The use of Android in embedded devices is heating up and along with that comes demand for developers skilled in embedded Android, say analysts and service providers within the embedded industry.

The impending demand isn’t immediately obvious; beyond smartphones and tablets, few products running embedded Android can be found on the market today. The automotive industry, where Android has been a staple of in-vehicle-infotainment systems, has so far seen the most prolific use. But that’s about to change.

Companies that have been experimenting with embedded Android for the past two to three years will finally soon begin releasing their products, said Jared Weiner, a senior analyst at VDC Research, which has done an annual survey of the embedded market since 2003. Android build systems have also greatly improved in that time period, making it easier and thus more likely for engineers to make the switch. And vendors have become more skilled at engineering and porting embedded Android services, he said.

“We might be close to the tipping point,” Weiner said. “It certainly seems like things are primed for the dominos to fall quickly.”

OEM-Approved and Ready

The Linux-based Android OS was designed specifically for phones and tablets but companies have increasingly been using, or considering using it, in other embedded products in recent years. Because of its custom design, it makes the most sense as an embedded operating system in connected devices that have a user interface, and most often, a touch screen.

Some 16 percent of embedded engineers surveyed said they are currently using Android as their operating system in UBM Tech’s 2013 Embedded Market Study. While 28 percent said they plan to use it in the next 12 months. By comparison, 27 percent were using embedded Linux and 20 percent were likely to use it in the next 12 months.

“We’re just at the start of Android in the embedded space,” said Jay Lyman, a senior analyst at 451 Research. “We may see a real fast growth of Android once it begins making some inroads. Android will be key where there’s a user interface and we’ll see more user interface technology with the ‘Internet of Things.”

Over the past 18 months, Wind River has seen strong growth in demand for embedded Android among its customers in the automotive, aerospace and defense, medical and industrial markets as well as in consumer devices, said Darshan Patel, director of open source platforms at Wind River. Companies that were once merely exploring the option through proof-of-concept designs are now launching real Android-based initiatives, he said.

“These areas that typically have more safety and security-focused devices are now feeling comfortable with open source software in general,” Patel said. “Linux has led the way and Android is piggybacking off of that.”

It’s a relatively hidden trend right now because these industries have much longer product cycles than the typical consumer device. But when they start rolling out, it will come quickly, he said.

“It’s just starting to change. It takes quite a bit of time to overcome inertia and what companies are using,” Lyman agreed. “It’s important that Android has earned trust and support from large OEMs, and not only handset makers, but their component suppliers.”

Embedded Android Training Recommended

Along with Android OS in all manner of devices — from rugged handhelds to kiosks, treadmills and cars –comes the demand for skilled programmers to develop them.

“Embedded Android is another one of these places were there will be great demand for experience and expertise and that goes all across the Linux operating system because we do see it being used in quite an array of things,” Lyman said.

Embedded software and services companies such as Wind River, Enea and MontaVista have already begun gearing up for the change, developing Android-based platforms for their customers. And at least one, Wind River, employs engineers who are entirely devoted to embedded Android.

Because of its existing network within the embedded industry, Wind River is able to find developers with embedded Android experience, but it’s difficult, Patel said. So they look not only for trained engineers, but those willing and able to learn quickly.

The opportunity is ripe for developers who may know embedded Linux, for example, but haven’t done much yet with Android.

“(Embedded Android) is definitely something that can’t be ignored and should be explored by engineers,” Weiner said.

“From the survey data we’ve collected and interviews we’ve done, everything points to Android being a big presence going forward,” he said. “Getting engineers trained up to repurpose or use Android in a project that isn’t going to be a smartphone or a tablet is definitely a piece of the puzzle.

Editor’s Note: In recognition of the latest embedded industry trends, The Linux Foundation is offering a new embedded training course, “Inside Android: An Introduction to Android Internals.”  The next course will be held Nov. 18 – 22, 2013 in the Silicon Valley. 

