Digital Twins and AEC
Article by yours truly, originally published under the title Urban digital twins, in DomusAIR N.5 | 2022 - Digital Connections, Go beyond physical infrastructure, pp. 66-72.
Se volete leggere l'articolo in italiano, fate clic qui
Digitalization is an epochal phenomenon, the structural and physiological process of contemporary, the salient anthropological trait of today’s globalized population. The digital permeates the social realm in every sector, and some of the most interesting new things that have been generated in this process are what are known as “digital twins”.
As they say in Latin, nomina sunt consequentia rerum (names are the consequences of things), and the name “digital twin” is already fairly suggestive. It is essentially a digital alter ego of some real object, which can be a room, a building, a piece of infrastructure, or even an entire neighborhood or city. Digital twins serve the purposes of their physical counterparts, helping us better understand, predict, and optimize performance through the use of mathematical models, similar to those used in physics, engineering, economics, weather forecasting, etc. In technical terms, digital twins are computer models that receive and process data, a cataloged series of specific analysis techniques (Data/AI-driven development, predictive analytics, etc.), and a set of instruments that are capable of increasing and updating the knowledge base of the digital twin over time (AI, IoT, sensor systems, machine learning, etc.). These elements can then be incorporated into a Common Data Environment (CDE).
The concept of digital twins first emerged in the 1970s with the advent of computers capable of modeling digital objects, but it would be more than 40 years before technology evolved to the point of allowing for practical applications. Some of the earliest examples of digital twins were seen in the consumer products manufacturing industry in 2015 but they soon expanded to other sectors, including construction and infrastructure. Although the concept spread is relatively rapid and welcomed by construction operators, the technology has not developed as quickly, meaning that digital twins are emerging tools still in the definition and comprehension phase.
Conceptually, we may think of a digital twin as an evolutionary progression of computer models used in BIM and/or GIS processes, although, according to current standards (UNI EN ISO 19650-1:2019), there is theoretically no distinction between a computer model and a digital twin. This technology brings virtually unlimited potential. One of the hottest research topics, especially in the post-pandemic period, is the creation of digital twins of entire cities as urban planning tools for crafting the smart cities of the future. Digital twins of urban areas are theorized as entities able to incorporate and process the enormous amounts of data and information that can be accumulated in a city (regarding buildings, various networks, buried utilities, traffic circulation, transportation, how people use space, etc.) in order to produce, using such techniques as predictive analytics, advanced simulations of all possible direct or indirect outcomes of one or more planned or plannable alterations of the existing urban environment (e.g., street closure, new metro section, entire new district).
Another area of application for digital twins is the multidisciplinary management of large-scale infrastructure such as airports, train stations, and ports. In this case, the digital twin consists primarily of a computer model that represents the infrastructure in a highly accurate and complete way, coupled with a database containing information about its assets that is updated in real-time via full-coverage sensor systems, complemented by continual acquisition of data by instruments used by operators on the ground and sent to the CDE. In this case, the digital twin tends to have a predominantly management role, helping the facility manager in daily operations and in identifying future issues, thus facilitating predictive maintenance operations.
However, as often happens when technological terms come into vogue, the term digital twin is now often abused or poorly interpreted. Quite frequently, two-dimensional drawings, three-dimensional models, computer models not keyed to the world and thus frozen in time, or even Enterprise Resource Planning (ERP) applications are presented as digital twins. However, a digital twin cannot be constructed out of 2D PDFs. ERP applications are certainly based on large databases but they include only digital information of known building conditions, information that requires constant human intervention for validation, maintenance, and updating. Currently, most ERP systems have only basic functionalities for monitoring assets, tracking performance, and planning operation and maintenance. They also require high levels of customization to meet the needs of the owners and are relatively costly to operate and maintain. In any case, this type of digital representation is not a digital twin.
While still relatively few, there are praiseworthy examples of digital twins already in use for infrastructure or urban applications. The Amsterdam Airport Schiphol is one and the San Francisco International Airport is another. But the subject is currently witnessing great ferment and very interesting projects are on their way, such as the creation of a digital twin for the Los Angeles International Airport. As for urban areas, a CNR team led by Giordana Castelli is currently developing digital twins for the cities of Matera and Catania.