DIGITAL TWIN DOES NOT BEGIN WITH THE MODEL.
IT BEGINS WITH INFORMATION OWNERSHIP.

The problem I kept seeing

For years, I thought Digital Twin was primarily a technology challenge.

Working inside complex EPC project environments changed that view.

A project can have sophisticated engineering applications, an EDMS/CDE, 3D models, databases, dashboards and Digital Twin requirements — and still struggle to produce connected, reliable asset information.

Why? Because the Digital Twin consumes information that other people create.

• Engineering creates design information.
• Procurement connects equipment with purchase orders and vendors.
• Vendors generate equipment data and technical documents.
• Quality verifies records and conformity.
• Document Control governs formal information exchange.
• Construction creates installation and field information.
• Commissioning needs all of these information threads to converge.

If these functions create information independently, the Digital Twin team eventually inherits the gaps between them.

That led me to a principle that now shapes how I think about Digital Delivery:

Digital Twin does not begin with the model. It begins with information ownership.

The Digital Twin Information Governance Framework

I developed the following framework from my practical experience working with project information, engineering data and Digital Twin deliverables in EPC environments.

It is still Version 0.1. I am publishing it unfinished because I believe frameworks become stronger when practitioners challenge and improve them.

1. Start with the digital requirement

Every Digital Twin requirement should be translated into an information requirement before the information is created.

The question should not only be: “What does the Digital Twin team need?”

The better questions are: Who creates this information? Who owns it? At what stage is it created? Who validates it? In which system is it controlled? How will it connect to the asset later?

This changes Digital Twin from a downstream deliverable into an upstream project requirement.

2. Assign information ownership across the EPC lifecycle

Digital information cannot belong to one Digitalization team.

A simplified lifecycle looks like this:

1. 01CLIENT / PROJECT INFORMATION REQUIREMENTS
2. 02ENGINEERING
3. 03PROCUREMENT
4. 04VENDORS
5. 05QUALITY & DOCUMENT CONTROL
6. 06CONSTRUCTION
7. 07COMMISSIONING & HANDOVER
8. 08OPERATIONS

Every stage creates, modifies, validates or consumes information.

The governance model therefore needs clearly defined information owners and Data Stewards across these functions. Without ownership, the Digital Twin team becomes responsible for correcting information it did not create. That is not scalable information governance.

3. Connect the information layers

In an EPC environment, a Digital Twin is not simply a 3D model.

The information foundation may include:

The real challenge is not producing these information types individually.

The challenge is maintaining the relationships between them.

A tag appearing in structured data should correspond correctly with its engineering drawing, model object, vendor information and relevant documentation. That is where metadata and information governance become critical.

4. Govern the connections

My proposed model can be summarized as:

1. 01Digital Requirements
2. 02Information Ownership
3. 03Data Stewardship
4. 04Metadata + Tagging + Standards
5. 05Validation & Quality Controls
6. 061D + 2D + 3D + Documents
7. 07Connected Asset Information
8. 08Digital Twin
9. 09Handover + Operations + Data-Driven Decisions

Technology enables this chain. Governance holds it together.

5. Move quality upstream

One of the most expensive mistakes in Digital Delivery is discovering information problems too late.

A missing attribute during Engineering may appear small. An inconsistent tag between systems may appear manageable. A vendor document with incomplete metadata may still be accepted as a document.

But downstream systems do not understand organizational boundaries. During integration and handover, those individual gaps begin to interact.

Commissioning does not create information problems. It exposes information problems created earlier in the lifecycle.

The objective should therefore be progressive information assurance. Do not wait for final handover to ask whether the information is complete. Validate it when it is created.

6. Treat metadata as infrastructure

Metadata is sometimes treated as administrative information around a document.

I see it differently. In a Digital Twin environment, metadata becomes part of the project’s digital infrastructure.

Found → Classified → Connected → Validated → Exchanged → Reused

Poor metadata leaves technically correct information digitally disconnected.

This is why tagging, naming conventions, classification, revision/status controls and asset-document relationships should not be viewed merely as Document Control requirements. They are part of the architecture that makes project information machine-readable and reusable.

7. Build one information ecosystem

Traditional EPC organizations often operate in functional silos: Engineering owns engineering, Procurement owns procurement, Quality owns quality, Document Control owns documents, Construction owns construction, Commissioning owns commissioning.

Organizationally, that makes sense. Digitally, however, the asset does not recognize those boundaries. The same equipment tag may travel through every one of those functions.

That means Digital Delivery needs to create an information layer that connects departmental processes without removing functional accountability. The goal is not to centralize every activity. The goal is to establish shared information governance across the lifecycle.

From traditional project delivery to Digital Delivery

The operating model behind the technology

My experience has led me to think about Digital Delivery through five connected dimensions:

Remove any one of these dimensions and Digital Transformation becomes significantly harder.

Digital Transformation is rarely blocked by technology alone. It is blocked when people, processes, information ownership and accountability are not aligned.

What I would establish at the beginning of an EPC project

If I were establishing the Digital Twin information foundation from project initiation, my priorities would be:

1. 1Translate client Digital Twin requirements into discipline-level information requirements.
2. 2Establish an Information Governance operating model.
3. 3Assign Information Owners and Data Stewards across functions.
4. 4Define the project metadata and tagging framework.
5. 5Establish system-of-record responsibilities.
6. 6Map information relationships across 1D, 2D, 3D and documents.
7. 7Embed information validation into Engineering, Procurement, Vendor and Quality workflows.
8. 8Measure completeness and consistency throughout project execution.
9. 9Integrate Commissioning and Operations information requirements early.
10. 10Treat handover as a continuously prepared information product — not an end-of-project recovery exercise.

Why I am publishing this unfinished

This framework has existed mostly in my project work, observations and thinking. I had not previously turned it into something public.

While completing the #GKMBA exercise inspired by The Man Who Knew Infinity, I encountered a question that stayed with me: What is sitting in your box?

Ramanujan’s story is an extraordinary example of work whose value was not immediately understood. His story made me think about a much smaller but personally important question: what knowledge have I accumulated through experience but never made visible?

This framework is one answer. I could keep refining it privately. Instead, I am publishing Version 0.1 — not because it is finished, but because it is ready to be challenged.

My working thesis

I intend to continue developing this framework around Digital Twin, Project Information Management, Data Governance, Digital Delivery and AI-ready project information.

If you work in EPC, infrastructure, Digital Twin, Information Management, Engineering Data or Data Governance, I would value your perspective: where do you believe Digital Twin ownership should sit — and who should ultimately be accountable for the quality of the information feeding it?