Since the National Security Agency’s PRISM surveillence program was leaked to the press in June, the public and corporate backlash has some analysts estimating billions of dollars in losses for the IT services industry. In this context, developing an open source alternative to commercial cloud platforms becomes even more important, argues Brian Aker, a fellow in the HP cloud services division. 

In projects such as OpenStack, contributors have the opportunity to build cloud software that’s not only transparent, but also highly customizable “so that there is no option to recover user data, for example,” Aker says.

OpenStack, which is HP’s cloud platform of choice, is still a work in progress, however, and has many technical challenges as well as opportunities. In his keynote talk at LinuxCon and CloudOpen in New Orleans Sept. 16-18, Aker will lay out the project’s specific successes and challenges. 

Here, he addresses open cloud skeptics; the role of Linux in the open cloud; HP’s commitment to OpenStack and involvement in the project; and how PRISM has affected enterprise cloud strategy. 

Why does an open cloud matter and are there still people who need convincing?

Brian Aker: If we go back in Linux or open source history, people didn’t necessarily have a vision beyond providing tools. But some folks didn’t buy that; they wanted a fully fledged operating system. The vendors tried to say you don’t need to do that, and yes we’ll charge you thousands of dollars for a feed compiler — tools that nowadays people wouldn’t pay for. Then we had FreeBSD and Linux come out and yet the vendors continued to say you don’t really need that. They were dismissive and hostile and then they went out of business (or changed their strategies to adopt Linux).

When we look at cloud right now it’s interesting to read some of the apologists who say we just need open source tools, not an operating system. They say, “Amazon is great, just use their API. Why do we need to do everything different?” It’s the same problem decades later. The cloud is like a distributed operating system that requires many different systems and if we don’t wish to all be sharecroppers again, it needs to be open source.

What is the role of Linux in the open cloud?

Aker: Linux is one of the most integral pieces to the whole thing. It’s the base OS that is the one of the main functioning building blocks of this whole thing. The bigger picture is that there‘s a new stack, and LAMP is a component of what is a much larger architecture.

What is the biggest challenge facing the open cloud today?

Aker: There’s a lot of support for OpenStack but we have to engage in it to turn out a high quality implementation that can compete with what Amazon has today. Historically there was MINIX. Linux went a step further to turn it into a system you could rely on. You saw companies that are very commercial step in to help evolve it in that direction. It’s the same with OpenStack. There are missing components — pieces that still need to be written. It’s a wide-open opportunity.

Is the post-PRISM debate over security making enterprises rethink the open cloud?

Aker: This is where open cloud becomes even more important. It gets people thinking about whether they want to do a public or a private cloud. And it drives home that they want some transparency in the software that’s being used.

If you look at the big cloud competitors, they’re everywhere from proprietary to extremely proprietary. That’s concerning. PRISM is the most proprietary cloud built, there’s no transparency about what’s happening there.

With the open source cloud we can bake solutions into this that are extremely useful long term, so that there is no option to recover user data, for example.

What is the role of Linux and open source in HP’s cloud strategy?

Aker: OpenStack is one of HP’s strategic pillars moving forward. It’s the same thing that occurred a decade ago when the company decided to point toward Linux. I’ve been at the company for not quite 2 years and it’s fascinating to watch an entire company move in a set direction. You can see the OpenStack question is in almost everything HP does right now. There’s no second system. This is the road map.

How is HP advancing OpenStack going forward?

Aker: Look at stuff we’ve done in Trove or Ironic. We contribute across the board; it’s not just a single effort. We are touching pretty much every bit of the code and have people working in all aspects of it at this point.

And we are a primary supporter of CI (continuous integration), the basic idea that you don’t commit code to trunk versions of software that hasn’t been run through testing before it gets there. It shouldn’t be innovative but it still is. There are many open source projects at this point that have yet to embrace it. Not in OpenStack, though. It’s a pure CI project.