Finally a car manufacturer realizes it can't out-compete a software company at making good software.
The reason infotainment sucks so hard is that car companies want to "own the experience" and also want to cut costs, so they underinvest and create poor experiences.
Nope. There's a current mismatch between automotive hardware life cycles and consumer software life cycles.
This Volvo hardware is going to be a minimum of 3 years old before it ships and will have to function for 10 years+.
How good is your 10 year old phone? How does it match up to the average consumer experience today? How's the speakers, camera, audio, OTA software, and display compare to what's available today?
Not owning the software ecosystem is a good first step but will be severely hampered by not aligning hardware development cycle expectations.
Counter-example: My vehicle's infotainment unit (new in 2016) has received exactly zero updates. If Android is supported for "only 2-3 years" that's still a step up, not a step down from the status quo.
PS - I should clarify that the engine's computer received at least two recall updates. But Android doesn't try and replace that.
This, Android things, for example, is only offering guaranteed updates for 3 years which is not nearly enough support time... and people usually Volvos for 15-20+ years... Even Microsoft understands this better.
Everything in a car has to be certified for safety. Infotainment systems integrate deeply with cars - they consume and expose a lot of information that isn't guaranteed to be expressed over standard protocols, and most of them aren't even in consistent sizes or shapes.
An Android Things IOT device is guaranteed 3 years of updates. This is hosted on GCP and Google picks up the cost for this. However, if an OEM wants to extend the service beyond 3 years they have the option to do so (for a price I think).
Which used to be the case also for cars -- we even have a standard, ISO 7736 (aka DIN mount), for car stereos that can be swapped when the consumer wants to upgrade.
So in my 2002 el cheapo car, I can buy and install in ~ 15 minutes a 2018 infotainment system with a 7" IPS touch screen, Apple CarPlay, Android Auto, Waze, Spotify Connect, all the bells and whistles you can imagine.
But of course car manufacturers realized this was costing them sales, both in form of less incentive to buy a new car, and less incentive to spend $2k on the "upgrade" option for the stereo when a better aftermarket unit is <$1k, so they moved away from this standard.
You can generally get replacement centre console surrounds that add DIN or Double DIN holes to cars that didn't come with it. I've done it to a few cars for friends. Search eBay for your car make/model + "double din". It's not available for every car, but it's available for a lot.
You can also 3D print your own. There's free downloadable models of the DIN slots that you can add to a 3D scan of your cars centre console. Then depending on your car's interior trim, fibreglass it, or trim it in leather/whatever you want. Or you can just buy the DIN/Double DIN housing, dremel out where your stock headunit is embedded, and mould it in. Plenty of people on car forums do this and there's lots of guides online.
Considering the absolutely insane premium prices for sub-par entertainment systems in modern cars, I've personally found the above effort completely worth it.
I'm currently adding a RasPi and 7" touch screen with GPS for Google Maps, 1TB SSD of media, Spotify (via Wifi tether to my phone), USB ports in glovebox for USB sticks of media or charging phones, and more, to a 2012 car that came with an embedded CD/MP3 player with no nav, no screen, and no DIN slot.
True but now they can offer their own systems that integrate far better with the car electronics. Such as controlling the HVAC or offering rear view camera in the in-dash system. Those are good opportunities for lock-in which they'll no doubt take advantage of.
The TV ecosystem solved this by splitting the screen apart from the set-top/streaming box (and then put them together again in the form of smart TVs, but most people I know just disable the internal smarts and use their own stick anyway.)
Why not just specify a "dashboard control tablet" form-factor, and have an empty spot there, expecting the user to supply their own (and replace it every year or two, as people tend to do with other mobile devices)? It'd be this decade's tape/CD player module.
This is yet another example of a car company unable to realize that they just don't make good car infotainment dashboards. This will just lead to further fragmentation of the Android ecosystem.
Maybe a dumb question, but do companies often support -both- AndroidAuto and AppleCar play? Or do they do lock-ins and partnerships...because if they lock you in, it really defeats the purpose.
My experience is the same. For day to day, bluetooth phone<->car connection beats Carplay or Android Auto. Wireless autoconnect is key, should've been in the V1 of those standards.
I never bother with Carplay for less than a 1 hour trip.
We use CarPlay all the time, it works almost instantaneously after plugging the phone and it is always responsive. The primary beef that we have with it is that you cannot use Google Maps. Though Apple Maps is definitely many times better than the built-in navigation system.
As you say, it is probably a problem with some hardware.
It seems based on other comments that it varies wildly by manufacturer. Several others in here mentioned issues with the same vehicle I have. Really unfortunate, it’s clearly better than the built in software. Maybe there is hope that Volvo will issue an update that improves stability.
Eh, it’s instant on in my car (a Citroen) but I suspect they start booting the electronics when I unlock the car. In the 3 months I have had the car it have never crashed but twice CarPlay refused to find the phone and required a reboot of the phone.
You need a better, stronger lightning cable for your car. Apple OEM cables can’t take the wear and tear in a car. As long as my lightning cable is good, so is CarPlay. You can also go the wireless route as an alternative
Unfortunately, mobile processors seem to still be undergoing the same rapid increase in power that desktop systems used to, and the OSes and applications are being built for top-of-the-line phones. Even my iPhone 6 struggles with modern versions of iOS, despite being under 4 years old.
Hardware improvement cycles are not some the same curves they were 10 years ago. So, short of a massive breakthrough, the improvement we can expect in the next 10 years will much, much smaller than what we saw in the last 10 (which was already much smaller than what we saw in the 10 before that).
They can still shoot themselves in the foot by intentionally buying underpowered hardware, but that won't be the fault of hardware trends, it'll be the fault of making a bad initial choice.
I mean, the iPhone is only 10 years old, and the difference between the first generation and the iPhone X is astounding.
That said, I generally agree with you: I think a lot of the advancement we'll see in the coming years will be in the VR/AR space, which will present a different enough experience that users won't be comparing it to existing interfaces.
I think there's a larger story, which I'll explain through a personal anecdote. I purchased the Samsung Gear 3 LTE smartwatch because I've always hated how mobile phones get in the way of personal interaction, but then was very disappointed to find out it ran Tizen and not Android Wear, which means no Google Assistant.
So one the one hand you have Google Assistant: great voice recognition and queries against the Google search engine, and on the other hand you have Samsung which doesn't own a search engine and can't compete in the voice search market.
So it looks like Google is going to dominate the entire hands-free computing world (smartwatch, cars, and home assistants)
Google I/O is tomorrow so hopefully they'll announce some Android Wear with LTE and voice assistant.
Wear OS currently accounts for about 18% of the wearable market share. There's also a WearOS session scheduled for I/O so they're still interested. The major hurdle has been the SoC. WearOS devices are still using the antiquated 28nm Snapdragon 2100 SoC which was created in 2014.
The reason infotainment sucks so hard is that car companies want to "own the experience" and also want to cut costs, so they underinvest and create poor experiences.