Co.Design

Why Lego Design Principles Don't Work On Smartphones

By now, you might have heard of Phonebloks, a modular smartphone design concept that is taking the Internet by storm. Created by Dutch designer Dave Hakkens, Phonebloks proposes a better smartphone that is made up of Lego-like modular components be upgraded individually. According to Hakkens, such a system would decrease electronic waste and cost consumers less money on smartphone upgrades over time.

At best, most of us have a dim understanding of what is going on inside of our smartphones, but the Phonebloks concept breaks out each separate function of our smartphones into its own module. There's something intensely satisfying about this approach: is this not the way we secretly wish all technology worked? And don't we all wish that when a new iPhone comes along that we could simply upgrade the modules that matter, like picking pimentoes out of a sandwich and replacing them with olives?

It's a shame, then, that Phonebloks is a pipe dream, a concept by its creator's own admission is not even achieveable within the next 10 years. "Making Phonebloks a reality is probably impossible with current technologies," Hakkens told Co.Design in an email, but didn't elaborate further. However, it's easy to extrapolate why.

Phonebloks' core concept is that every technology in your smartphone should be capable of being broken out and upgraded as an independent module, similar to the way a PC motherboard works. You should be able to upgrade your display independently of the CPU, independently of the graphics, independently of the RAM, independently of the Bluetooth, and so on. That's a nice idea, but it ignores a bunch of practical problems.

Smartphones Are Integrated For A Reason

Within your smartphone, data whizzes between components at speeds that are nearly impossible to imagine. Every milimeter's distance between these components comes with a speed penalty attached, which is why smartphones tend to put as many components as possible on a single chip. Consider, for example, the iPhone 5S's A7 processor, which has the iPhone's CPU, graphics and RAM clustered together in a sandwich-like wafer.

Breaking this trinity up to allow for modular upgrades wouldn't just make the device run slower, though. It would make your iPhone consume more power and triple its physical footprint. The result would be a bulkier device, or a device with less room for other components (such as a bigger battery). Even if you could live with that, though, Phonebloks would require expensive sockets so that the CPU, graphics, RAM, storage and modem could communicate with one another at high speed.

Phonebloks Would Increase E-Waste, Not Eliminate It

So even at first glance, we can see that a Phonebloks smartphone would be bigger, slower, and more expensive than a regular smartphone. But maybe it's worth it if we're not throwing away our smartphones quite as often because they've become obsolete, right? Unfortunately, no. One of the little illusions the Phonebloks concept pulls off so well is that it fools us into thinking we're seeing a simpler way of designing a smartphone. It's sleight of hand. The reality is that the Phonebloks design is a more complex way of building a smartphone, and there's a lot of things that can go wrong.

How do most smartphones break? They break because of user error: you drop them, or you smash them, or you do something you're not supposed to do with them. The Phonebloks concept would take the human error factor that ruins so many smartphones and multiply it with every module. As users constantly pry modules off, replacing them and switching them around, there's just a million ways a Phonebloks smartphone could break that an integrated smartphone is not susceptible to. Instead of throwing away our smartphones because they were merely obsolete, then, we would throw away our Phonebloks because they were breaking all the time. And because these modules would be bigger than regular smartphone components, we would actually be increasing e-waste, not lessening it.

We don't need to labor the hundred other ways in which Phonebloks is a designer's dream but an engineer's nightmare. Simply put, Phonebloks is the opposite of what it appears. Phonebloks makes an appeal to our love of order and simplicity, while actually being significantly more complex. Phonebloks tells us smartphones can cost less, while making each component within them cost more. Phonebloks says that we can upgrade our smartphones without being wasteful, while making it significantly more likely that we'll have to throw away our phones because they're broken. And so on.

Phonebloks: Not A Prediction Of The Future Of Smartphones

Phonebloks appeals to the many facets of the modern ego-id at once: the part of us that wants the universe to be neatly ordered and precisely aligned, the child within us that wants technology to work more like Legos, the guilt that follows from throwing hundreds of dollars of electronics away every year because they are "obsolete." Phonebloks makes us feel good about technology, not confused, covetous or remorseful. Practicalities aside, it's easy to see why the Phonebloks concept went viral. But the whole point is that practicalities can't be put aside. They need to be dealt with and overcome.

"In the 1960s, Intel's Gordon Moore said that every two years, processor clock speeds would double, and he was right. Later on, he said that Moore's Law wasn't a prediction, but a goal," says Hakkens. According to Hakkens, that's what Phonebloks is: a goal, not a prediction.

