Why We Need To Stop Calling “Reality Capture” And “BIM” A “Digital Twin” On Its Own
Before I dive into this topic, I want to admit that when I first heard the term “digital twin” I was quick to associate the term with reality capture. It just felt right. The connotation seemed to work. It felt logical that our projects are “Digital” and they are a “replica of the living thing” = “Twin”. Right? Wrong!
This is a huge oversimplification of what a digital twin represents. Reality capture simply represents a single static component of the makeup of a digital twin. And often is not even part of the twin itself, as its commonly used to create the model used in the twin. Although, I would argue that this is potentially going to change and there is a real conversation to be had about using reality capture in place of a model. And a discussion around when a model is really required or when reality capture can fill the need (I will leave that topic for another blog). The BIM model itself is simply a 3d model used to support the digital twin. A digital twin is a complex living thing. It’s connected to many processes and data threads in a synchronized way. Ok let me slow down for a moment.
Don’t get me wrong, I am an evangelist for reality capture – 3d lidar, SLAM, photogrammetry, 360-imagery, subsurface, and aerial remote sensing – I love it all. Like most of the community here, I literally get excited about seeing a fly-through, a complex 3d model, or some cool application captured with technology in our field. I am quick to congratulate the hard work and efforts to produce these works of art, and I truly understand the value they bring to the community of modelers, as-builting, reverse engineering, fabrication, virtual design construction, AR/VR/XR, and all the many applications and use cases that are derived from the technology.
But I am much more cautious with the term today after working with complex projects and working in the space. Diving deep to build a college level course on the topic of Digital Twins- fundamentals, applications and techniques was also very helpful to flesh out my understanding. Interviewing and collaborating with many subject matter experts in manufacturing, nuclear, oil and gas, and complex facilities who live and breathe this space was one of the most helpful exercises in deepening my understanding and applications of digital twins.
I think you start to fully appreciate a “digital twin” when you address the manufacturing industry as this is where the term originated from – the Product Twin. But before I go there, let’s first look at a few helpful definitions the industry is providing for us to consider. They are not identical, but I think they have some very common elements that most leaders in the space would agree on. Please forgive me if I paraphrase some of the language.
“A Digital Twin” as defined by the Digital Twin Consortium:
“A digital twin is a virtual representation of real-world entities and processes, synchronized at a specified frequency and fidelity.
- Digital twin systems transform business by accelerating holistic understanding, optimal decision-making, and effective action.
- Digital twins use real-time and historical data to represent the past and present and simulate predicted futures.
- Digital twins are motivated by outcomes, tailored to use cases, powered by integration, built on data, guided by domain knowledge, and implemented in IT/OT systems.”
Thinking through this definition you can break apart many components. The 3d model and reality capture are only part of the makeup of a digital twin. But complex 3D models that link to operational processes have been around for a long time, not officially coined “digital twin” until John Vickers of NASA in a 2010 Roadmap Report.
So, what are the driving forces that are giving us this new understanding of digital twins? David Socha describes it wonderfully in his 3 blog posts on digital twins: the new big data? , digital twins 1.01, and the digital twin maturity continuum. But I would argue there is a critical dynamic missing, which is the evolution of reality capture (accuracy, speed, cost, processing power, web solutions…) and the advancement of 3d modeling (advanced model libraries, AI, rapid prototyping…) which are pushing the envelope at the same time.
When reading scholastic articles, you will quickly find many references to digital twin origin in manufacturing and aerospace industries where the “product twin” was first modeled, simulated, and both operations and maintenance built into the product lifecycle digital integration (LCDI). Lifecycle digital integration is the way to engineer the entire product life cycle to ensure the highest possible levels of performance, reliability, and safety in operating the product. The fidelity of these models is typically very high as the accuracy is needed for simulation and physics-based modeling. It must include simulation-based engineering of the process instrumentations and control system (hardware/software).
The purpose of a digital twin is lifecycle management. The purpose of a product twin is the eventual assembly of multiple product twins into a large complex product such as a Boeing Jet or NASA space shuttle, or for the modeling of products used in a system of products like a pump in a factory.
The criticality of these product twins is essential to ensure i) all parts perfectly fit together in assembling the final product and ii) we engineer optimal product operability and maintainability (credit to Amadeus Burger).
This is where digitally enabled integration of distribution product systems (DPS) allow complex systems of parts, components, and products to be built globally, shipped, and assembled accurately like Lego. The process has the highest level of repeatability, and the key value is to modularize production for multiple further products and systems.
“The digital twin of a complex physical product is a lot more than the product physical configuration in 3D; it includes the digital definition of the product operations process, and the definition of the process of making the product. This combination makes the digital twin a dynamic environment that makes possible process simulation functionality. Product graphical visualization in 3D and 2D can be created at demand to communicate the design intent as required to support product engineering and product manufacturing efforts. A digital representation of an object is not truly a digital twin unless it incorporates feedback … the two-way connection between the physical world and the digital world. A Digital Twin needs to have a distinct purpose; with understood value to be gained from it; accompanied by a continued effort to sustain that value.” – Dr. Neculai Tutos
Other great examples of digital twins come from the oil and gas, petrochemical, and nuclear space where the cost to access remote and dangerous sites, or the cost of getting it wrong can be catastrophic. In these spaces, digital replicas, combined with sensor data, and real-time and historical information are used in the critical decision-making of complex and precise processes. Some of these processes require machine learning and AI to compute or synthetic data is used to game out optimal outcomes prior to completing a task in the field. This is where a digital twin lives.
When you think of oil and gas deliverables, you need to consider dozens of 3d models, hundreds of piping and instrumentation diagrams (P&IDs), data sheets, hazards and operability, layers of protection reports, and root cause and risk-based inspections to name a few. None of these on their own are digital twins. They are data sources that feed into a living digital twin with O&M data that help us make sense of complex things, processes, assets, systems, and facilities.
Below is a very helpful model for determining digital maturity on the path to digital twin. You can see Reality Capture is a Level 0 on the path to Level 5 (Atkins Model – not the diet).
So, a digital twin is made up of many components, must have bi-directional feedback, has many potential applications, and must drive value to a specific business case aided by a virtual representation of real-world entities and processes, synchronized at a specified frequency and fidelity. The below framework of understanding and delivering digital twin projects on use case capabilities was recently published by Digital Twin Consortium and I feel its very helpful when exploring data threads to build out a specific use case for digital twins.
So, I likely confused a lot of people with this blog post, and likely irritated some people in the reality capture space, which was not my intention, but hope for others I have opened your minds to think past reality capture and BIM when using the term “Digital Twin” and understand the use case and business application for why we are deploying these complex living things to solve real world problems. The journey is just started. And now we have to deal with the MetaVerse!! This is going to hurt my brain I am sure. I will leave you with 2 great digital twin examples from NVIDIA who are providing a platform for digital twins called Omniverse. If you have not had the opportunity yet to check it out, I think you will find these interesting.
Happy to hear your thoughts in return.
Kelly Watt | CEO – Visual Plan Inc. | www.visualplan.net
Contract Instructor at Mohawk College: Digital Twins – Fundamentals, Applications and Techniques