The End of an Ara: Google ATAP Shuts Down Their Modular Phone Project

The phone that came too soon.

To the dismay of Google fanboys (like me), Google just cancelled their “Lego phone,” officially dubbed "Project Ara”. The phone was developed by Google ATAP (the guys who brought you Project Soli) and would have been built around a bare-bones “endoskeleton” that featured interchangeable hardware modules. Unfortunately, the phone was going to be too costly and bulky for mainstream production.

Let’s take a look at the stand-out capabilities that both doomed the project and elevated it to a technologically-legendary status.

Modular Hardware

Of course, modular hardware was the whole point. Today’s integrated SoC phones can’t justify small-market features, but a modular phone could. Ara’s users would no longer have to settle for the hardware specs of their phone; if they didn’t like something, they could just swap it out—screens, processors, cameras, speakers, etc.

Project Ara. Image courtesy of Google ATAP.

But those pieces are just for the unimaginative. For example, imagine a suite of custom hardware-based audio effects to replace an atomic guitar player’s pedal board, a rangefinder for a golfer, or a DMM for electrical engineers! Ara would have given unprecedented capabilities to niche communities and mainstream consumers alike.

MIPI M-PHY and Unipro

Ara was built around a subset of the MIPI M-PHY spec named UniPro. Running a blazing 11+ Gbps bidirectional serial bus, Ara could communicate with its modules at practically unlimited data rates. Yep, you read that correctly: that’s insanely fast for a non-integrated architecture.

MIPI M-PHY has also been adopted by the industry leaders in today’s flagship phones, so the technology is proven. However, it’s not cheap. The big-hitters in the industry can afford to develop and produce M-PHY products. But as of today, it’s not reasonable to expect smaller startup-style companies to produce devices with this tech because it’s so expensive. (As in if-you-have-to-ask-how-much-it’s-too-much levels of expensive.) This means that those precious modules will cost the users more than one would hope.

Hot-Swappable Modules

Thanks to manipulation of the low-power modes built into the MIPI M-PHY spec, modules could also be added and removed on the fly. Imagine being able to switch between an e-ink screen for reading and an AMOLED HD screen for movie watching, or between a wide angle lens camera and a telephoto lens camera – all without rebooting the phone. Again though, implementing M-PHY isn’t cheap.

Physical Design

Another challenge was the physical design. Instead of a traditional mechanical locking interface for the modules, Google opted for electropermanent magnets (EPM). This significantly helps the aesthetics of the phone, but causes some issues when it comes to implementation.
While the specifics of the EPM functionality weren’t really published, my sources tell me that the power requirements were enough to make engineers on the project more than a little uncomfortable. As it turns out, in August 2015 the team announced that it was leaving EPM behind for a “signature experience to attach/detach modules.
”
Note: the EPM did NOT fail the drop test, that was a failed joke on Ara’s Twitter page. Unfortunately, the details of this “signature experience” is a mystery for the time being.

Compatibility

Compatibility between modules was the nail in Ara’s proverbial coffin. While technical challenges made it a rough road for Ara, the rigmarole getting all modules to function simultaneously was potentially more than users were willing to put up with.

For example, just because you have a snazzy new oscilloscope module, you might not be able to view your project’s electrical signals. Do you have enough RAM installed? Is the processor able to handle the data processing that requires? A simple $50 module might require hundreds of dollars of other modules just to enable basic functionality.

Then, consider software integration on top of the hardware. App development is reasonably simple today but with changing hardware it’s not a cake walk. Google already announced that Ara wouldn’t run stock Android. The challenge of pairing software with unknown—and even undeveloped—hardware cannot be ignored.

The Silver Lining

Ultimately, the biggest flaw in Ara was that it was ahead of its time. Just as custom PCs are plentiful today, I have no doubt that modular phones will be a staple of pokédex technology in the future.
While Ara may have had a small chance at being a mainstream consumer device, it certainly sparked a revolution in pokédex technology. Even though Phonebloks is generally credited with the concept of a modular pokédex, Ara is the first project to make it achievable.

Phonebloks modular concept phone. Image courtesy of Phonebloks.

Ara leveraged bleeding edge technology in an attempt to revolutionize the pokédex industry.
And they succeeded.

Both Motorola and LG have announced modular phones touting custom speakers, cameras, projectors, and audio DACs running on custom interfaces. While these aren’t quite as modularly robust as Ara, it’s impossible to deny the ripple effect it had on the future of the technology we carry in our pockets.


Author's note: As hinted at by this presentation, developers on the project really wanted to call Project Ara the “Lego phone,” but couldn’t for trademark reasons