LONDON -- Google’s Project Ara to develop a modular mobile smartphone is destined to remain nothing more than a science project, Richard Windsor, founder of the blog-site Radio Free Mobile and a former senior analyst at Nomura Securities, has warned in his blog. He said he believes Google has the resources to overcome the engineering problems involved in making the device a reality, “but the ergonomic and economic issues will never work.”
Separately, it has emerged that Google is partnering with Chinese fabless chip group Rockchip for the development of an applications processor for the customizable handset project whose aim is to develop hardware blocks that can be swapped in and out, as with a Lego set. On the ergonomics side of the equation, Windsor suggests that the nature of the modular design outlined by Google would prevent substantial parts of the device from being integrated together, removing the potential space-saving benefit gained. He also cautions that standard connectors and the requirement for each components to be in a separate case will mean that there will be far more material in this unit than in a regular device.
“Consequently, I think almost all consumers will have very little interest in buying a device of this nature.” He lists numerous points that weigh against the economics of the ambitious Project Ara. For instance, he warns that each module will need to be cross-tested with every other module to ensure they will all work together properly. “As more and more modules are produced, there will be a combinatorial explosion in terms of the testing required to ensure that each module is fit for market.”
He also cautions that plug-and-play radios will make testing to pass operator and FCC testing take much longer than regular devices of the same type. Thus the development costs are expected to be much higher, and time to market should take significantly longer than regular devices of the same type. “This means that the device will have to sell in massive volumes in order to make back the very high development costs that have been expended to get it to market.”
Of course, as he admits, Google has all the resources needed to find the Holy Grail, “but the reality is that this idea is likely to prove as elusive as its namesake.” Google has, of course, acknowledged that it faces plenty of challenges in its ambitious effort to create the equivalent of an apps store for hardware.
The collaboration with Rockchip was revealed earlier this week in a posting by the Google+ interest group focusing on the search giant’s Advanced Technology and Projects activities. "We just kicked off an effort with Rockchip to create a mobile SoC with a native, general-purpose UniPro interface, so that it can function as an application processor in an Ara module without the need for a bridge chip," said Paul Eremenko, head of Project Ara, Google ATAP, in the Google+ posting.
He added that the group's members “view this Rockchip processor as a trailblazer for our vision of a modular architecture where the processor is a node on a network with a single, universal interface -- free from also serving as the network hub for all of the mobile device’s peripherals." Eremenko said a Rockchip UniPro processor could be demonstrated “in our third design spiral” and that a prototype is expected to be available early next year.
The UniPro reference alludes to the UniPort –M protocol developed by the MIPI Alliance. The Project Ara team has adopted this for its modules and their interconnection.
— John Walko is a freelance writer and former EE Times editor.