How Industry Software Expertise is Helping to Solve the Android Fragmentation Dilemma
It is no secret: While Google continues to push for wide adoption of smartphones driven by its open source Android platform, most industry pundits know that manufacturers typically need upwards of a full year of development and productization effort (and often Google’s help) to develop an Android phone. Not surprisingly, it takes much longer for an OEM to reach a level where they can irrigate the operator ecosystem with differentiated phones based on Android (this of course applies to most new mobile device software stacks). At the same time, we are all aware that there are multiple parallel efforts to enhance Android that have led many observers to believe that fragmentation is inevitable. Ultimately, it is said, this will become the bane to developers who will struggle to reach an economy of scale on a moving target where multiple variants of the Android OS co-exist.
That said, a handful of companies are seeking to close the gap in Android fragmentation diversity by providing the ecosystem with tradecraft skills, tooling and testing that allow for an unprecedented degree of platform validation that can act as a common denominator on the latest versions of Android for the many combinations of open source configurations, IP changes, and hardware revisions. Indeed, one of the biggest advantages of the diversity in Android has been that manufacturers are not tied to the success or failure of an individual hardware platform, meaning that today there are dozens of hardware programs in downstream Android development.
This, when overlayed with Google's official Android versions and integrated with the multiple moving parts of open source code and 2nd party/3rd party subsystems and applications, has created a "thicket" of software code and hardware that is difficult enough to manage inside of one manufacturer, much less between device makers and developers. Indeed, there is a growing recognition that companies are challenged to cost-effectively create a commercial handset product based on something that's inherently unstable or changing too rapidly, especially for mass market.
Wind River is one of the companies with the deepest credentials in the Android compatibility space, having been Google's lead Linux commercialization partner for Android starting in early 2007 and having earned a seat at the OEM and operator support table by creating the most comprehensive Android “test and fix” program in the industry. This has included deep modifications that conform with the OHA governance model as well as global support and maintenance to ensure reuse across a true Android lifecycle management program (in fact Wind River just announced its latest commercial Android platform optimized on the TI OMAP3 Platform).
Although there have emerged a long tail of professional services software companies that have announced OEM support programs for producing Android phones over the past two years, few appear to have recognized "the pain" that open source causes the manufacturer ecosystem and why a commercial platform eases that pain. They miss the opportunity to solve the business risk problem of open source by increasing predictability – many times the key issue with open source. Using a commercial platform, Wind River is offering that predictability, but without losing the accelerated innovation rate that a dynamic platform like Android is providing the market. Only time will tell which approach will prove out.