CarDroid: Android built into cars. Please vote so we can get funding | Hacker News Reader

CarDroid: Android built into cars. Please vote so we can get funding

nissaninnovationgarage.com

59 pointsromonopoly50 comments

Can already do some of this, though not as integrated - an OBD2 to bluetooth adapter plugged into the OBD2 port and Torque running on an Android phone. I don't like the idea of tying something that will obsolete so quickly (an Android device) to a car, though - look how badly supported 2 or 3 year old Android devices are. Are ones in cars going to be supported any better? I'd rather a standard dock so that a smart phone can be slid in, held in position securely (not a stupid suction cup on the windshield), get access to power, OBD2 data, audio, steering wheel audio remote controls, and a GPS aerial (given it won't be under the windshield anymore)

Aren't we currently in a situation where your car software / hardware just doesn't update - ever? Apart from very recent models, I don't know of many systems that are even possible to upgrade. And if they do, they often require payment.

I don't think Android situation is much worse here, even if you were stuck on only one major release.

Newer Fords come with a USB port for firmware updates. I don't remember the system they use (MyTouch? Ford SYNC?) but I remember my brother-in-law mentioning doing a software upgrade on his F-150 before hooking up the trailer to go camping.

I'm somewhat surprised (or perhaps I shouldn't be) at the generally negative tone of the replies here so far. Maybe this specific project isn't the winner or the exact right idea, but I am _extremely_ bullish on the basic idea of getting car nav / "infotainment" systems all running a common OS (be it Android or iOS).

I think it is an absurd inefficiency and possibly one of the most anti-DRY things out there that nearly every manufacturer has its own totally-different way of doing things on the HUD. Every radio operates differently, nav systems have varying degrees of quality, and overall the UX is about 50 shades of crap :) Reminds me of the days when every phone manufacturer had a different os -- and we see how well that worked out for everyone.

I think the future of automotive UI rests on the adoption of a standard os. Though I'm partial to iOS on my own phone, I'm happy to see either iOS or Android in our cars (or both).

Imagine this: You get in your car -- the UI is familiar and comfortable (though adapted for automotive use). You have access to your music library (sync'd from iCloud) instantly, you get your choice of GPS apps (Waze, Google Maps, Apple Maps -- lol, sorry just kidding), you can use radio services like Pandora or Spotify, etc. And yes, the manufacturer will have to make a custom app for the "manual stuff" like air conditioning controls, etc., but that's still a lot better than essentially having to reinvent the whole os.

Anyway, you get the idea. I'm very excited about this stuff and am fairly confident it's the way things are going. Whether or not this specific project is a winner is somewhat irrelevant - its a step in the right direction.

Coming from my experience with my Android phone, I would not want Android controlling any mission critical function.

If its for non critical functions, thats different. There was a survey about this sort of thing done very recently.

http://www.fiercedeveloper.com/story/report-automakers-shoul...

Mission critical? That thing in the console of most cars is already anything but mission critical.

They want to replace the entertainment (including the navigation) software, nothing more. Android should be capable of doing that just fine and might bring some life into a rather dull field.. These systems have so many flaws and limitations that make them look like they're from the 90s (based on my experience with brand new Audi, BMW and VW [1] cars in the last 3-5 years). These systems easily set you back for more than 1000 EUR (usually far more) and are limited in terms of supported filesystems (like 2GB limits on SD cards), numbers of files or folders (truncating your playlist) and just .. so .. slow..

I'd love any 'mainstream' component that is integrated into the car, running on hardware that isn't far worse than my phone.

1: References: Had an A3 with ~every option~ as a company car in 2007, a VM Golf GTD with ~every option~ in 2010 and rented a lot of decently sized BMWs when travelling.

Here is the idea, please vote if you like it: Most cars already have a big LCD in the center console for navigation or entertainment, but each manufacturer writes their own software for it. My idea is to have it run Android, which would allow anyone to make apps specifically for cars, entertainment or information-based, taking advantage of special API like vehicle speed, fuel, error codes, etc (from OBD2 or ECU). This would bring huge value for car manufacturers(offsetting the development cost) and customers would love the limitless customization possibilities. Here is how it would look like: http://s13.postimage.org/8bt3a7kdj/Car_Droid.jpg

For what it's worth, a lot of manufacturers use .NET actually. Microsoft started pushing for integration into those systems back in the WINCE days, and has made significant headway. The "SYNC" based players are just the latest iteration.

Figure the car development cycle is about 3 years, and realize Andriod was just starting to exist in the minds of car designers when this year's models were being designed. It'll be a while, but I wish you the best.

Bet you dollars to jelly beans that the Google driverless cars have Android though =)

I know from a friend that a big German manufacturer uses .NET in their cars.

