Jun 26 · 5 min read

Having looked at a large number of so called “Industrial Internet of Things” early stage startups over the last 3 years our main conclusion was that — at least today — there is no such thing as the Industrial Internet of Things.

The most obvious difference between a large majority of the IIoT projects presented to us and the “existing” Internet (the one we are using daily) lies in the objectives for architecture and governance. On a high level this difference could be described as “horizontal and open” vs. “vertical and closed”.

The existing Internet is characterised by generally agreed universal communication protocols (mostly use case and vendor independent), universal representation of data and free access to vast amounts of information and technology. On top of that, the increasingly popular concept of APIs allows services built by one party to expose clear and open interfaces in order to invite other parties to connect and build the next service on top.

Industry on the other hand is still characterised by a multitude of proprietary communication protocols (use case and vendor specific), infinite different representations of data and closed doors (and network ports). Instead of defining “API type” interfaces most industrial players seem to be stuck in a narrowly defined decision point to either open up (all?) their data to cloud services or not allow external data connections at all.

Why “all-in-one” solutions are probably not the answer

We have seen several startups trying to overcome those shortcomings in today’s IIoT by offering an “all in one” combination of retrofit sensors, communication modules, hardware gateways and cloud application. While those approaches certainly are a faster way to first proof of concept projects and pilot installations in an early market, we see a huge risk of building vertical / single use case solutions which in the end just reinforce the above described shortcomings of today’s IIoT world. At the latest when industrial firms move from pilot to factory wide or even global roll-out there should be a moment of consideration to make sure the chosen IIoT architecture and governance is the right one. Aspects to be considered are:

  • future proof (will it build a solid foundation to support future services & extensions like edge computing modules, “API type” connections, etc.),
  • scalable (will it deliver the required performance once rolled out to larger numbers of machines/factories/users ?),
  • manageable (will the IT team be able to keep it up and running/ deliver the required uptime?),
  • secure (can it prevent outsiders from stealing critical data or manipulating the system?), and
  • efficient (will associated costs with roll-out, daily operation, maintenance and future updates be supported by the generated value add?)

Industrial companies want to own their data

Some Industrial firms have started to take this thinking about architecture and governance very seriously (we have seen very detailed RFPs emerging which describe exactly those aspects of “Industrial IoT architectures”) while others seem to be stuck in a strategic dilemma around “ownership” — ownership of individual data as well as ownership of the IIoT infrastructure as such.

Every player in the industrial value chain seems to have read by now that “data is the new oil”. Almost every single one of these — from component OEM, machine OEM, system integrator to machine operator, predictive maintenance service and IoT cloud platform provider — tries to create lock-in by owning data, controlling and even operating as much as possible of the IIoT architecture. Future will tell if really all those players are well equipped to design & operate future proof, scalable, manageable, secure and efficient architectures to cater to their own long term needs and those of their respective customers. Quite a few seem to be building siloed, non-scalable, non-secure and costly systems in the attempt of owning data & IIoT solution.

The race for the dominant IIoT cloud platforms is already decided

And while other players are still debating about ownership, the well known cloud infrastructure operators Amazon aws, Microsoft Azure & Google are clearly the players that have gained already a very dominant market position also in the Industrial IoT. They have started very serious efforts to dominate Industrial IoT cloud offerings quite a while ago. Most of the other “Industrial IoT platforms” are actually running on those dominant infrastructures and therefore rely to a large extent on their architectures, governance concepts as well as business strategies (a relatively basic one being: the more data is uploaded the more business for the infrastructure operator). As the infrastructure business is primarily driven by scale in datacenter operations as well as available capital for acquisitions of innovative technology startups and developer talent, there is probably no case anymore for European companies to compete head to head with those US based operators. The question is not anymore “IF one should be open to connect to those dominant IIoT cloud platforms” but rather “HOW to connect to those IIoT cloud platforms and maintain sovereignty”?

Paving the way for something worthy of the name “Industrial Internet of Things”

We think that there is a certain urgency to now define a horizontal, open and use case independent IIoT architecture which is future proof, scalable, manageable, secure and efficient — before today’s suboptimal situation is cast in stone by larger roll-out projects. Cybus’ Connectware Software and the Cybus team’s vision have convinced us to be one step in the right direction. Their software based data gateway or data broker is usually installed “on-premise” in a factory’s IT system and enables:

  • to build a factory- or enterprise-wide semantic data model on top of a heterogeneous machine park,
  • to add edge computing capabilities,
  • to manage data flow & data access rights for each machine or datapoint, and
  • to act as a single connection point for any external cloud service, any data storage & data analytics solution — be it cloud based or installed in the enterprise IT system.

This allows any factory to actively distribute data across IT systems, share selected data with external suppliers, always stay in control, work with a multitude of cloud platforms and switch between them when useful as well as reduce the cyber attack surface substantially. Cybus is particularly focussed to enable efficient roll-out and operations through automated deployments and central configuration capabilities. Cybus will never own any data but enable component OEMs, machine OEMs, system integrators, machine operators, individual service providers and IoT platforms to exchange data in a controlled way, establish clear governance & trust and pave the way for something worthy of being called the Industrial Internet of Things.

Our Industrial Technologies team will continue to look at many more IIoT companies and pursue additional investments linked to the topic — as whenever there are shortcomings, there are opportunities! Please do get in touch with us to discuss or challenge our perspective and let’s collaborate to build the most valuable Industrial Internet of Things possible in Europe!

brains to ventures

btov thoughts from the digital and industrial technology worlds


Written by


Industrial Tech at btov partners . Early stage investments in innovation across the industrial value chain (IIoT, Data&Analytics, Additive Manuf., robotics,..)

brains to ventures

btov thoughts from the digital and industrial technology worlds

Welcome to a place where words matter. On Medium, smart voices and original ideas take center stage - with no ads in sight. Watch
Follow all the topics you care about, and we’ll deliver the best stories for you to your homepage and inbox. Explore
Get unlimited access to the best stories on Medium — and support writers while you’re at it. Just $5/month. Upgrade