First the Cloud, Now Fog Computing

[Arthur Hanson]

Cloud computing was just the start, now we have Fog computing, the layers between IOT and the Cloud, in short computing is now getting omnipresent in everything and everywhere. This is going to not only add opportunities, but severe management and security challenges from what will become a diverse data stream of multiple languages running over multiple communication protocols. Add to this the greatly varying quality of ecosystems involved in all aspects and the probabilities of failure increase dramatically. In a case like this the most streamlined, direct architecture should present the fewest challenges. It will be the ecosystems that put forth the best model that should win the game. Systems that just plug into a wild west model will do so at their own risk. Below is a link on the subject.

https://arxiv.org/pdf/1611.05539v3.pdf

 

[count]

What is getting lost in the hype around IoT is that IoT is not a new technology. It's control systems engineering, where some of the layers of the control system are moved to the cloud.

Take a look at this picture the GE is using to market it's Predix platform.

If you have any background in control systems engineering, you'll notice it looks an awful lot like a control loop diagram.

This is not a coincidence by any means.

Now look at the following control system architecture. Looks a lot like what people today are calling fog computing, except this sort of architecture has been around for 30 years.

IoT is 90% hype. Companies are simply taking old ideas, old technologies, putting new branding around them, and trying to sell it in a service model. That's fine, and there might be some advantages to the service model and moving some functionalities to the cloud. However, what people need to realize is this is a 5% growth market, and new branding isn't going to turn it into a 20% growth market.