This is definitely already happening - I tangentially work on a number of projects embedding Android into transport, and it's really just a matter of time before manufacturers adopt Android as a matter of course.

For example, the top two manufacturers of in-flight entertainment systems are already transitioning to Android (Thales and Panasonic). There are a number of benefits using Android like this, least of all you can take advantage of the hardware expertise that already exists for building Android backed devices, and leverage the larger amount of software talent available.

Of course, there are problems. All these embedded systems have very long shelf lives, much longer than a typical handset. A IFE unit in a plane might be expected to last 15 years at least, a car is hopefully going to spend a decade or more on the road. The hardware is not going to be upgraded - don't expect updates to later Android versions.

This presents problems for third-party development, assuming manufacturers offer it. Ideally, you'd hope for a manufacturer supplied SDK that offered you easily consumable information and telemetrics (vehicle speed, fuel level, etc) and an open app store. Car companies may not see an advantage to offering it at this time.

So really, the challenge isn't to convince car manufacturers to use Android: most are almost certainly already considering it for their high-end models. Instead, it's to get a tech eco-system that's open.

> For example, the top two manufacturers of in-flight entertainment systems are already transitioning to Android (Thales and Panasonic).

Hmmm, on a flight a couple of years ago my IFE screen (Panasonic I believe) crashed, and I was treated to linux kernel boot messages and then a bunch of SDL (the hoary old linux graphics/game library) error messages as it messily failed to restart properly...

It definitely had a bandaid-and-chewing-gum feel, so using something like Android would probably be a good thing.

Any more details?

It got rejected by Nissan, not sure why...

That's strange since it's in the top 10 with votes.

Something worries me about commodifying driving distractions. Every day I see several people fiddling on their phones in the front seat, and I know I'm no less guilty. How much thought have you put in about safety criteria for these car apps?

Haven't thought about safety to be honest. Reason being cars already come with proprietary operating systems in the center console, so this wouldn't be any different safety-wise.

Sure, but most proprietary systems have only basic interactivity, which you might consider to be a feature rather than a flaw. Some even disable interaction while driving. An average driver already can't control much more than the steering wheel and pedals at once.

That said, I'd be totally down with some passive applications for a console. Maybe a navigation plugin that helps find the cheapest gas without wasting extra miles to get there, or a roadtrip meal planner that helps direct you to restaurants you like at proper eating hours.

Hopefully no Facebook though, and definitely no GTA.

Right, so maybe a moderated app market would be a good solution, with apps being placed in two categories, one category that works whether driving or parked, and the other would only work when the car is parked.

How about not trying to solve this problem by artificially crippling the software up-front and instead just leave it up to national governments to regulate this sort of thing? We don't have special software in mobile phones to make them turn off when they're in a car, we've just made it illegal to fiddle with them while driving.

The same should work just fine for certain dashboard interaction.

I have Android installed on my HP Touchpad, and it has a driving mode where it restricts you to limited features. I believe it is maps, music, phone, and maybe something else.

Not sure if this is a standard Android feature or something put in by Cyanogenmod.

Won't be much of an issue when we have self-driving cars.

If you want to know why a car running android is a bad idea, take a ride in an 8 year old BMW/Mercedes with (back then) state of the art navigation system. While the car itself is probably still nice, the "smart" part just feels horribly outdated - because it is. The length of these two innovation cycles differs by a factor 3-5. For the same reason, a "smart" fridge running Android isn't, either.

If, on the other hand, you propose introducing a reasonably standardised protocol for interfacing your phone with the in-car audio, video and input systems (plus readouts from the internal sensors), I definitively agree. This allows you to update the "smart" bit every two years on contract, but you can use the controller and screen (with longer innovation cycles, even though they are still below the typical ttl of cars) integrated in the car.

Tesla just pushed OTA updates to its in-car software yesterday.

If anyone else with more than 50 tabs open freaks out about weird bird songs and ambient noises: Disable the sound on _this_ page, which is auto-played..

Drove me crazy for 5 minutes..

For things like OBD monitoring I sort of prefer my own Android device. There are already bluetooth based dongles for reading OBD. Car manufacturers would probably lock their OS down hopelessly.

Yeah I've seen one, pretty nifty (but messy with all the wiring). The idea would be that the manufacturers don't lock it down to allow outside developers to bring value to the platform and utilize additional APIs. I made a mockup of how it would look like: http://s13.postimage.org/8bt3a7kdj/Car_Droid.jpg

It has always been a dream of mine to get desktop-like computing wherever I am; in the car, in a phone, on a bike, all sorts of craziness. The biggest drawback that I've always come up with is boot up and shut down times and when boot up and shut downs are allowed to happen. You can turn a car off and the radio/nav/etc. instantly turn off. That isn't true with computers/phones/tablets.

