Word comes this week that Google has given up its plans to build modular smartphones with interchangeable parts, having cancelled all ongoing Project Ara efforts. According to reports by Reuters and Re/code.it appears that the price of the final handsets was going to be so high as to make viability questionable, as the modular phones lost some of the cost and performance advantages of hardware integration. The company reportedly plans to concentrate on other hardware projects, including its Chromebooks and various Android-based devices.

Google began to explore the concept of modular smartphones in 2012, and publicly announced its Project Ara in late October, 2013. The company thought that for many people it would make sense to configure their smartphones themselves and then upgrade modules, as new ones come out, instead of getting entirely new handsets.

Initially, Google considered building a fully modular smartphone with a PC-like architecture in a bid to enable upgrades of core components like SoC, antenna, sensors and so on. However, this required a lot of efforts in hardware standardization, interconnection, compatibility as well as software support. Back in May, the company announced changes of the Project Ara concept. Google said that core components of modular smartphones would not be upgradeable, but users would still be able to switch camera sensors, speakers and even add secondary displays. Google promised to introduce a working Project Ara model this fall and then release a commercial product for consumers in 2017.

Modern smartphones are self-contained, cannot be upgraded and their repair is often tricky at best. However, such integration allows manufacturers to make them sleeker and cheaper. By contrast, Google’s modular design appeared to be rather bulky. Moreover, Re/code reports that Google struggled to come out with a modular smartphone that could perform up to expectations and come in at a viable cost. The price of modules themselves was also a potential concern, as they'd need to be built to handle the modular system and wouldn't necessarily enjoy the high volume sales of a solitary phone design.

As it turns out, Rick Osterloh, the new senior vice president of hardware at Google, decided to cancel Project Ara even in its “limited” form announced back in May. The company will no longer invest in the project, but may license technologies and patents it developed over the past several years to parties interested in building their upgradeable smartphones, according to reports. Keep in mind though that Google yet has to confirm its intentions regarding Project Ara.

Recently Motorola came up with the Moto Mods idea to sell add-on accessories for its Moto Z smartphones. The add-ons can transform the handset into a camera with decent optics, a projector or a stereo system. Apparently, even though Project Ara is gone for good, the concept of add-ons for smartphones lives on.

Sources: Reuters, Re/code, The Verge.

Comments Locked


View All Comments

  • zepi - Friday, September 2, 2016 - link

    Anyone with any electrical engineering competence knew from the get-go that this was doomed.

    Regardless, I'm pretty sure they figured out the smartphone HW business via cheaper way than MS. $7.2 billion so wasted...
  • ddriver - Friday, September 2, 2016 - link

    It was silly from the get go, and moronic from engineering standpoint. You don't do that with magnetic blocks. You need standardized sizes for SOC "motherboards", batteries, displays and enclosures, so that you can actually put something together. Tried and proven concept, all that was needed was a new mobile device form factor for it. But noooo, let's use magnetic blocks, because stupidity!
  • boozed - Friday, September 2, 2016 - link

    I'm not sure the choice of magnetic connectors is the only (or even the major) problem with this idea...
  • ddriver - Friday, September 2, 2016 - link

    So then what would be the major problem in your opinion?

    Also note I did not blame "magnetic connectors" but the "magnetic blocks" as a whole. Were they trying to make a device, or a children's toy?

    "this required a lot of efforts in hardware standardization, interconnection, compatibility as well as software support"

    BS, more like it only required a form factor standard, which itself incorporates connectors. All the rest apply equally as much to their silly magnetic block concept. However, with actual modular hardware you could have some choice in size and design rather than being stuck with this rectangle with more rectangles on it. And that whole "lot of efforts" would actually have taken a few days tops.

    But as I already noted in a comment below, from the looks of it that project was simply not designed to succeed, but to show a "noble effort" while at the same time convince people they are better off with planned obsolescence devices.
  • boozed - Saturday, September 3, 2016 - link

    I had written up a reply with a lot of my thoughts before I read your post below and realised that you'd already covered most of my own thoughts on the matter.

