berry-1238249_1920

Everywhere is internet and everything gets connected now. What we need is secure personal hubs that control our data with generic software.

In this article I would like to look at some aspects on the technical side.

A networked connection simply means that more than one device is involved to form a combined solution of some given problem. This means that two things usually happen:

  • Devices exchange data, in a conversation pattern like humans on a telephone. In technical terms this is about APIs.
  • Computing itself becomes distributed: The classic elements user interface, processing and storage run on separated and specialized places. They exchange data over the network.

Let’s focus on the second point. The mobile revolution brought us computing power in abundance. It’s not restricted to data centers or the good old mainframe any more.

But there is one tiny problem: What software should run on those powerful devices? Where does it come from and who takes care? Standalone devices like phones (besides the “smart” aspect), wearables or the IoT devices usually come with their software included, right out of the box.

With smartphones and tables the distribution is different – apps can do anything, they are “generic”. And yet they are managed in a perfect way. The developer loads his program into an “appstore” and the mobile’s owner just uses it. Every other aspect like download, installation and even the most parts of maintenance are hidden and cause no effort on any side. This has led to an amazing effect in software distribution:

Selling apps scales perfectly.

It is a bit more complicated for example in the emerging IoT market. Most devices come with two software options:

  • A dedicated app (including SaaS-like web-interfaces) that controls every function. Easy!
  • An API that lets more generic software – that is not included – control the device.

Let’s focus again on the second point. And also again: Where will the controlling software run and how is it distributed?

IoT vendors already established the concept of “hubs”, but those are usually restricted to controlling devices of their own brand or at least a certain sub-standard. And the selection of software is limited: A system for central heating will probably not be designed to process datasets from a fitness tracker on the same network.

So if one clever startup founder has the idea of combining those two, so that the temperature in the shower is comfortable whenever someone returns from jogging?  Where would the controlling software reside? What if the next project would base on security camera inputs, with huge data streams and machine learning?

The obvious solution here will be the generic personal hub. I’d define it like that:

  • It could be compared to the classical server, as far as the hardware is concerned – be it a standalone device or something virtual in the cloud. Whatever it is – it’s personal.
  • Software modules will load automatically following the appstore/mobile pattern. To achieve the same level of perfect scaling.
  • The performance factors like computational power, storage capacity or network access will be chosen by the owner.
  • Security and privacy are essential. Designed to eliminate potential flaws of the controlled devices.

Those hubs won’t be restricted to IoT. There are other areas, where a “blackbox” personal computing instance will be useful:

  • Everything that has to do with privacy.
  • Selling apps that share data between mobiles, tablets or web-browsers without the app-producer’s need to provide a costly cloud service (so he can sell without a subscription).
  • It’s clear who is responsible for the data – the hub’s owner and not the app creator.

There many more interesting aspects if you start thinking about it.

Potential problems as well, namely the notorious standardization. Maybe also IP pitfalls in more restricted regions outside Europe where patent trolling is common. It will change the jobs for infrastructure experts and push the still new devops movement to new challenges. But all those problems won’t stop it from happening.

I believe that the Web of Hubs will come – soon. It will grow fast and it will stay for a very long time.