The Next Battleground for Open Source Is Your Car

It all seems upside down: a major toy company releases its first tablet; a major search company works on its first car. Yet all of this makes sense when you realize everyone just wants to be – or already may be – in the mobile device business. Including car companies. But as automakers get into the computing business, the biggest hurdle they have to overcome isn’t each other...
Image may contain Transportation Vehicle Automobile Car and Driving
gorbould/ Flickr

It all seems upside down: a major toy company releases its first tablet; a major search company works on its first car. Yet all of this makes sense when you realize everyone just wants to be – or may already be – in the mobile device business. Including car companies.

A friend recently showed me his shiny new luxury sports car. Did he rave about the 333-horsepower, six-cylinder engine, or 14-speaker, noise-cancelling stereo system? No. His first point of pride was the car’s ability to become an internet hotspot, powering Wi-Fi devices throughout the vehicle. This makes sense when you realize cars have become our portable offices and homes, a shared mobile experience for the entire family.

#### Jim Zemlin

##### About

Jim Zemlin is executive director of [The Linux Foundation](http://www.linuxfoundation.org/), a non-profit organization and the home of Linus Torvalds. Zemlin works with the world’s largest technology companies and others to help define the future of computing. He has worked at Western Wireless (later acquired by Deutsche Telekom and renamed T-Mobile USA), Corio, and Covalent Technologies, and was executive director at Free Standards Group.

In this brave new world, connectivity is king. The engine is almost an afterthought.

The big question, however, is: Couldn’t my friend have achieved the same result with just an iPad and Velcro? He paid thousands extra for the “in-vehicle infotainment” (IVI) system: two screens in the back to stream content, one up front for navigation, Pandora, etc. Custom navigation? Great, but it will quickly get out of date since there’s no way to update software over his hotspot connection. Pandora? Nice, but I prefer Mog – and unlike with mobile devices, I can’t choose my in-car apps: It’s not a computer (yet). And who knows what options I’ll prefer as the market and technology changes.

I’ll definitely have a new Android device in one year, but I prefer to keep my cars for 10 years.

As automakers get into the computing business, the biggest hurdle they have to overcome isn’t each other – it’s consumer expectations driven by the rise of ubiquitous mobile computing. This is where I’d argue the battle between open and closed is going to play out the hardest in coming years ... the next OS wars.

When Automakers Become Software Makers

A luxury automaker recently told me its IVI system contains about 1,900 use cases – “of which we only consider about 3 percent unique to our products; the other 97 percent are common across all car companies.” Let me emphasize that: THREE percent. Can these companies really afford to pour a lot of time and money into such a small amount of differentiation?

I’m not sure how many research and development (R&D) resources the car company spent on my friend’s car software, but I bet it was a lot. And it was redundant, because the result could never compete with today’s Apple, Android, or Amazon app-and-content ecosystems: the new "AAA" of the automotive landscape.

Yet ... the iPad-on-Velcro option can’t deliver the fully integrated experience expected for future car systems, either. Because cars aren’t just mobile devices, they’re mobile sensor networks.

>The result could never compete with today’s Apple, Android, or Amazon app-and-content ecosystems: The new "AAA" of the automotive landscape.

My friend’s car probably has about 100 electronic control modules (ECMs) for things like engine control, heating, air conditioning, lighting, sound, braking, and more. For each ECM, there are multiple sensors that capture everything from sunlight, rain, and temperature to outside noise, proximity of nearby cars and objects, oil levels, and tire pressure. A car with a fully integrated, connected IVI system can take advantage of this sensor data.

Picture this: Your car’s systems will be able to sense how much fuel is left, compute your average consumption, and combine this information with your GPS navigation data. The result? Your car will tell you which gas stations you should stop at before you run out of gas, and since it’s connected to the internet, which one has the lowest price. It may even drive you there.

An iPad-on-Velcro or even better smartphone in-car connectivity can’t compete with that. It simply requires too many separate apps and inputs and hands manually connecting the dots.

So the R&D investment matters. It’s the only way car companies can deliver on the promise of a fully integrated experience. It’s the only way they can overcome customer expectations primed by the consumer electronics industry. And it’s the only way they can differentiate themselves against each other, even if it’s just for that 3 percent.

When a Closed Industry Goes Open

But here’s the paradox: The automotive industry is going to have to collaborate in order to differentiate.

Is this as strange as cats and dogs living together? Not really. IBM and HP did it with their server businesses. That’s how open source works: Competitors collaborate on the code and requirements to produce a common base, upon which they differentiate and compete with each other.

For the automakers, it’s a way to ride the shoulders of computing industry giants like Google, Intel, IBM, and many others who have already invested $10 billion in open source. These companies made open source suitable to run the connectivity and infotainment needs of a car, and the automakers can close the gaps by adding functionality unique to cars.

>Is this as strange as cats and dogs living together? Not really. That’s how open source works.

Last month the Linux Foundation launched an “Automotive Grade” Linux workgroup founded by the one of the world’s largest automakers (Toyota), chipmakers (Intel), and mobile device makers (Samsung), along with Jaguar/Land Rover, Nissan, Harman, NEC, Nvidia, and Denso.

I think the fact that Toyota is now going outside its ecosystem by joining with competitors is a significant move toward openness. We know they already collaborate with a complex network of suppliers in one of the most sophisticated just-in-time production systems in the world: the Toyota Production System (which, by the way, inspired lean startups). I'd argue that examples like this show that the automotive industry does know how to standardize and collaborate. The question, of course, is how the automakers will balance their need to collaborate with their desire to provide a branded infotainment experience to consumers.

The telecommunications industry, a traditionally (and many would argue still) closed industry took a similar workgroup approach in the early 2000s, eventually displacing the proprietary systems of old: Linux is now the dominant operating system in the telco market. This industry, much like the automotive industry, requires 99.9999 percent reliability in real-time software. It's not like you need to be closed to get great performance, no matter what Steve Jobs would say. (And when pondering WWSJD when it comes to being open, please don’t keep saying Apple’s the exception: There’s a lot more open source there than you realize.)

Yet getting traditionally closed industries into a more open mindset will take time. For example, what if automakers try to prevent customers from modifying source code?

People have been modding cars for years (horsepower, turbo capability, hydraulics), and law enforcement reacted with “street legal” definitions and enforcement. How will we be able to hack cars when hacking the software could lead to unintended, disastrous consequences? Do we need “hackable” distribution for our cars, much like CyanogenMod provides the Android-modding community?

And what happens if malicious hackers insert malware into automobile source code? The prospect of security becomes all too critical when we’re strapped behind a four-wheeled killing machine that someone else can control – even if that someone is the car itself.

With greater connectivity comes greater vulnerability. Right now, a car is an island. But that also means it’s a lonely place.

Open source can help connect those islands. Think about it: automakers, parts suppliers, developers, car tuners, and consumers working together to address security, define new specs, and build the ultimate mobile device ... whatever form it takes.

Disclosure: Jim Zemlin is the CEO of the Linux Foundation, which hosts the Linux Kernel project and the Automotive Grade Linux, Carrier Grade Linux, and other open source projects and initiatives.

Editor: Sonal Chokshi @smc90