Digital thread adoption in the EPC sector

Paul Connell, senior strategy and enablement consulting lead at Octave, who has previously worked in global projects and development at ExxonMobil, discusses the data visibility challenges and opportunities for today’s EPC organizations:

What’s the state of play with digital thread adoption in the EPC sector today?

I discussed this very topic with several EPC organizations and owner operators in the US recently. Without a doubt, everyone wants this capability, but they’re also feeling that it’s a monumental effort to implement when there are already long-established processes in place to execute projects. When you have projects that are ready to be executed and you have constrained resources, it can become difficult to find the moment to move digital thread maturity along.

Where are most EPC companies today in regard to IDC’s definition of the three maturity stages for digital thread capability?

The majority are in that middle “moderate maturity” group. At some organizations, there are also variations in capability within projects. For example, design and engineering might be pushing forward toward mature digital thread capabilities. Further down the project lifecycle, however, teams may be utilizing some fully cloud, totally SaaS capabilities, but with significant gaps in the processes and workflows where they end up supplementing with other point solutions to fill in. There are also still a few organizations at the “low maturity” end, where there’s a lot of manual data handling, Excel spreadsheets and so on.

What benefits do organizations see in making the effort to increase maturity?

It can fix existing problems and enable many new things. One of the key issues that it would help mitigate is the validation of data coming out of engineering before it gets downstream. If you have the ability to connect all data through a centralized pipeline, if you have a kind of birthmark on each piece of data that allows it to be tracked through its lifecycle, then that becomes incredibly powerful. When you’ve gone through the pains that can be experienced in projects when you have bad data, that becomes a real motivation to do something about it.

What are the most common internal barriers to transformation?

Project teams who want to develop digital thread capabilities will often face resistance from a much higher level in the organization, such as “we can’t invest the people and the time right now,” “we don’t have the budget,” “corporate has standardized on technology X,” and so on.

In other organizations, they might commit a large group of their IT professionals to actually develop a custom software solution for data connectivity, which can work initially, but then often hits scalability and reliability issues – then they backslide from moderate maturity back to the low maturity stage.

Does industry-wide collaboration have a part to play in advancing maturity?

Absolutely. Talking to those EPCs and owners in the US recently, they were all concerned about that data validation issue. For example, how do you ensure that the data you’re producing during engineering is controlled and validated, so when it leaves one of the authoring tools and is disseminated through other systems and programs, it remains true and correct and there is only one source of truth? There’s an imperative here for the industry to come together and have an open and honest discussion on what goodlooks like – for instance, “if we can all align on what this type of pump is called, across the world, then we’ll always have the taxonomy right.”

Is this somewhere that AI can also help with?

Sure. If you train an AI agent on a specific type of taxonomy or a specific project data naming convention or structure, you could then deploy it to carry out clash detection – for example, identifying and correcting naming anomalies such as extra spaces, too many numbers and so on. It has the potential to significantly speed up the harmonization of data in support of a digital thread.