And a good thing too. The parameters of goals fluctuate according to the demands reality place upon them, while predictions are usually the domain of cheap fortune-tellers who are depending on some supernatural force to intervene and pull it all off. As a goal, Phonebloks might get us all thinking more seriously about electronic waste and how we go about disposing of our devices. As a prediction, though, it will require magic to make a reality.

Add New Comment

70 Comments

  • d.free

    From a hardware perspective, it's impossible.

    Look at the modular design and all those connections. Do you honestly think that the mobo of the phone can magically wire the connections together to the modular cpu, screen, etc. properly? A modern PCB on a phone has many layers and careful engineering to get all the wiring right, this Phonebloks system is simply plug-and-play. Good luck with that.

    Even if you could wire everything right, think about power. Different screens, chips, cameras, and more have different requirements. Unless the Phonebloks manufacturer intended for us to SMB solder minature resistors and such to keep the voltages in check, the phone would most likely be fried.

    And drivers? Nice try. It's hard enough to build a PC and get 100% compatibility (and the DIY PC market has been going on for years), now try that with a phone.

    Magic, yes. Reality, no.

  • Avi Bernshaw

    True the technology to make this doesn't exist yet. The point of the project in its current phase, if you've read the entire website, isn't to make a working unit. This isn't a Kickstarter.
    Hakkens isn't saying he's got something to go to market with, he's saying wouldn't it be nice if, with all the computational power we have now, companies started to make phones like this so individuals can choose what they want.

    And he's totally right.

    My phone, a Galaxy S3, has more computational power than my now 7 year old desktop work computer, yet when it comes to anything but playing games (my only upgraded component since purchase has been the graphics card) they both work at entirely the same speeds - MY SPEED.
    If I'm making a phonecall, I speak at the same speed on Skype as I do on the phone. If I'm typing an email, I type 118wpm on a keyboard or 62wpm on predictive text. If I'm making a new track, it takes me as long as it takes to listen dozens of times and swap beats around, but the processing of the audio is a couple of seconds on either device.
    Who cares if there's extra latency? Can people type so fast on their mobiles that it matters? Who's doing 1080p video editing on their phones or creating vector images for billboard advertising?
    Unless your phone is primarily a gaming device losing speed isn't going to matter, and if it IS primarly a gaming device, why not just get a PSP and install a hacked version of Skype to make calls?

    Hakkens whole idea is to get people to say that they want this as a product so that manufacturers actually start to think about it.
    Once the basics are done, who knows what other components will get made for it.
    Maybe it will herald in the age of wearable micro-component computing. Maybe we'll just get a smartphone made from LEGO.
    Either way we'll never know if people just say "That's childish and you're childish for wanting it." With attitudes like that, we'd never get anything new or exciting. Ever.

  • ern0

    Great article.

    I have another argument against modular design: the interface between parts must be specified and engraved into rock, which makes tehcnological jumps impossible. You just can't plug any processor and memory into any motherboard. Okay, we should change motherboard. But we never do that: ask your friend, how much he was spent on changing the mohterboard in his Mac. Also, upgrading the motherboard requires processor and memory upgrade as well.

    You will upgrade your phone just as you upgrade your notebook: never, nope. You buy a bigger HDD or SSD for your notebook, and bigger SD card for your mobile.

  • Michael Baker

    I don't think your argument is entirely fair.

    You say that you can't drop any RAM or any CPU into a motherboard, that's true, however you can drop any supported CPU or RAM. It doesn't entirely null how upgradable the phone is, just limits it - exactly how modern PCs are limited in the respect.

    Using a Mac as an example for replacing parts isn't a fair example of how expensive a motherboard is to replace. Macs are designed so that you have to get replacement parts from them, and apple isn't known for being affordable. Also, what's better? Replacing an entire mac or swapping out that motherboard. Personally I'd prefer the latter.

    Provided that you keep the same sockets on the motherboard you won't need to upgrade your RAM or the CPU if you replace it. The only time you'd need to replace them is if they're built into the board or you change the socket type.

    The reason you only upgrade a notebooks RAM or Harddisk is because they're not designed to be modular. The idea of lego-type modularisation is akin to that of the modern desktop computer - aka parts can be swapped.

    Having a standardized interface isn't bad as it can still be upgraded and updated. If for example they release a more powerful motherboard or upgrade the firmware to support new hardware the old interface can be supported and a new interface extending it. Meaning that you replace the motherboard or update your firmware, while keeping your blocks and allowing for more blocks.