    From my perspective the problem is that a given flagship smartphone costs what it does today because 10 million identical ones are made. All of the hardware and software development is undertaken once and the development costs are amortised over those 10 million devices. Then there is the huge economy of scale when you're ordering parts from suppliers and manufacturing that many of something. In the electronics industry, these discounts can be on the order of low to medium double digit percentages, each time you increase the size of the order by 10x.

    A modular smartphone necessarily has connectors that a conventional smartphone doesn't. These introduce additional complexity, size, mass and points of electrical and mechanical failure. It will not be possible to make a modular smartphone down to the same performance, cost, reliability, size and mass as a conventional smartphone. You might hit two of those targets but only by (severely, IMO) compromising the other three, and at that point, I can't see many people being bothered buying one.

    I had also written about planned obsolescence, to the effect that you can bet your arse that it will still be a factor! These companies are in the business of continuing to sell you things, after all, and I'm not sure that any conspiracy theories are required there. You'll probably see software compatibility issues between old and new modules, new modules such as CPUs or screens that require more bandwidth or power and thus a new connector, and all manner of other things that are already seen in every other modular system.
  • ddriver - Saturday, September 3, 2016 - link


    "probably see software compatibility issues between old and new modules"

    If a 10 year old video card can work on a brand new motherboard I don't see any reason why a future-proof connector can't be created. There is an easy way to do this - allocate the first two pins to initiate a good old serial connection to negotiate the actual connection protocol. Make a module with 40 "pins" and a basic switch fabric, and modules can negotiate pin allocation as needed - pcie, usb, lan, video, whatever. Naturally, even if this is a possibility and not too complex or costly either, you could also get away with something a lot simpler.

    "connectors that a conventional smartphone doesn't"

    I don't know if you have every taken a phone apart, or at least watched a teardown, but phones have plenty of connectors. And they are not the "sturdy and easy to use" ones either, they are frail things which often break. Conventional phones are VERY MUCH MODULAR, however, the modules are not designed to be interchangeable and are put together in the utmost modification and repair unfriendly ways.

    "costs what it does today because 10 million identical ones are made"

    Interchangeable standardized modules would make for a MUCH LARGER scale production, as the parts won't be limited to one specific brand and model. It would actually bring down production and design costs significantly. The same module can remain useful for years even in its original form, and furthermore there is the possibility to upgrade the chip only, as many SOCs have proven that it is very much possible to release pin-compatible newer and better models.

    "not be possible to make a modular smartphone down to the same performance, cost, reliability, size and mass"

    Performance would not be a problem if SOC makers sold their chips freely and without years of delay, for the cost I agree - it will cost a little more, however, as mentioned in the previous post, large scale production will bring production costs lower, cost size and weight would be a little higher, but nothing too severe, a realistic number is 5% for all the three metric. As for reliability - it could be just as good with the right design, plus the option to easily repair and upgrade. And 5% is a very small price to pay, and will be entirely negated by the possibility to easily upgrade and repair the device.

    "I had also written about planned obsolescence, to the effect that you can bet your arse that it will still be a factor"

    There are two types of planned obsolescence, when a product was designed to be useless after a short time, which IMO is BS, because given a more efficient software infrastructure, a phone can remain perfectly useful for over a decade easy - after all it is a phone, and the second type - which is what is the real problem - when a product was designed to fail after its warranty period is over. That kind of planned obsolescence is usually the product of intentionally bad design, and is something every maker does, more or less. For some makers it is actually quite a lot - such as apple, their engineers put a lot of effort to create weak points that will in time lead to a device failure, and it is being devised not to justify "certified service repair" but a brand new purchase that would otherwise be needless. This kind of behavior would not be a problem with standardized modularity, as the standard itself will enforce quality and durability, and people will have a choice and simply NOT BUY modules which don't stand up to the task.

