ISO and Digital Twin Definitions

[Michael Mulquin]

In IEC Smart Cities Systems Committee we are working on a definition from our perspective. Our draft is:

Digital Twin

formal, explicit, computer-readable and computer-executable digital representation of an object or system that allows a computer to understand that object or system

Note 1

Digital twins can be made of material objects and systems (house, city, human, electrical distribution system etc.) and of immaterial objects (process, project plan, etc.)

Note 2

Digital twins can be primary, when the digital representation is developed before the physical object or system is built to help in the design or construction of that object or system. Or they can be secondary, when they are developed to represent an object or system that already exists in the physical world.

Note 3

A digital twin of an existing object or system is created from data gathered from Internet of Things (IoT) sensors, with scientific and social-science modelling, machine learning and other advanced analytics applied to replicate the behaviour of that object or system, supporting decision making.

Note 4

A digital twin of an existing object or system is more than just a model: it must have a two-way connection with that object or system and should provide dynamic as well as static representations of that object or system. Data needs to flow between the physical twin and the digital, enabling insights and creating the opportunity for positive interventions within the physical twin.

Clearly Note 4 has some “should”s in it, so we need to correct this!

[Timothy Ståhle]

@DRossiter87, thank you for this post. Your explanation of the requirements for a definition to meet the ISO criteria are very helpful. I would like to approach this topic more from the perspective of the business use case and the purpose of the digital twin. Although the ISO/TS 18101:2019 captures certain aspects of the digital twin, I would argue that it doesn’t fully capture what a digital twin is. I think the definition of the digital twin should be driven by the purpose one wants to achieve with it. One could argue that a “digital asset on which services can be performed that provide value to an organization” is a CAD or a BIM-file (digital asset) that is used to export drawings in PDF (services) is a digital twin. From my point of view a BIM in and of itself is not a digital twin. However, the data from BIM can used to create a digital twin. As an asset owner I often have discussions with architects about digital twins and design engineers often confuse the concept of a digital twin with BIM.

• I would say that the power of the digital twin is the ability for the digital representation of the physical object to connect to it’s physical environment through a function of input and/or outputs e.g. by geographically referencing the model and using meteorological data to perform simulations.
• In the operational phase, when the asset owner receives the AIM (according to ISO 19650), the AIM is not yet a digital twin. It’s only when the asset owner connects the digital representation of the object to its physical environment that it actually becomes a digital twin. That could be achieved by e.g. connecting the AIM to an IoT-platform using data from sensors, booking systems etc.

With these criteria in mind I would propose the following definition “A digital representation of a physical object and its interactions with the environment that creates valuable information.”

What do you think?