Android is enjoying an unprecedented march, and there’s no sign of this momentum slowing. Thanks to its quality, breadth, and open source Linux heritage backed by Google, Android is dominating the embedded operating system market. In a recent survey of embedded developers, 34% claimed that their devices leverage the Android OS. The next closest OS had less than 15% endorsement in the same survey.
While many perceive Android as being a mobile device operating system, it’s actually being used in a much broader range of devices, including tablets, e-readers, set top boxes, media players, cars, medical devices, and even cameras. As Android is an open platform, any vendor can enter the market and begin offering Android devices. To remain competitive, Android device makers are shrinking their release cycle times. Release times for Android devices from the time a new version of Android is released, has collapsed from ten months to under three months in just four years.
With an average Android development cycle of about three months, time to market often determines the winner in this marketplace - a few weeks delay in the release cycle can spell the difference between a hit and a dud. With newer Android operating systems being released at a rapid pace, the delivery of devices is still gated by how quickly device makers can build, test, and release their software for each Android release. To add to the complexity manufacturers face, the OS is innovating at an incredible pace, with very frequent release cycles continually introducing new functionality.
To keep ahead of competition, Android device makers need to accelerate their device delivery times to market. Three fundamental techniques that can help are:
· Accelerating Android software builds
· Speeding up Android testing cycles
· Fully integrating and automating the end-to-end software build-test-release process
Imagine if builds were completed in minutes as opposed to hours – not only would the entire cycle time be accelerated, but developers could focus on new innovation instead of spending time on operational tasks. An ideal solution would support hundreds of developers by leveraging the power of clouds as resources instead of dedicating expensive hardware for each developer.
Once the build processes finish, the next step involves running tests on the Android devices. Unfortunately, it typically takes a significant amount of time to complete these testing cycles. The process is very complex and tests like Google CTS can take upwards of seven or more hours. It’s imperative for device makers to reduce these long testing cycles to be able to increase the quality of devices shipped in a timely manner.
Lack of well defined processes commonly plague organizations building Android devices. Each in-house or third party tool must be integrated into the overall process, necessitating budget for time, effort, and resources solely focused on supporting these tools and processes.
As the market continues its massive adoption of Android, device makers will continue to offer Android on devices that can’t even be imagined at the moment. Companies servicing this growing marketplace must be prepared for mounting time-to-market, profitability, and competitive demands.