    As you see, you are about 95% wrong and have very little valid point. Which also explains why you bring in "conspiracy theories", as apparently you are one of those people who is being influenced by the "its a conspiracy theory" approach used by many to "refute" their own conspiracies. This kind of a conspiracy doesn't even require for tech corporations to come together to collude and conspire - it is implicit and goes without saying, it makes things very easy and very profitable for them as they well know if people don't have a choice for anything better they will settle for one of the available as bad as it may be (US elections 2016 LOL). They do a lot to maximize their profits, and those profits come out of consumers' pockets. They have just about zero interest to sell affordable flexible and interchangeable modules, which would allow people freedom of choice, upgradeability and repairability, as that would simply slaughter their profits. Many of them "struggle" enough as it is because of their dumb business models. Modular phones are ENTIRELY 100% possible and the design and execution are not all that complex as google's intended failure aims to convince us, and it would be a great thing for all people as well as the environment - we are talking about the whole planet here, but corporations don't care about mankind or the planet, they care about their profits, and they all have a long history of actually phucking up people and the planet deliberately for the sake of profit. Corporations would need to restructure and shift their models and priorities, and the people behind them are dinosaurs long obsolete but refusing to die off. Allowing a modular phone to succeed would be tad amount not to shooting themselves in their leg, but tho shooting their head off. Under the current corporate model that puts profit before everything they are simply not in the position to allow this, much less to make it themselves. And they have the manufacturing complex and the patents in their hands, so they can easily prevent anyone from doing it.
  • boozed - Saturday, September 3, 2016 - link

    You need to take a step back and a deep breath.
  • ddriver - Saturday, September 3, 2016 - link

    So now that you are faced with a logical cause-effect chain, you are down to insinuating things to bring my judgement into question? Keep up the great job you are doing...
  • close - Monday, September 5, 2016 - link

    ddriver, I know "hardware standardization, interconnection, compatibility as well as software support" takes a few days tops for you and your 5.25" hard drive because you're a one man team doing everything from googling about hard drives to posting comments about hard drives and showing aggression when people call you a "chronic bullshitter" (skills that take ages to master).

    But when it comes to building any kind of modular system in which many manufacturers are involved things get "a little" more complicated. Their people simply don't have the engineering, management and negotiation skills that you have so it takes just a little bit longer for everyone to agree on the standard and actually turn it into something marketable.

    Maybe you should submit your CV. I'm sure they can't ignore your kind of genius.
  • ddriver - Monday, September 5, 2016 - link

    Yep, take it from a broken record. You do know your stuff LOL. Let me guess, "close" stands for "close minded", right? How very suiting.

    Standardization is only an issue when it involves standard committees made of competing interest corporations. Such committees are known to be extremely slow to produce extremely poor standards, as their primarily task is for each to pull things in his own direction. And in the end the standard is in favor of their companies, not the consumer. Standard committees have ruined more than a few great things. Coming up with something as simple as a reusable form factor for mobile devices is quite literally a few days of work, and most of that would be refinement, not "figuring it out". It is actually an extremely simple problem compared to the stuff I deal with on daily basis.

    Join google? Or any of the big greedy corporations? Don't make me laugh, there is a fundamental rift between my mindset and their so called "aspirations". They will never be willing to execute my ideas, as those will put them out of business, my ideas are all in the best interest of the consumer. As a matter of fact I have already refused a seven digit offer for one of my projects, and will never sell out at any price, because the only reason for them to want it is to drive it into the ground, because those relics know it is something they simply cannot compete with. There is not a single large corporations which can be trusted with or deserves my ideas. They have already ruined so many great things, including the vision of Nikola Tesla - probably the smartest person in human history. Only a compete fool will fail to learn from the past.

    It is finny how you keep advising others to "not bother" with me, yet you keep bothering, and only to spew out the same silly stuff. And of course it makes sense that when you are silly, smart things will sound meaningless to you, because you are oblivious to their meaning. You wouldn't know smart if you hit your face in it. I will actually take you up on the "do not bother" advice, as you, poor soul, have time after time proven to be the intellectual equal of a parrot. Cherish this last generous gesture of pity for you and good luck with being complacent subservient dummy ;)

Log in

Don't have an account? Sign up now