Enabling bigger thinking with Artificial Intelligence

Enabling bigger thinking with Artificial Intelligence

Artificial Intelligence may seem exciting, but without consistent standardisation, problems will arise

Caspar Herzberg |


This article first appeared in the Autumn 2017 issue of The Record.

If you are a regular reader of The Record, you will no doubt be familiar with many of the new technologies that are permeating the enterprise business world today.

These technologies may seem somewhat glamourous, and can be used to help reimagine the way we work, make decisions and collaborate. But in the same way you can’t build a house without solid foundations, it’s important to not overlook the basic requirements for success.

OPC Foundation is the organisation behind OPC Unified Architecture (OPC UA) – the standard for industrial interoperability between different vendors, different devices and different vertical markets today. OPC UA scales from the embedded world up to the cloud world, and Microsoft has not only deeply integrated this standard into the Azure cloud platform, but is also the world’s largest open source contributor to the OPC Foundation.

One of the most exciting technologies to emerge in recent times is artificial intelligence. Artificial intelligence requires data, and involves working with this data to gain insight. OPC UA acts as the interface to the data sources companies need to gain true insight.

In this scenario, OPC UA is the enabler. Think of a USB stick. The USB part is just the enabler, but the functionality of the device is beyond the USB part itself. The same stands for OPC UA. OPC UA is the enabler that allows data to be moved left and right, but the true functionality of this data is the real value. USB is a physical connection between two devices – OPC UA is the ethernet connection between ethernet-connected devices. OPC UA delivers the meaning and semantics behind the data in a digestible format, which provides the true value companies seek.

Artificial intelligence requires easy, fully standardised access to data sources – be it from robots or other machines – to let users take advantage of all the technology has to offer.

In the days before artificial intelligence, for human beings to understand how different machines worked, we had to know everything the big handbook told us about them. We needed to understand different IT protocols, and then understand how the robots work and the commands they would have to follow. All these descriptions were different for each vendor, providing a complicated challenge.

In today’s world however, robots can have interfaces which can be easily understood by human beings. Everybody should be able to easily understand these interfaces and interact with the robots thanks to standardisation.

Let’s look a little further afield. Technologies like voice recognition, speech machines and Microsoft HoloLens are helping companies reimagine their processes today. But getting true insight from these technologies, or predicting when they may fail, is another step. Intelligent services can offer support here, but without the right levels and standards of connectivity, they cannot be utilised.

Having standards obviously therefore makes sense in this area. There’s a push and pull in the market to agree on a specific standard that meets the requirements of Industry 4.0.

But that’s exactly what OPC UA offers right now. It’s the one and only mandatory required standard for Industry 4.0. The Verband Deutscher Maschinen- und Anlagenbau – better known as the VDMA – is one of the largest European associations for machine builders and robotic companies. The VDMA understands the need for standardised, fully-secured access to certain technologies if artificial intelligence is to truly take off. Everybody will then have the same understanding of the parameters at play.

Are companies in the artificial intelligence space fully understanding this need? That’s debatable. I remember being quite shocked when I was sat in a room with 18 leaders in the robotics space at a recent event in Germany. They were talking about the parameters within which they operate, but there seemed to be different meanings in different companies. This complicates matters, and certainly doesn’t make things easy for end users. This is why there are high engineering costs in plants to integrate these robots. The good news is that companies are starting to standardise the semantics now, and agreeing on the same interfaces for data and services based on OPC UA. The results will be available at the Automatica Conference which takes place in Munich, Germany, in mid-2018.

Is your robotic supplier supporting this initiative? Ask them, as you could be connecting your robot in 20 minutes to general IT enterprise and Microsoft Azure cloud solutions.

Stefan Hoppe is global vice president of OPC Foundation

Subscribe to the Technology Record newsletter


  • ©2024 Tudor Rose. All Rights Reserved. Technology Record is published by Tudor Rose with the support and guidance of Microsoft.