IoT Highlights from Embedded World 2019

Nuremberg fair reveals shifts in operational technology

Embedded World focusses on the computing needed within all types of machine. This spans a huge set of suppliers, covering everything from insulin pumps to oil rigs, collectively known as operational technology (OT), as distinct from IT. As more machines are connected, Embedded World has become Europe’s premier trade show for enterprise Internet of things (IoT).

Unlike the glitz of MWC in Barcelona, which takes place in the same week, Embedded World is a more down-to-earth, technical affair and addresses what’s achievable today. The clearest example of this is 5G. This was a huge theme at MWC, with numerous exhibitors and speakers presenting their visions about how the technology will transform industries. Meanwhile in Nuremberg, we found only two stands that mentioned 5G.

One of the clear messages from the show was that the OT area is changing quickly, thanks to increasingly rapid adoption of IoT and a huge focus on artificial intelligence and machine learning, as companies strive to get maximum value from the amazing quantities of data now being generated. Many of the stands highlighted edge computing and edge intelligence, and the show’s theme this year was embedded intelligence.

Developments from the show are nicely illustrated by an announcement from Microsoft using its new Azure Kinect developer kit for robotics.

Neat @Microsoft demo of Azure Kinect and windows working with Robot OS. Hololens-developed 3D positioning with orientation is a big step forwards. #embeddedworld2019 pic.twitter.com/OGc5o1U7rL

— Martin Garner (@martin_garner) February 28, 2019

Another highlight came from UK start-up GreenWaves Technologies, which demonstrated machine learning on battery-powered devices.

Developing theme at #embeddedworld2019 is AI on battery powered devices. ⁦@GreenwaveSys⁩ demo self- navigating drone is a good example. pic.twitter.com/pSfLRfEQF0

— Martin Garner (@martin_garner) February 28, 2019

These developments bring two major shifts that OT suppliers and users are navigating:

• A move from discrete, disconnected, specialist machine controllers to a distributed and connected system, with all the issues of distributed data, distributed computing, system optimization and security that go with it.
• Companies that have historically been hardware-driven now need to learn modern software, cloud and machine learning practices. They’re moving into a new world with more off-the-shelf hardware, where most of their differentiation will come from software.

One analogy is that OT suppliers are facing their “iPhone moment”, when their world risks being turned upside down by the arrival of new ways of doing things that they fail to adjust to quickly enough. The choices they make in software, and the way they use it, could have a huge effect on the future of their business. But this analogy isn’t perfect. Industrial IoT is less concerned with single devices, it has longer replacement cycles and it has to deal with a greater number of compliance requirements from specific sectors. Nonetheless, the challenges and potential for disruption are real.

The big technology, cloud and artificial intelligence players also have some learning to do, to help speed up adoption within the OT world. Today their services are mostly horizontal, and they need to find ways to package those services for use within specific vertical markets. Some companies have already made a start on this, but there’s more work to do. They also need to help OT suppliers and users to meet and demonstrate compliance with relevant requirements when using those services. We increasingly expect these players to take a vertical approach to marketing their services.

The speed at which these adjustments can be carried out is likely be the major factor dictating the pace of progress in industrial IoT from here on.