Michael Tremer

The fairytale of the "Internet of Things"
by Michael Tremer, March 27, 2015

Arne, Hannes and me visited an exhibition called “Embedded World” last month. A really nice one with lots of interesting new products it seems. Lots of companies took their latest developments and presented them to the public.

Exhibitors and visitors came from countries all over the world. It was not like other exhibitions where you have mostly marketing people any more who try to get other marketing people into buying their stuff. It does not really matter what the stuff is they are selling. It is just about the sale and making money. With what ever sells.

This is not completely new. I have noticed this at many occasions and in different forms. It just has never been that obvious to me and it kind of seems to me that some people are not even trying to hide it any more. They think that their customers are that stupid and it seems that they are right. They are stupid enough to buy that. Both the story and the products.

By that I am talking about one very specific point:

There is no innovation any more

My point I want to lay out here is that companies do not have either the potential or need (or possibly) both to innovate new products any more. At exhibitions you will find people who are working at the booths and are supposed to present “new” things. If you ask them what the new hot things are, there is usually no answer. You will certainly be shown some things and you will certainly hear a long answer to your question. If you remove all the marketing phrases from that answer you are usually left with not that much any more.

The “Internet of Things” (IoT) is the label for that. Let’s quickly dig into what the Internet of Things actually is: Marketing tells us that in the future everything will be connected to the Internet. Every device will be able to talk to the other devices on the network. Let that be your thermostat talking to your light bulb or coffee machine. We can all image what this communication can be good for – or not. Marketing people also tell us that we need to prepare for that. Why? Because it is coming! How? By buying their new products of course.

I personally think that we have already reached a point where most of those devices are already connected to our networks. Computers, phones, printers, TVs and what not. The most important devices are already on the network. There may be many to come. I am just not too sure in what way I can prepare for my new WiFi-enabled light bulb.

The IoT label (I am sure you have already seen it) is now put on almost any device. So that we know what a thing is. Mainly the vendors of routers/firewalls seem to have “invented” something new. They did not just put the label on their products. No. They made their product “ready”. But in what way? The answer is very very simple: In no way. There is nothing you can do on the firewall to make your network ready for the IoT. If at all you would want separate networks for your home automation devices and entertainment devices. That is something you would do on a completely different layer of the network. Not the firewall. So we can conclude that even they just put the new label on their product. Now it is enabled for the Internet of Things.

You know what else is enabled for the Internet of Things? IPFire is. It has been ready for the last decade. Really. That’s how 2015 we are. We enable IPFire for trends before we even know that this will be a trend. You want to know what we did for that? Nothing. We just implemented what a firewall does. Nothing else. Nothing fancy that costs extra.

I find this is a pretty easy example for the absence of innovation. Sometimes products are done. You cannot add too much interesting stuff any more that is important for a huge number of people (and therefore customers who bring you the money). You can improve things; you can always add features. Those are just not that ground breaking sometimes.

Exhibitions & Conferences are dying

I think you draw a long line from here. Because of the lack of ground breaking features exhibitions are dying. CeBIT is almost dead for me. We had our own IPFire booth back in 2010 where it was really fun talking to all the people who were interested in the project. The years after that those people did not come any more. They send their sales and marketing people. Those do not really have much interest in many things. They just want to make money. If you have something that is free that is a good start for them because all of the money they can get from their customers is their own profit.

There are also many other conferences closing down or at least they have difficulties attracting people. Why should people come when there is nothing to see? Companies and Open Source projects announce their new things online and do not need an exhibition for this any more. I think that the purpose of those exhibition and conferences has been shifted from presenting products to meeting people. Unfortunately not the people who engineered the products.

No engineers needed any more

Maybe companies do not even need those engineers any more. This would just support my point.

One of my interests was to find out if there will be solid products based on the ARM architecture. Most of the major companies got something based on ARM nowadays. I was really surprised that they all tried to talk me out of it. ARM wasn’t going to be that good as x86 and what not. They also do not have much software that runs on their hardware. Interesting. I concluded that they just had to add these things to their product portfolio. Nothing says “Internet of Things” more than tiny devices. ARM boards are tiny.

Almost all of those ARM-based single board computers were based on the Freescale reference design. Very similar (if not completely similar) to the Wandboard. The selection of connectors was the same every time and there was not too much difference between all of them. It was the same System on Chip (SoC) with some periphery. Designing such a board is a pretty easy task if you use a SoC. The most important bits are already in there and you simply need some power and connectors. Done. Students can do that. It is even more easy if you have a reference design you just need to copy. Why pay an engineer for that?

The SoC cannot be changed unless you need a huge number of them. Even then it does not make too much sense as there are other solutions and still is very expensive to do. That will also become true for the x86 architectures where SoCs are just coming up. I fear that these systems will become that similar if not identical as they are in the ARM world. There is no value that the board manufacturers can add any more. They will take the SoC, put it on their boards, maybe add some peripheral devices like an extra network adapter and they are done. Everything else is in the SoC which is the same for everybody. I guess this is also the main reason why the sales people tried to talk me out of the ARM products: There is no value they add and therefore not as much profit as on other products.

The end of the story will be that there will be again no innovation any more. The board manufacturers cannot do much any more. The market share of x86 processor vendors will unlikely change in the near future. So there is just that one player who will drive innovation. But why should they do that if there is no competitor? They have not done it in the last couple of years and they won’t do that much in the future either. Why? And what for?

The Internet of Things is just a huge fairytale that helps people to sell their old products once again in a new shiny look. After that there will be a completely different hardware market.

Maybe I am right with all these assumptions. I hope I am not.


Posted: March 27, 2015 • 1670 views