Understanding the Digital Twin in AEC

Meaning and Understanding

The interesting thing about words, is that they have “meaning”, and can convey and carry complex concepts of understanding, very quickly, as we communicate with each other. They are like “codes” of communication. Of course, that doesn’t mean that everybody has the same understanding of a word. It depends on your maturity of understanding of the topic or domain, and also the context in which the word is used. Simply uttering a “word”, doesn’t necessarily mean the other person receives the intended message. Take for example simple, and well-used words like “cat” or “dog”. Most people would have a clear understanding of those words. They typically describe a type of 4-legged animal, with fur, a tail, protruding ears, which are, or can be, domesticated as a household pet. In that sense, they are similar, but they are also different enough to warrant a separate “word” or category to convey “meaning”. Of course, in the context of a discussion about wild animals, “cats” could refer to lions and tigers, and “dogs” could refer to wolves and foxes. So knowing the “context” is important. Also, a 3-year-old, who is less mature in their understanding, may incorrectly use the wrong word, when they see a domesticated furry animal with 4 legs, tail and protruding ears. They are unlikely to say, “...look mommy, I see a domesticated furry animal with 4 legs, tail and protruding ears, but I’m not sure if it’s a cat or a dog…”. They are more likely to utter something like “ca..” “do..”, but mommy will see what they are pointing at (the context), at get their meaning.

 So, why am I talking about “words”, “meaning”, “maturity of understanding” and “context”? Well, when you come across a new word or phrase, like “Digital Twin”, then you have to be sure that you, and the other people that you are trying to communicate with, have the same understanding of the complex concepts that the word or phrase is trying to encapsulate and convey, otherwise, there is going to be “mis-understanding” and confusion, which will degenerate into poor communication, poor cooperation or collaboration, and ultimately poor results or outcomes. And “digital twin” is a relatively new word or phrase. It seems it was first anticipated by David Gelernter in 1991, in his book “Mirror Worlds”, and the concepts were first applied by Michael Grieves of Florida Institute of Technology in 2002 in manufacturing, as a conceptual model underlying product lifecycle management (PLM). The term “digital twin” was used in a report in 2010 by John Vickers of NASA.

 What is the Digital Twin?

The “digital twin” concept consists of three distinct parts: (1) the physical product, (2) the digital/virtual product, and (3) connections between the two products. The “connections” between the physical product and the digital/virtual product, is “data” (digital information) that flows from the physical product to the digital/virtual product, and information that is available from the digital/virtual product to the physical environment (ie a 2-way flow of information). This flow of information can be supported by technologies such as sensors, IoT devices, and other data-capture and processing methods. The “digital twin” concept was later expanded to include the “prototype”, “instance” and “aggregate”. The “prototype” is all the designs, analyses, and processes required to make or realize a physical product in the first place (ie information that exists before the “physical” product exists). The “instance” is the information about the product once it had been manufactured. And the “aggregate” is all the subsequent information that is gathered, used or produced during the ongoing use of the product through interrogation, analytics, and learning etc. These phases broadly equate to what we call “design”, “construction” and “operations” in AEC.

 When it comes to applying these “digital twin” concepts to the AEC sector, of course, things get a little more complicated. We are no longer talking about a single manufacturer, producing and monitoring a single “product”, in a neatly organized and controlled environment. “Buildings” and “Infrastructure” are complex things. They are vast collections of many materials and products, planned, designed, procured, assembled, and operated by multiple parties, throughout the extended lifecycle of these built assets. This high level of “fragmentation” of the supply chain, makes the collection of data (digital information) equally “complex” (or mostly chaotic, in my own experience to date).

 WHY “Digital”?

The first question that really needs to be answered, is “WHY?” would you want a “digital twin” of a building or infrastructure asset in the first place? (and maybe “who” wants this?). If you don’t know “why”, there is no incentive to go forward, or do anything different. You can only answer this question, if you understand the “value” of information. Unfortunately, the AEC sector has been dealing with “poor quality information” for a very long time, both in creating built assets, and operating or maintaining those assets, and consequently the “players” in the AEC sector don’t put a high “value” on information. The industry has learned to work around the issue of “poor quality information”, in order to still “get the job done”, no matter what the cost, in time, money, safety, or impact on the environment. The way the industry is structured, the business models and incentives, are all based on the presumption of “poor quality information”, and passing the risk associated with “poor quality information” down the supply chain. Nevertheless, we still “get the job done”. It is expensive, slow, stressful, dangerous, not very profitable, but it works. So, what is the “value” of information? Why pursue better quality information? “Why” change anything? As they say, “…if it isn’t broken, then why fix it..”. Well, it is “questionable” as to whether it is broken or not. Yes, buildings and infrastructure get delivered, and people make a living (or existence) in the sector, but is it meeting the needs or demands of society? Even in Ireland, where we would consider ourselves a first-world country, the AEC sector can’t deliver on the needs of society, in terms of housing, healthcare, education, infrastructure, etc. Building is too slow, too expensive, the working conditions in AEC are stressful and dangerous. And, the impact of AEC on the environment is enormous (AEC contribute over 40% to energy usage, CO2 emissions, and use of raw materials). The truth is we need “better quality information” to help us make significant improvements in how the AEC sector works, and by “better quality”, I mean more accurate, more timely, more accessible, more organized information. “Digital” provides the way to achieve “better quality information”, and opens the door to better results or outcomes, better performance, more efficiencies and productivity, less time and cost, less delays and disruption, better and safer working environments, better understanding of the impacts on the environment. Yes, we can continue to work with “poor quality information”, as we have become accustomed to doing, but the “results” will remain the same. (as Einstein said “…insanity is doing the same thing over and over again and expecting different results…”).

