Same or Similar? Specifying a Digital Twin

How do you respond to a client who asks for a Digital Twin as a deliverable in the Architectural Engineering, Construction, Owner Operator (AECOO) Industry? In the following blog, we set out the major aspects that need to be understood, defined, and addressed.

Firstly, what is the Client’s understanding of a Digital Twin?

Do they understand what it is that they are asking for? Is there a discrepancy between your understanding of the term and theirs? Currently, the term is being used in several totally different contexts and, as a result, there is considerable confusion as to what a Digital Twin is. The following are three possible definitions:

Digital Twin: An exact geometric and visual representation of physical reality, a building, site, or town.
Digital Twin: A virtual geometric and data-rich representation of physical reality, which links or shares data to drive downstream processes.
Digital Twin: A virtual representation capturing the physics of structures and potentially changing internal and external conditions, as measured by myriad connected sensors driven by cutting-edge computing. A digital twin can also run simulations within the virtualizations to test for problems and seek improvements through service updates.

In addition to the confusion about the term digital twin, there is also the issue as to whether the twin is ‘identical’ or ‘fraternal’. While this might seem a minor distinction, it is hugely consequential. The difference in cost and time involved in generating a ‘fraternal’ rather than an ‘identical’ twin is huge. It is critical to determine where the value lies in what you are requesting. Does the digital twin need to be highly granular, down to the nuts and bolts, or is an approximation of the overall asset sufficient? Does the twin need to be a mirror image of reality? What accuracy is actually required in terms of size and placement to enable the performance of downstream functions? Does the data transfer and feedback between digital and reality need to be automatic? These are some of the issues that will determine whether your digital twin is an identical replica or, like fraternal twins, different but closely related and linked.

Planning for a Digital Twin:

Assuming that your client is interested in a digital twin, then, based on the above, the first question is Why?

WHY? – Why a Digital Twin?

It is vital to consider the end at the beginning; i.e. one must understand what it is that the data and geometry is required to do. What system does it need to communicate with? What outputs are required? What visualization is needed? Does the geometry need to be a mirror image, matching reality, or merely a symbolic representation?

Once a common understanding of all the desired goals and uses for the geometry and data has been established, one can move on to a very important discussion point: Has the Client understood the consequences of what is being asked for?

More specifically, has consideration been given to the work effort required, to achieve the stated goals and uses? Is there an understanding of the potential increased costs and time to achieve this outcome? Has a cost-benefit analysis occurred? Has the maturity of the industry been considered, along with the potential risks associated with the achievement of the stated goals?

Once one has a clear understanding of why a digital twin is being requested, the next step is establishing exactly ‘What’ is required.

WHAT? – Defining the Digital Twin.

Using the ISO 19650 framework will help organize and assign scope and outputs, leaving design and construction teams to determine who will be responsible and record how they will achieve what has been asked for in the BIM Execution Plan (BEP).

The specific goals and uses that are required to achieve the stated goals need to be clearly articulated. What data fields are required? What processes are required to be undertaken? In other words, what are the goals and uses for the geometry and data? What is the Level of Information Need for all required components of your digital twin? (Note, this is not LOD.) What is the data format required as the final deliverable? Will the Client be able to access the data easily or need a team of skilled staff to operate the software? Who will be responsible for linking/associating design intent and installed asset information? Is it to be held in a common data environment? Who is to host the environment? Is it aligned with international standards? When must the required data and geometry be available within the datasets to drive the required processes?

Have you generated a robust document that can be used to define precisely what is required? Are the required quality control processes in place to audit the data set to ensure compliance and reduce risk? Do you have any clauses to manage non-compliance of the requirements?

With a clearly defined ‘What’ objective, the next question is ‘How’ the digital twin is to be delivered.

HOW? – Delivering the Digital Twin (including the importance of the Client’s role)

Clients need to define their outcomes, per ISO 19650, the OIR, PIR, and AIR, and allow the selected teams, Design and Construction, to define how they will achieve their goals. In addition, to reduce the risk that the final delivery will not meet their needs, there is a need to assign an Information Manager as part of the team, to monitor the process and undertake data quality checks.

Things to consider:

Do the BIM Requirements address the consequences of non-compliance?
Has the BIM Execution Plan (BEP) been reviewed to check that the proposed strategies, processes, and workflows will generate the geometry and data required for the Client’s definition of a digital twin and the downstream goals and uses?
Do the proponents have the required BIM maturity? Do they have experience in building robust data sets to drive defined downstream goals and uses?
Can they provide examples of previous projects where a digital twin was a requirement?

Defining Quality.

It is the Design and Construction Teams that are generating the underlying data and geometry that will be the foundation of the digital twin. To mitigate risk, it is vital that data audits are undertaken to ensure that any potential problems are addressed in a timely manner. Like knitting, if a problem is not addressed early, resolution becomes progressively more and more problematic. We address this issue in a recent blog, Data Quality: What is it and why is it important?

Conclusion

Most often, in the AECOO industry world, the above process will reveal that what the Client requires is a fraternal, rather than identical, digital twin. The generation of a fraternal digital twin is not difficult, it does however require additional work effort and an explicit specification of what the expectation is along with clearly defined goals and uses, so that everyone understands what it is that they are being asked to undertake.

The bottom line is the teams are being asked to build a robust data set rather than just a 2D document set. That is an additional work effort requiring additional time and cost. However, what is most important is the value to the Client downstream for the life cycle of the facility. Conventional wisdom is that, in lifecycle costing, excluding the cost of staff, Design = 2%, Construction = 18%, and FMO = 80%. Based on this, design time is cheap time, and it is here that effort should be focused.

The collection and uploading of installed asset information, as mentioned earlier, is associated with the design information. When undertaken in a common data environment (CDE), this is not a difficult or time-consuming task. Summit BIM utilizes a CDE BIMFMi^© to manage the data upload. Once the data is extracted and in a CDE, robust, organized, and future-proofed, it is available for all sorts of potential future uses not yet envisioned.

We are here to help you. If you would like to learn more, please reach out to us.