself, it does a somewhat good job of just passing audio through and you have fewer sessions with a crash.
That said; i have noticed a disparity in what the infotainment gps navigation can do and what carplay can do- carplay is often limited to the center console, but the built-in navigation can go into the HUD of the car. An annoying difference.
On the other hand, the voice command system works as badly as it did when I got the car, while Siri has gotten worse.
I discovered not later than yesterday that if I plug a Nintendo Switch to one of the USB ports (to recharge it), the interface disappears.
And this is still one of the least garbage infotainment I had to endure to this day.
The insane lag to enter anything on google maps on my 2024 Yukon is ridiculous. Worse, the fact it makes me TYPE in the touchscreen instead of letting me do this on the phone is yet another reason to completely ditch it. Google could easily let me control it through the phone (as they do with youtube on my tv) but I doubt they care enough to do that.
It should be straightforward, pick a direction on my phone, beam to car, car opens it automatically in google maps. That is not difficult.
Result is that it's smooth like a tablet, even with it's quite large 12" screen. Been very pleased with it so far.
It also supports Android Auto and CarPlay though, something I had as a requirement.
[1]: https://9to5google.com/2021/09/06/android-automotive-renault...
Also, the integration with the iphone was top-notch. It understood that the iphone provided navigation, so if I pressed the nav button while on the radio screen, it would activate google maps from the iphone. The only thing which wasn't great was typing in some long address, which is why I connected the iphone.
Rented some kind of Audi a few weeks later, and it had the usual crappy integration. "nav" is the car's laggy navigation, etc.
I installed a new stereo for my old car today, intentionally avoided CarPlay and got a simple non-touchscreen JVC. $60, 15 minute DIY job, BT paired instantly, has aux too, and sounds great.
The point being, further hurt the sales of any cars like this, because really, it could be true.
That said, a car without CarPlay is simply out of the pool of possible cars I would buy. It's a requirement.
I don’t remember which specific brands had CarPlay and which ran their own infotainment systems, but I remember a couple years ago visiting dealerships of BMW/Mercedes/Honda/Toyota/Kia/Subaru/Ford and walking away either relieved that the car had CarPlay or internally cursing at their first-party infotainment system and promising myself to not even bother trying next time.
CarPlay and Android Auto give you the choice of your own app and full integration with the rest of your listening life.
And to be fair I prefer waze, not just because it makes driving simpler, but the community feeling you get by marking stopped cars or other obstructions on the road for others to pay attention to.
Unless automakers organize and come together to build a common os/ui layer, I don’t see apps like that ever being present in official infotainment systems.
And audio book apps change all the time too - local variants with local libraries, language learning apps etc. There is quite the variety of stuff you could do in a car.
And even if you want to use only audio and maps, then there are still a selection of audio streaming providers and map providers around.
Plus CarPlay it's pretty unstable and requires cable on some cars.
camping hiking etc bring me places without reception.
driving between towns here (BC, canada) often mean long stretches of road without reception.
After all, it’s just a protocol to let your phone use the display (or part of it) as a second monitor.
It’s been proven time and time again that vertical integration is a massive money earner and you can bet that if Tesla were to release a smartphone then support for that in their cars would be a software update that follows immediately after. But since there’s no plans for a phone, the next best thing is a subscription-based infotainment system.
I pay for premium data only because of Spotify and live traffic. Which I wouldn't have to do if it had CarPlay.
Source? I cannot find a single reference to Apple collecting fees, much less onerous fees, for implementing CarPlay functionality.
Automakers don’t like CarPlay because it makes cars fungible. I like CarPlay because it makes cars fungible. Apple seems like it would give away CarPlay (at least for now) to broaden the reach of its platform.
I could care less if my car has Apple or android auto at this point. I would rip it out immediately if it had one. Any always on connectivity would be removed.
Car manufacturers are increasingly selling off your private data and leveraging all of these technology upgrades you paid for to do it. As soon as the car is connected to the internet, it’s shipping off your private data and selling it to data brokers. Manufacturers are hiding behind their wall of text called “terms of service” to do so [1]
In some cases the manufacturers are reporting your driving history to insurance companies so they can get any reason to bump your rates or deny you coverage . [2]
My dream car is now a “dumb” car.
Give me a car with a simple backup camera, manual transmission, and regular sized vehicle (no trucks or suvs, fuck that).
[1] https://foundation.mozilla.org/en/privacynotincluded/article...
[2] https://www.nytimes.com/2024/03/14/technology/gm-lexis-nexis...
So what exactly are you ranting about here? Why this rant when you don't seem to know what those technologies are?
Why can't we stop companies selling our data? You'd think it's easy:
--- very clear opt-in method for having your data sold
--- rejecting cannot prevent regular use of services
--- heavy penalties for breaking these rules
Problem is that no politician wants to touch this because
--- manufacturers sell data to subsidize the product
--- if they can't sell data, costs are going to shoot up
--- if they do this in response to a law, they get to raise costs even more because it affects the whole industry at the same time and there's a clear scapegoat
Consumers care a LOT more about their cheap, connected devices than their privacy. Because getting by your data like [2] happens to individuals, but costs affect the group.
EDIT: To clarify, the MP only suggested that costs would go up and people don't care. The rest is my personal speculation.
I'd much rather see stronger privacy legislation that bars companies from selling your data to third parties -- period, no exceptions, no opt-in terms that offer customers more features or services if they'll agree to privacy violations.
And then actual enforcement, with fines high enough that they can't just be shrugged off as the cost of doing business.
While I don't think we'll ever get quite there, I do think we're going to be moving in that direction. I do have a car with connected features, and I live in California, and I've already instructed them that they may not sell my data to third parties. I don't know if they actually follow the law on that, but it's a start.
This isn't meant to be a dig of Apple vs Android - I think I would feel the same about CarPlay if i was on iOS.
I think we just value different things. As someone who wants to be able to drive, navigate with google maps, listen to a podcast, and handle incoming messages, but do it safely, these integrations are incredible.
Obviously good. Driving is incredibly dangerous to other people and insurance companies are the most motivated to make it safer.
CCTV dates back a hundred years. The vast majority of dumb phones have had cameras since before the advent of the smart phone.
Mazda, cx5 <= 2016. Small. Has usb sd, simple knobs, backup camera. It reminds me of when apple removed the escape key, I had to learn how to buy a used mac.
And by 'rip out' you mean "disconnect your phone"
Every partnership Apple enters must result in Apple winning and the other party simply being a subservient cog
So, no one wants to do business with them in new markets where Apple doesn't have leverage
This is why they couldn't get a manufacturing partner for the now-dead Apple car...no one wanted to be the Foxconn of cars (they do all the work, Apple gets all the credit)
I think Apple failed on cars because they are significantly more complex then hobbling off-the-shelf electrical parts together and slapping an OS on it. Think of how many companies can assemble electronics and make OS. Its easy, literal sub 18 year olds make prototypes like this.