 WHY “Twin”?

While “digital” opens the door, and provides the potential for significant improvement in performance and outcomes, it doesn’t by itself guarantee improvement. “Digital” information can also be poor quality, incorrect, late, inaccessible, and disorganized, and can still hinder or obstruct performance and results. This is where the “Twin” concept becomes really important. The information must be an accurate representation of the physical product (appropriate to it’s time in the lifecycle of planning, design, construction or operations), there to support all the activities and decision making during the lifecycle of the physical asset. This information must be made available and accessible to the people (and machines) who need it, and it must be carefully managed, organized, and maintained over time, so that it doesn’t deteriorate, deplete, or become “poor quality information” in the future. The “Twin” needs continuous effort, attention and care, and in return, it will provide a useful, accurate, timely accessible resource of information to support all the activities associated with planning, designing, constructing and operating building and infrastructure assets. But, if ignored, the “value” of the “Twin” will continually diminish over time. (like all things, left to their own devices, will deteriorate and die). Of course, if you don’t understand the “value” of information, as discussed above, then you are not going to see the reason to invest time, effort and attention into developing and maintaining the “Twin”, and you might simply decide to focus on the maintenance of the “physical”. As if it was easy to operate and maintain a physical building or infrastructure asset without information? This attitude demonstrates a poor level of maturity or understanding of the relationship between information and performance outcomes. It is probably worth noting, that the “twin” is not an “identical twin”, because the digital model is a “representation”, not a “replica” of the physical. The “physical” and the “digital” serve different purposes, and have different characteristics depending on the needs they serve.

 What is the “Digital Twin” for AEC?

The “digital twin” is a set (or collection) of data (digital information) that represents a physical building of infrastructure asset. But, since physical buildings and infrastructure assets are in themselves, complex entities or collections of many systems, materials and products, it follows that the “digital twin” will also be a complex set of data, representing all those systems, material and products that make up the physical asset. What is important is the “relationship” of all those system, products and materials, and how the come together, and function together. There are “physical” relationships – components are located in space, and have a “physical” presence (their size, shape, etc). There are “functional” relationships – components work or act together as part of a system (eg. a fan connected to a ventilation duct). Some of this information, is best represented “graphically” (ie to show the size and shape of a component in 3D space), and help humans “visualize” components in the context of other components, and some of this information is best represented “non-graphically”, or as digital properties and attributes of information, that computers/software can act upon (eg. adding all the outputs of all the fans in a ventilation system, to calculate the air flow and exchange to each space, or reading the temperature of a sensor etc). Of course, the AEC sector is also highly regulated, with laws and contracts that require certain information needs to be provided as record “documentation” to demonstrate compliance with regulations (eg. certificates, reports, technical assessments, installation and operating instructions, etc). So you could say, for every physical component of a complex building or infrastructure asset, there will be 3 types of data: “Graphical Data”, “Non-Graphical Data” and “Documents”. Of course, all this data is being produced by many different parties and people at different times. Some during the “prototype” phase (planning and design), some during the “instance” phase (construction, fabrication and assembly), and some during the “aggregate” phase (ongoing operations and maintenance). And all this information has to be brought together in an organized and carefully managed way, where it can be continually maintained, and made available, and accessible to other people (and machines) who need it. This is where the “Common Data Environment”, or CDE, comes into play, as described in ISO19650, an international standard for “organization and digitization of information about buildings and civil engineering works, including building information modelling (BIM)”. It is not anticipated that the CDE will be a single “system”, but rather a inter-connected group of systems that are able to share the relevant (common) information as digital data.

 Where Next? (How)

This discussion raises some interesting questions, like, can a “digital twin” exist, without a Common Data Environment (CDE), as described in ISO19650? And, if BIM (building information modelling), is the technological means by which we digitally represent the “graphical” and some of the “non-graphical” information about buildings and infrastructure, showing the “spatial” (3D) and functional relationships of all the different components, then can a “digital twin” exist without BIM? Of course, there are many people, without a good understanding, or capability and skill in BIM, information management or CDE’s, that will argue that the “digital twin” can exist without those. But when you hear some of those arguments, it is a little bit like hearing a 3-year-old trying to describe a “domesticated furry animal with 4 legs, tail and protruding ears”, when they don’t have the vocabulary, maturity of understanding, or full context. I’ve been involved in the world of Architecture for over 30 years, and in the world of BIM for over 12 years now, and I’m still astounded by the lack of maturity of understanding amongst many AEC sector professionals, as to the “value” of information, and the function of BIM, information management and CDE’s. So I’m sure we will be fully entertained by the ongoing discussion on “digital twins” for the next decade.