Following my post regarding the differences between Windows CE and Windows Mobile, I’d like to discuss the responsibility share among the Mobile Operator, device manufacturer and Microsoft on a Windows Mobile-based device, especially when you consider that this device has customizations from each of these companies (This article is also available in Brazilian Portuguese here).
Microsoft (yellow region)
Microsoft, as I mentioned earlier, is in charge of the core Operational System (Windows CE), the shell (user interface), a set of applications (Office Mobile, IE Mobile, WMP Mobile, etc.), RIL, and so on. We also expose several APIs (from both Windows CE as well as those specific to WM) and we include .NET Compact Framework/components to allow developers to extend the Windows Mobile platform.
This is the kit we deliver to the device manufacturer – as you probably have already figured out, at this stage we still don’t have a cell phone or a PDA. In fact, we are far from having a functional device, it is just a bunch of software. It varies a lot among device manufacturers, but I would guess this kit represents only 60 to, no more than, 80% of all software inside a WM-based device. The other 20 to 40% (or even more) are from the device manufacturer and the mobile operator.
Device Manufacturer (red region)
The Device Manufacturer (OEM/ODM) receives the kit and starts integrating the software with their hardware. Most of this integration is done through the development of devices drivers, allowing the hardware to talk to the software – different from the PC model, the hardware is much more specialized (less standardization) and OEMs spend a lot of time in this process.
Beyond drives, the OEM also develops and integrates the cellular technology into WM connected devices – we call it the radio stack. Although the OEM is in charge of developing all the communication with the cellular network, the OS comes with the Radio Interface Layer (RIL) allowing the radio stack (GSM or CDMA) to interact to Windows Mobile. From my point of view, this is one of the most super geek complex areas on developing a connected device.
I’d risk saying that at this point we have a functional device, and it is time to begin adding value to the platform. OEMs are able to add additional software developed by themselves or by their partners, like voice recognition, JVMs, home screen plug-ins, games, media players (other than WMP), MS Office viewers, backup software, etc. Beyond adding value to the end user, these applications allow them to differentiate their products from competitor ones.
We now have working device! In a non-connected word, this device would be ready to hit the retail stores , however, on the connected world, this would be considered a vanilla device and would still need to go through the Mobile Operator for further customizations, test and certification.
Mobile Operator (pink region)
The Mobile Operator (MO) receives the vanilla device, described above, from the OEM and starts two processes: customization and test/certification – these processes can be serialized or run in parallel according to the way the MO works.
These customizations generally are:
Visual: branding, colors, screen pictures, water marks, boot screens, sounds, etc.;
User Interface: Start Menu look-and-feel, menus position, IE Mobile favorites, speed dial, etc.
Applications: extra applications that the MO adds to add value to the device, differentiate it from competitors and best align the device to its current strategy/services;
Network: specific cellular network settings like: GPRS/EDEG (APN, user, password), WAP Gateway, SMSC address, proxies, PRI, etc.
After fully configured, the device is submitted to the final test & certification process. Test & certification is used to make sure the device complies to MO’s requirements on usability, network, reliability, etc, and it can be very complex on some MOs. It usually goes between 2 to 6 weeks and, in the case a major issue is found, the whole process is started from scratch (OEMs are in charge of customizing the device and supporting it during certification).
Once certified, the handset is finally ready to be commercialized!
End User (green regions)
Well, the end user is in charge of using the device :-). And, on using it, the user ends up personalizing it, changing the home screen picture, installing applications (business and consumer oriented), creating new network settings (proxies, VPNs), etc. ISVs, developers and enterprises can also develop corporate applications and deploy it to the device; hardware partners can also integrate barcode scanners, card readers, etc…
I hope you have had an idea of ecosystem’s roles and the complexity of making these devices available into the market – please, let me know your comments and questions.