There are android-based radios for my car that I was considering, but the first question was how long it would take to boot up when I turned my car on. My i[Phone|pad]/nexus7 takes a good minute, id go insane waiting on that every time I turn my car on/off.

I've always wondered what are different about computer-like appliances and computers when it comes to booting them up and randomly shutting them down, even the the middle of preforming a task?

Seems to me it would be easy for the android console to go into standby while the car is off. Basically turn off all the radios and just keep ram alive so it can come up instantly when the car turns on again, using minimal battery. You know, like your laptop. ;)

Yeah but that would die out eventually, like your laptop. These newer cars with sophisticated systems probably do something like you're talking about. My macbook will write the contents of ram to the hard drive once the battery dies and load up form that image next time you boot it up, maybe that is a way to handle it in cars.

I've had a custom-built carputer in my vehicle for 5 years now. It first ran Windows Embedded, and now Windows 7. The front-end UI is a commercial application called Centrafuse(http://www.centrafuse.com/) at the moment, but there are open-source options available, namely Ride-Runner (http://guino.home.insightbb.com/roadrunner.html).

From a hardware perspective such a system is simple - I have a nano-ITX board, a 12V DC-DC power supply, and an SSD. You can package this into a double-DIN opening without issue. OBD-II connection is via Bluetooth to an adapter plugged into the diagnostic port under my dash. GPS comes from a USB receiver. Currently I use wifi to connect to my home/work network and load content, but a 3G/4G dongle is an easy mobile solution.

However, by far the best solution suitable for commercial, rather than hobbyist, use is from QNX. Check it out: http://www.qnx.com/products/qnxcar/ It's what powers most OEM systems, such as BMW's awesome iDrive system.

Don't forget to add NFC. I want to tap a location through to the onboard Nav app from maps, contact, or an appointment on my phone.

Everytime I want to go somewhere new I have to type with that damn on-screen resistive nav screen keyboard. When 80% of the time I've already got the info in a contact, meeting, or search result on my phone.

There are already android devices that fit where a car stereo does: http://www.parrot.com/usa/products/bluetooth-hands-free-car-...

The only difference would be the integration of key data points from the car itself. But then we stray from navigation and entertainment into core functionality of the vehicle. And at that point I'd want the device to be proven to still be fine for the operating parameters of a car, which requires a different level of engineering/quality than an entertainment and navigation system.

Just out of interest does anybody know what OS these car interfaces typically use? iDrive etc?

Is it Linux underneath or something totally bespoke?

Ford (don't know about others) use this: http://en.wikipedia.org/wiki/Windows_Embedded_Automotive

Update: Fiat, Ford, Kia, Nissan all listed as references: http://www.microsoft.com/windowsembedded/en-us/evaluate/wind...

My buddy created KillNanny, an app that hacks a touch screen car radio named AppRadio, so that you can run your android directly through it. His still-beta efforts can be seen here:

http://appradioforums.com/Thread-KillNanny-Video

It is very cool, even though it is still in an early state.

im starting to get really fed up with sites that require a facebook account to do anything.

"Log in to use your Facebook account with Nissan Innovation Garage."

I don't exist.

Submitted to Reddit to try and spike the contest in your favor. Good luck!

Totally agree. This would be ideal for infotainment systems. No one is arguing this should be running the ECU.

It would drain your battery before a full day is up though ;)

I've often thought a similar thing like this would work well for entertainment systems on flights too.

It would be built into the car, instead of the navigation/entertainment system and feed off the alternator, so no battery drain :)

Cute trolling. Take it elsewhere.

Sometimes it is hard to tell the difference between a troll and a joke.

Flight entertainment systems are moving to Android, but a number of factors (primarily the A380 and 787s both being severely delayed) mean this roll-out is taking a lot longer than originally planned. Both Thales (http://www.thales-ifs.com/topseries.aspx) and Panasonic (http://www.zdnet.com/blog/gadgetreviews/panasonic-testing-an...) will be moving to Android IFE systems.

The A380 has been available for a while and I've flown on them with Qantas a few times.

They use a Linux based system not an Android one.

Was probably meant to be the A350 - which is the Airbus competitor for the 787 Dreamliner:

http://en.wikipedia.org/wiki/Airbus_A350

what a bunch of bullcrap....facebook login required to vote = no vote

Pass - the partnership with Nissan is the turn off.

Which Android antivirus will it have? I'd rather not get malware on my car.

Since when Android needs antivirus?

Since it has been targeted by more malware families than Mac and (GNU/)Linux combined?

Reply on news.ycombinator.com