You are Home   »   News   »   View Article

Is 'digital twin' still a useful term?

Thursday, April 21, 2022

Does the term 'digital twin' still mean something specific and useful? Or is it time to replace it with a term like 'industrial software,' since so many different types of software are called 'digital twin' these days? Cognite held an interesting debate.

What does the term digital twin mean today and is it useful? Cognite put together a panel of three industrial software experts as part of its global Industrial Digitalization conference 'Ignite Talks' to discuss this.

The participants were Jose Maria Nougues, director of technology with Inprocess Technology and Consulting, a company developing digital twins for monitoring chemical processes; Johan Krebbers, a former head of IT strategy with Shell; and Christian Møller, chief product officer with Aize, an industrial digital twin company in Norway which is a spin-off from the Aker Group. The moderator was Petteri Vainikka, VP of product marketing at Cognite.

Mr Vainikka introduced the session by saying that use of the term digital twin can be traced back to 2002, defined as the digital information about a physical object.

Any digital twin probably involves a 'stack' of digital technologies, with the data itself at the base; the data integration, models and analytics; and the data visualisation or application experience at the top, he said.

Digital twins being sold today usually focus on either the application experience, such as a 3D visualisation, or the integrated data model, he said.

A 3D visualisation can itself be a powerful data discovery tool. It is also 'really easy for humans to relate to,' he said.

Jose Maria Nougues

'Digital twin is a very powerful name, it describes a very high level concept that fits many applications,' said Jose Maria Nougues from Inprocess. 'It is very useful for marketing. If someone told you about a digital twin, you understand the concept very quickly.'

But the term may be overused today, he said, with the term used to refer to 3D models, data driven models, and simulations. Another confusing use of the term is in the telecom industry, where people talk about a 'network digital twin.'

In the same way, the term 'intelligent' was over-used with digital technology a few years ago, being used to describe anything, he said.

So when we use the term, it is important to talk about what sort of digital twin we mean.

Of the three elements of the digital twin (data, models, visualisation) Mr Nougues thinks the visualisation element is the most important.

To develop the technology further, Mr Nougues thinks that it is important to develop digital twin applications focused on specific functionalities, rather than trying to do many things at once.

There is a lot of interest in developing digital twins on a first principles (physics) model. 'This requires a lot of development. This is something that [might] start to happen in the coming years.'

An area digital twins can provide more value is in helping to diagnose what is happening, such as with a complex chemical process plant, and support decision making, he said.

Johan Krebbers

Johan Krebbers, a former head of IT strategy with Shell, said he sees that digital twins could be an important part of autonomous operations.
For example, it is very important for an autonomous car to have a sophisticated digital view of the world around it.

The same applies for autonomous systems for managing industrial facilities. This could be called 'Digital Twin Release 2,' he said.

Of the various elements of a digital twin, Mr Krebbers thinks the most important is the data integration layer. At this point operations and engineering data is brought together.

As a general point, digital twin user interfaces are not very well developed, Virtual reality is 'totally underdeveloped in this space,' he said.

To develop the technologies further, Mr Krebbers thinks it is important that digital tools can demonstrate what they can do. 'In my experience, people can only work from it if they see it,' he said.

'Select a critical workflow with a number of data sources, not too many,
implement in a small proof of concept, and then show it. Then people start to think of ways to extend it, what other data sources they can add.'

One area where digital twins can provide value is in helping improve uptime of equipment, and enabling maintenance to switch from reactive to predictive, he said.

Christian Møller

Christian Møller, chief product officer with Aize, said he thinks digital twins 'follow a certain pattern, have certain characteristics.'

'I think the name in my mind fits quite well with that pattern.'

'The term has been a little diluted because it is used a lot, it is powerful marketing, which causes a bit of confusion.'

Of the various elements of a digital twin, Mr Møller thinks the data integration layer is the 'essential piece.' Its maturity varies from company to company.

The digital twin itself could be seen as a data integration, he said. 'You have lots of data sources, the model is about trying to build relationships between them.'

'This 'model layer' is also quite immature in a lot of systems.'

It is important that people remember how important proof of concepts (POCs) are, he said. They 'have got a bit of a bad reputation. Managers say, 'we're doing too many POCs and not getting anything valid.''

'Sometimes we forget what a POC is. It is a test. [For example] can I use my phone to scan and get sub millimetre precision. It is a concept you try it out. As you get more confident that it can work - you go into the MVP [minimum viable product] stage.'

Turning your POC into something adopted by the organisation is a task often underestimated. 'Sometimes I wonder if people think, you just throw them out and people grab them, like an unhatched egg.'

The biggest value from a digital twin is that it helps companies improve safety and efficiency, like many other industrial digital technologies, he said.



Associated Companies
» Cognite
comments powered by Disqus

CREATE A FREE MEMBERSHIP

To attend our free events, receive our newsletter, and receive the free colour Digital Energy Journal.

DIGITAL ENERGY JOURNAL

Latest Edition March-April 2022
Apr 2022

Download latest and back issues

COMPANIES SUPPORTING ONE OR MORE DIGITAL ENERGY JOURNAL EVENTS INCLUDE

Learn more about supporting Digital Energy Journal