    I do admit that modularisation has many flaws, but so does integration. Integrated systems are comprised of a static and unchanging system of parts. It's smaller and the speed boost is insignificant. If any part breaks you have no choice apart form where are you buying a new system from.

    I myself am in favour of modularisation, despite of potential pitfalls. From a programming perspective a modular system is normally a requirement, since you can update and upgrade parts without replacing the whole system. In-fact you could say that operating systems are, in a way, modular. That is that you can add or upgrade programs rather than being stuck with a default set and this has been chosen over a default set because it makes it adaptable at the cost of a little speed.

  • jj

    Well I disagree with just about everything in this article.  Go back to working for Microsoft.

  • Andy

    Do you disagree that it would makes phones MUCH slower and bulkier? Because that is simple fact.

  • Marcos

    But let's think the simpler way: our PC or notebook is exactly what our smartphone is: an already modularized system, with independent assembled modules, build by a lot of companies. This design would be possible if scalability changes things for better, while inputs and outputs keep the same. Screen, memory, processor, camera, these 4 could be replaced with no problem, like the video showed, if a modular-design is the main engineering goal.

  • Marcos

    And I'm a computer engineer (btw, authority fallacy isn't the case). I'm sad about what electrical engineers are saying. If a project runs perfectly on a CAD simulator, the only thing you need to do is to change constraints and test... why would it be IMPOSSIBLE to made it phisically? What could prevent us, for example, to program electronic circuits to run at certain clock speeds, depending on tech evolution? Add memory? Ok, just do it... Could the wiring array be a new sort of advanced FPGA, so different modules can fit? Yep!! 
    All of that is possible to implement.

  • Cedric Mamo

    An electrical signal travelling at the speed of light (it doesn't in the real world...) would take almost 1 ns to travel from one end to the other. double that for sending a signal and receiving a response. That adds up to almost 2 clock cycles. In practice it'll be more.

    It would introduce additional latencies and slowdowns, never mind the fact that you can't just route high speed signals randomly on a board. There are a ton of factors to control.

    Don't work under the assumption that all electronics involves is connecting the right wires together. If you're a computer engineer, you should know better

  • Marcos

    All problems can be contoured. Design can be so far we can't even imagine, and durable, resistent, easy-to-install, energy-efficient and recyclable pieces could be the norm.

    This is happening today with furniture industry (3D printers will simply skyrocket customization, for example).

    Unfortunately, none of these ideas are "economically efficient" or "profit driven". That's a sad world we live in. Capitalism is essentially against technology.

  • günay geyik

    If you insist on keeping the current fashion of smartphone design then yes, it is impossible. but the real question is, Why is it too hard to see that this also brings another idea  of implementation of smartphones? it cant be hard to implement as the article claims, there might be imcompentencies among the blocks but this can be easily overcome and not a serious issue, not at all. All in all, the idea is nothing but a modernization and compactization of the classic computer design (cpu ram, mainboard etc), so no need to re-invent the wheel.
    Th other claim on the cost is another hoax. The cost of a phone will totally depend on the user in that case cause he'll be able to choose deeper. If he chooses the most exclusive ones, then yes, cost will even be ahead of iPhone. Otherwise, not necessarily.

  • Cedric Mamo

    cost will *always* be much more than current devices. It's not a question of choice. Each block will need its own packaging (and I don't mean the box it comes in, I mean the plastic box which contains the actual component), which would need additional manufacturing costs. Then there are the connectors which have to be included with each component. Possibly some extra chips to communicate over the, as yet inexistent* magical bus which will carry everything (the technology to make this doesn't exist yet).

    So even if you choose low-end components, you could still get a similar (feature wise) low end phone for *much* cheaper

  • Mark Stevon

    This whole thing is crazy. It's a distraction. You could play music on these things at a picnic, or try playing it yourself and have a good time jamming your friends under the sun. You could avoid getting blasted daily by electro-magnetic radiation, or you could, well, do that. You could fund an expensive phone (even though you might already have one which is likely perfectly fine at its job), or you could fund a charity. Etc.

  • brsma

    Both the article and the majority of comments here miss the most crucial point. On the premise that this very well presented concept is a vision that relates to a time about 10 years in our future, the same critique should be applied that Bret Victor based his great http://worrydream.com/ABriefRa... on: this concept just extrapolates a current and meanwhile quite conservative solution to personal communication and computing without basically challenging the now already traditional idea of the smart phone at all. In other words: The concept's designer actually stays confined to the very box he is pretending to break out of.