So recently there has been a lot of debate and excitement about Phonebloks (http://www.phonebloks.com) which I too believe is an awesome idea. However, there is a reason why it is still merely an idea and why I think it will remain as such for some time to come. Here are some of the reasons I believe something as cool as Phonebloks won’t take off:

The average user doesn’t want to design and assemble their own phone:

Just look at PC’s how many PC owners actually know what goes on inside and what¬†components¬†do what. An example of this is the number of times you hear people say “Oh my PC is slow so I must buy more RAM.” No No No RAM is not the only aspect that makes a PC slow, what about latency in reading and writing drives etc. The real answer to why it is slow depends on what you are doing and what spec hardware the rest of the PC is made from. It doesn’t help you run an i7 Quadcore with 4GB RAM and a 32bit OS. The only thing extra RAM will do here is empty your wallet.

Hardware manufacturers don’t care about electronic waste:

Apple, Samsung, Nokia, whoever else, don’t really care what you do with your old phone as long as you are buying their latest model. The companies gain an edge by having products that look unique, which have hardware that has been chosen to work well together and by optimizing OS code to manage this hardware effectively. By minimizing electronic waste and having pluggable components you allow the user to chose what goes into a device, and this could prove detrimental to hardware based companies which focus on hardware design more than software. Furthermore, you are also likely to cause inefficiencies in the device as users might pair seemingly suitable hardware together which might not actually work well together.

Phones are highly optimized:

Ask any hardware engineer, cellphones are designed to be as thin, light and streamlined as possible while carrying the most features. This design process take months and requires highly skilled engineers to ensure the device is still functional while being made as compact as possible. Why do you think iPhones can’t have their battery taken out, it’s not Apple being full of shit it’s because that clip mechanism takes up space. By having a non removable battery you can leave the circuit board uncovered and thus fit it as close as possible to the battery and screen without worrying about a user touching it. For Phonebloks to work you will require every component to have additional hardware to act as an interfacing layer to the “base board”. This is due to the fact that every chip has a different number of pins, pin layout, communication standards, voltage standards etc., while most people will say “well let’s just standardize all chips to be the same” this is not at all feasible as there are reasons and IP behind all those standards. The problem with this interfacing layer is that some small sensor (gyro, accelerometer etc.) now becomes increasingly more expensive to buy and takes up much more space than when it is crammed onto that tiny little PCB inside a modern smartphone.

Software to match hardware:

Many phone companies design, or modify their OS of choice to manage the existing hardware. If you have user interchangeable hardware how will the OS know what communication standard to use and what hardware to manage. This is not impossible to overcome, but it will once again add to the cost and complexity of such a device. Let’s assume the “base board” has some type of FPGA dynamic routing architecture. The user plugs in their modules and edit the configuration file to say where each device is plugged in on the base board, and let’s assume this is all done through some very simple user friendly interface that even my mom could do it. Well now the problem comes down to speed. dynamic routing is expensive and is not all that dynamic. Certain buses and communication channels interfere with each other and thus are normally fixed to predefined locations on such chips. Furthermore, by implementing dynamic routing you will drastically affect speed as some type of lookup table will need to be used in order to route signals. This is why FPGA’s are useless for serial computation, their power comes from their ability to perform tasks in parallel.

Even if some highly efficient hardware routing system was used to route signal perfectly between different devices, there is still the problem of how do various OS’s manage this hardware as if there is no standardized communications standards between devices the OS will have no way of knowing how information should be formatted. And if a standardized communication system is chosen you lose the effectiveness and edge that some of the other standards allow. FOr instance who will become the deciding body between USB 3 or Thunderbolt? Or should sensors use I2C or SPI? Each of these standards have their own set of strengths and by adding a secondary layer of abstraction you generally lose the strengths of each individual technology.

Support:

The final point I would like to make, is who will provide support for these device. If a module breaks or the phone doesn’t work (as it is unlikely the average user will know what part is broken) who is responsible for fixing and managing the warrantees of such a device. Large corporations will probably not want to deal with claims of a single component here and there. It is much easier to replace an entire device and ensure a happy customer than fight over a single module which might have actually broken due to another module not abiding by the standards laid out in the specification of such a device. It is not uncommon that some components do not follow IEEE specification for the implementation of some signal bus. FOr example I have found USB ports which supply 4V instead of the 5V laid out in the standards.

While many other issues exist these are the main ones I feel will have the biggest impact on the realization of such a device. I like the Phonebloks concept of reusable, upgradable smartphones, but at the end of the day no matter how many people say they want them I just don’t think it will happen. Or not yet anyway. The tech industry has many other issues to solve, and a reusable smartphone isn’t one of them. People have PC’s which is essentially the same concept, however how many individuals or companies actually upgrade the parts in the PC? Not many in my opinion. Most just toss them out and buy the latest and greatest thing at the moment. Furthermore, mobile technology and chipsets are improving and changing way faster than desktop hardware and thus it is even more unlikely to be able to develop a plug and play platform that will allow users to quickly and easily upgrade their device without losing the performance of the latest hardware in the process.

But come on world prove me wrong!