Why should we connect digital twins together?

The value of individual digital twins is in making better decisions faster. But the value of

making connections between digital twins is that this can help us to understand larger complex systems better, and then to intervene more effectively. What we are  talking about here is an ecosystem of connected digital twins rather than one massive model of everything, and the connections are made by a secure information flow between them.

The Climate Resilience demonstrator, CReDO, that I helped to initiate and which is now being taken forwards by the Connected Places Catapult, is a really good example that shows the value of information flows across sector boundaries – in this case, water, energy and telecoms.

The idea behind CReDO is that climate-resilience decisions could be made better if asset owner/operators had access to more information about the other sectors on which their own assets rely.

We wanted to demonstrate, for example, that if the water network depends on the energy network, and the energy network depends on the telecoms network, then the water network could fail because something's gone wrong in the telecoms network. So, it’s about understanding those connections and transcending potential silos. This is essential to avoid  cascade failure – where climate-related shocks like flooding or extreme temperatures could make one system fail, causing another system to break down, which, in turn, leads to failure in another.

What was really interesting is that we got to a point where one of the asset owner/operators realised that it may be better for them to invest in assets in another sector in order to provide better resilience for its own customers. We need these systems-based solutions for big whole-system challenges.

Can you transfer this sort of thinking about interconnected digital twins to the oil and gas sector?

In a word, yes!  Digital twins are not just about modelling physical assets, you can also have digital twins of processes and systems. With this view, there is clearly potential for digital twins across all of the oil and gas industry, from pipeline operations through chemical processes to supply chain management and logistics.

When you see the potential use cases for all these individual digital twins across the industry, then you can imagine that it would be even more powerful if you could federate them, making data-connections between digital twins, even across organisational boundaries.

It makes sense that there could be an information flow that follows the whole end-to-end process, from well to petrol tank. Because in the same way that you've got the physical flow of hydrocarbons, you've also got an information flow that goes with it.

So, organisations should waste no time in setting up digital twins across their operations and looking for ways to join them together?

Yes. But we shouldn't expect to have a digital twin for every possible use case. It should be purpose-driven, so we should only develop digital twins where there's a clear business case. It’s the same for making connections between digital twins – if there is a purpose for it, then, great, connect them. If there isn't, then don't. This way, we envisage an ecosystem of digital twins being developed one purpose at a time.

However, in the market, we see a much higher growth rate in individual digital twins, rather than in the connections between them. This is mostly down to the challenges of interoperability.

It’s always easier to develop individual new technologies than it is to connect them. But if we want to get to a better solution, we've got to put the effort into that interoperability issue, especially where different organisations are involved. It requires hard work but promises a huge amount of value.

Of course, that’s easy to say but difficult to do, and one of the reasons why it's so hard, is because you’ve got to convince different organisations they need to collaborate on the rules to enable information flow, and that's a difficult argument to make, particularly in oil and gas where competition is so fierce.

Having said that, as so often, the oil and gas has been leading the way – for instance, the ISO15926 standard, which is all about data integration and exchange, grew out of the oil industry.

The oil and gas industry is notorious for working in silos within companies and wariness of sharing information with rivals. Doesn’t that complicate things?

Yes, but that is also precisely the case in every sector that we've looked at, so it’s not just oil and gas.

We are sure that the answer is not for everybody to put all their data into one big central pool. What makes much more sense is a distributed architecture in which everybody looks after their own data, and gives controlled access only to the data they want to share. It means sharing data standards, not sharing all the data.

So, what needs to happen to facilitate data-sharing between digital twins?

There are three basic things needed for a standard approach to data-sharing between digital twins:

First, you need a consistent approach to data modelling, because if everyone's got completely different data models, then you need to translate between them and that creates friction to information flow.

Next, you need shared reference data across the end-to-end process. So, to take a simple example, everybody needs to share an understanding of what is meant by ‘temperature’, otherwise it creates confusion and increases friction to information flow.

Then, if you are giving controlled access to controlled data, you need to have shared access, security and quality protocols.

But that's pretty much it, technically. Those three things are the key technical enablers for federating digital twins. However, of course, you also have to address the human and organisational factors, which are at least as hard as the technical solution!  These would include legal, regulatory and commercial solutions as well as addressing governance, skills and ethics.

So, it’s not about opening up all the valuable data on your company’s system to your rivals?

Definitely not. And it actually raises some really interesting possibilities about the value of data.

For example, if you have got your data that you're looking after on your side of your firewall, and I've got my data on my side of my firewall, but we've agreed that we will use consistent data models and reference data, then this means that our data is more ‘shareable’, but it does not mean we have to share it. If I request access to some of your data, then you can choose what you want to do about it – you may want me to pay for it, or you may want to do an exchange – a kind of bartering. By doing this we open up a data economy in which there is a flow of data and a flow of value (but not necessarily a flow of money).

If we were to put a genuine value on data, it would drive more rapid change, including in the key issue of interoperability and data-sharing – investors and insurers would require it, and boardrooms would demand it. It stands to reason that an organisation that manages its data well is of higher value than an equivalent organisation that does not. Being part of an effective data-sharing infrastructure is a massive source of value.

Given the challenges, how do you persuade people of the benefits of joining up digital twins?

Strangely, pretty much everybody we speak to recognises that connecting digital twins would be a very good thing, but they also say it would be really hard. The business case for individual digital twins is much easier because it comes down to making better decisions faster. The challenge is to take a step back and see the added value of federation.

One problem is that it’s often nobody's job to do it, because roles tend to be for specific projects, or specific parts of the end-to-end process. Not many people have the overview to say, actually, it would make sense if we just joined things up. There has to be a big enough vision to see the value in the first place and then the leadership to carry it through.