In November I posted an article called Is Apple ruining mobile virtualization for us or saving us from it? Shortly afterwards, I was contacted by Red Bend Software, another company that’s hoping that mobile virtualization will be the answer to BYOD and dual-persona devices. Will this happen? Let’s take a look at their specific solution, and also see if we can answer any more questions about mobile virtualization in general.
Red Bend was founded in Israel in 1999 to focus on doing remote software updates. They started with PC software over dial up, and then in 2003 they began doing over-the-air firmware updates (known as FOTA) for mobile phones. All that evolved to the point where they now support 1200 different device models (three quarters of them are phones with the rest being things like medical devices and utility meters) and over a billion devices use their technology for FOTA, device management, analytics, and similar tasks. In 2010 they acquired VirtualLogix, one of three original companies doing mobile virtualization—the others being Open Kernel Labs (recently acquired by General Dynamics and invested in by Citrix) and Trango (acquired by VMware in 2009 and now the basis for Horizon Mobile). Red Bend’s virtualization product was released as vLogix in February 2012, and up until now (similar OK Labs’ product), has only been marketed to carriers and handset OEMs for management and hardware abstraction purposes.
This means that so far all the things they do are at a level that's not visible to users, MDM vendors, or IT departments. Like the other mobile virtualization companies, Red Bend realized there’s another use case for their product (the one that we care about): creating dual-persona devices with separate spaces for work and personal data and apps.
Red Bend’s dual-persona product will be a Type-1 hypervisor for Android. There won’t be much communication going on between VMs, and when I commented that this sounds like it’s as close as possible to being two separate phones in one physical device, they said “Exactly!” While most of the other dual-persona mobile virtualization platforms can allow IT to control cutting and pasting or opening documents between environments, that type of communication isn’t allowed with Red Bend. There's also a third management VM that controls incoming messages and cross-VM notifications. It can route calls from a single phone number to either the work or personal VM based on context.
It will only be possible to create corporate VMs through the management APIs, not the user interface. To provision to a device, IT needs some identifying information from the user, but they don’t get to see inside the personal VM. It’s too bad users can’t make VMs on their own, because if they could, tech-savvy employees at companies with strict BYOD management policies could just let IT access a VM instead of their entire phone. I know I would absolutely do this if I were in the situation, which is one of the ways mobile virtualization can be really awesome. This use case is another small niche for sure, but one that’s not planned for Red Bend.
Keep in mind that carriers and OEMs can change the way that the features of Red Bend work to fit their specific needs. Like the other dual-persona mobile virtualization products, Red Bend uses a highly-customised version of Android, so bringing it to market requires licensing it to a handset OEM, though they do have solid relationships here thanks to their existing products.
A few more thoughts on mobile virtualization
I’ve hashed through the pros and cons of mobile virtualization before, (read this article for all of the arguments on both sides) but there are a few more factors that come into play when we add carriers and OEMs to the conversation
For carriers, mobile virtualization for BYOD represents new revenue opportunities. The days of a carrier having an advantage by exclusively offering certain phones are mostly behind us, as the iPhone and popular Samsung phones are pretty much available everywhere. While many people might be content to see the carriers be completely commoditized, obviously they don’t want that. A phone like this could be a good source of new revenue, especially if any one of them actually introduce the much-feared “pay double the amount for a single phone” plans (which thankfully hasn’t actually happened yet.)
What about the OEMs? On one hand, the leap from virtualization for hardware abstraction to virtualization for separating work and personal isn’t inconceivable. On the other hand, Samsung, for example, already has a set of special management features that they're marketing to the enterprise.
As for IT, mobile virtualization sounds cool but it's limited because IT has to support more than the few phones that use mobile virtualization. It's not like they can say, “Great, we’re going to support these two phone models from one carrier. Enjoy your consumerization!” IT has to find a way support everything, including iPhones, all versions of Android, and cheap tablets from Walgreens.
Red Bend compared to Open Kernel Labs, VMware, and Cellrox
With their existing manufacturer relationships, Red Bend seems to have as good of a chance as anyone else for actually bringing a product to market. Their position is pretty similar to that of Open Kernel Labs—both companies already market virtualization to carriers and handset OEMs as a way to make lower-level management and development easier.
The situation is a bit different for Cellrox and VMware. VMware seems to be the farthest along with actual devices shipping soon in Japan, but overall 2012 was a disappointing year for them and mobile. Cellrox is a startup completely dedicated to mobile virtualization, and they have what I think is the most compelling product (it has the capability for both users and IT to create multiple, granular virtual environments).
Mobile virtualization is a compelling idea, and it certainly has its advantages and use cases, but I’m still skeptical that it will be anything more than a niche. And as we’ve seen, it takes a long time to come to market, and even once that happens IT will still have to deal with all the other Android devices out there (and not to mention iPhones, too!)