Google Android: Beyond the mobile phone

By Stephen Shankland on 23 September 2008, 10:59am

The first phone using Google's Android operating system will debut today, a model from T-Mobile, and more are set come. But some Android partners say the software will use more broadly than just phones.

"We're starting to see Android get designed in on devices that extend way beyond the phone -- things that might go in the automobile or things that might go in the home," said John Bruggeman, chief marketing officer at Wind River Systems, a Google ally that helps phone makers build and customise Android for their phone hardware.

It's not clear yet whether Google shares this broader Android ambition -- the emphasis today is for mobile phones -- but extending into new areas could increase both the prominence and competitive threat of the project. Projects that spread wider also can be stretched thinner, however, and advantages such as broader developer interest could be offset by incompatibilities and other drawbacks.

Bruggeman declined to share specifics about which Internet-connected devices might employ the operating system, but he did mention TVs and set-top boxes as well as cars. And he was confident some will arrive next year.

"I don't want to pre-announce any design wins," he said. "I think you'll see them in 2009. I would be shocked if you didn't."

Google didn't immediately respond to a request for comment.

Of course, Android is mostly open-source software, so there's nothing stopping people from doing anything they want with it. But Wind River is a notable member of the 34-company Open Handset Alliance that Google gathered to build, support, and use Android.

Wind River has years of experience with so-called embedded operating systems, starting with its own VxWorks and eventually extending to include Linux, which underlies Android. It also has plenty of customers, and to beef up its Android support services, Wind River acquired mobile Linux firm Mizi Research in August for a price it said could reach $16m.

The Android business "was significant enough for us that we acquired a company so we had additional resources," Bruggeman said. Mizi is based in Korea, as are LG and Samsung, two notable phone makers in the Open Handset Alliance.

Much depends on how Google sees the effort. It has massive engineering resources, of course, but perhaps more important, it has a powerful brand, some pull with the programmers it's enlisting to write Android applications, and a strong will to spread Internet access far and wide.

But embedded computing is a tough nut to crack. Wind River, along with MontaVista Software and many others, have tried to spread Linux to embedded devices. And while they've had significant success, there's much fragmentation, with nothing as universal as Windows or as standardised as the iPhone.

Google has come up with a prominent brand and a strong developer programme for Android, which brings compatibility issues to the fore. With a brand comes an implicit promise that everything sporting the brand works well together. The broader the Android brand spreads, the more complicated it gets. For example, what if a programmer wants to take advantage of the considerable computing horsepower in an Intel mobile Internet device for a game -- would that work on a comparatively feeble feature phone with a smaller screen, no keyboard, a low-capacity battery and inferior graphics?

One convenient element Android brings to the compatibility challenge is that software doesn't run on the Linux component of the operating system. Instead, it runs on a Java layer from Google called Dalvik. That means programmers writing applications for Android need not concern themselves with the underlying hardware, such as whether a device is running an ARM-based processor or an x86.

Here again, though, there are some compatibility issues. Sun Microsystems' Java, already used widely in mobile phones, is a slightly different foundation. Software may transfer more easily from one domain to another, but there won't be any guarantees of compatibility.

One of the big elements of the Android sales pitch is openness, though, and that could have appeal in other markets.

Perhaps a developer might want to sell an application that shows photos that have been geotagged nearby on Flickr, on an in-dash navigation device without having to obtain General Motors' permission first. Perhaps a user might want to download a TV game from Google's Android Market without requiring clearance from Sony.

And of course, people might want to use Google to perform an online search. Which is why Yahoo, Microsoft, and Nokia probably shouldn't be too complacent about the possibility of Android spreading beyond phones. -Stephen Shankland