Semantic interoperability opens doors to the digitalized world
A universal IoT standard is yet in the distant future. Dr. Richard Soley, the Executive Director of the Industrial Internet Consortium treats the idea with skepticism. “I’ll be glad to be the first one to say that there will never be a single universal standard,” he states. According to his view, achieving interoperability is a much more important issue to tackle.
“No matter how hard we try, there will not be one standard, because people have different requirements and are trying different things and experimenting in different ways,” Soley argues.
He divides the world of IoT into two stages. The first one is middleware, which is about the input and output of bits and bytes; about moving information around. Numerous standards exist in this stage, and even though they may differ, no standard is necessarily better than the other.
In the second stage the main challenge is capturing the semantics of the different systems. “The question about the semantic representation is actually much harder. Now that you have moved those bits and bytes from here to there, what is it that they’re telling you to do?”
Jet engines, for example, all work the same way but differ from each other in the ways and units in which they capture data. The ability to translate that data – semantic interoperability – is a prerequisite for exchanging information between different machines.
“Achieving interoperability is our main task in the face of that differentiation and diversity between hundreds or even thousands of standards. How to bridge those standards is where we need to focus,” Soley says.
Big data analysis reshapes operations
Increasing efficiency is the most obvious opportunity created by digitalization. Adding sensors to different systems enables access to more detailed information about current processes, which usually results in improving them. The best opportunity, however, lies in finding completely new business models.
“When people buy air compressors, it’s not the machine they want – it’s compressed air. So it makes sense for the manufacturer not to sell the ownership but the outcome. They still install the compressor but maintain control over it and make sure it’s kept in perfect condition keeping the customer satisfied. In this so-called outcome economy, consumers get to buy what their hearts desire rather than machines that make them what they want,” Soley describes.
“Achieving interoperability is our main task in the face of that differentiation and diversity between hundreds or even thousands of standards. How to bridge those standards is where we need to focus.”
Another great opportunity in the world of big data analysis is discovering unexpected correlations that haven’t even been considered before. Furthermore, revealing hidden correlations can help reshape operations in a very concrete manner:
“In Southern Ireland, in one of our smart city testbeds the county’s ambulances are linked to national health data information, but in the future might also be moved to wherever the likelihood of an accident is the highest at the given moment. That’s something you can only do with big data analysis,” Soley says.
The way is paved in testbeds
With digitalization, as with all megatrends, come both opportunities and challenges.
“I think the biggest technical stumbling block at the moment is semantic interoperability especially in terms of analytics. In the past five years we’ve made a lot of progress but there’s still a lot of work left,” Soley points out, and continues: “Starting to create standards at an early stage where you haven’t even actually decided what it is that you’re going to build and which parts need to work together is really kind of crazy. In my opinion, a more fruitful approach is to build testbeds to learn about technologies’ best practices, what kind of analytics is required, how to build interoperability, and so on.”
To some extent, interoperability is sort of what applies to utilizing human resources as well. Fundamentally, harnessing digitalization is a collision between information technology and operational technology. In most companies it means combining forces of people with different kind of skills and competencies.
“In manufacturing companies, CIOs have to be open to finding new partnerships in terms of operational expertise to get a full understanding of how does it all fit together,” Soley says.
Nevertheless, Soley’s most important advice to CIOs as well as CEOs is to start small and with enough patience: “Don’t go and try to boil the entire ocean at once. Choose an application, start collecting sensor data, find out what it means and how it’s going to affect your operations. That’s what testbeds are for. Once you’ve done that, you can start building out across your operation and business, integrating information from unexpected sources and finding unexpected opportunities for your business.”
Trying to solve every possible issue at once is unlikely to pay off. Instead it may lead to a situation that Soley refers to as “analysis paralysis where you’ll try to analyze every single thing you do”.
Ultimately, just getting started is what matters. “For the CEO of the company I would say that digitalization is going to transform everything from business and manufacturing to transportation and healthcare – and this includes your industry too,” he concludes.
Dr. Richard Soley works as the Executive Director of the Industrial Internet Consortium, an international organization developing Industrial IoT testbeds; and CEO of the Object Management Group, a developer of semantic standards for industrial IoT.
Image credit: Christian Lagerek / Shutterstock.com