[Lee Myung-ho's Future Preview] Twin Worlds: Digital Twin

Cyber-Physical Systems Linked to Real Spaces
Simulation Results Reflected in Reality
Used in Vehicle Safety Tests and Clinical Trials
Accurate Physical Property Representation Is Key
Growing Importance of Natural and Basic Sciences

The hope to live tomorrow in advance is a wish shared by everyone from young children to successful entrepreneurs. If we can know the results of certain actions beforehand and prepare for upcoming events, the chances of success increase. When predictions align with real outcomes, we gain a lot. Digital twins are opening a world where our predictions become reality.

Internet of Things (IoT) technology involves embedding small chips into objects to enable recognition, communication, and control through sensors. IoT technology is linked with big data functions that collect and analyze dynamic information generated by objects and the spaces they occupy. Inferring patterns and predicting responses through big data analysis is the essence of object intelligence. The principle is the same from AI air conditioners to autonomous vehicles. The difference lies in how many sensors collect diverse information and how many functions are coordinated in complex situations. The intelligence of objects is evolving into the intelligence of spaces where multiple intelligent objects operate. This intelligence is based on the essence, characteristics, and dynamic information of objects and spaces. This is called a cyber-physical system or digital twin.

A cyber-physical system refers to a dynamic system where physical systems existing in real space and cyber systems existing in computers are interconnected and interact with each other. Cyber-physical systems have advanced beyond intelligence that perceives and responds autonomously based on monitoring and analysis of physical objects and spaces, enabling problem-solving and optimization through simulation. Pursuing the fusion of the physical and cyber worlds, cyber-physical systems are gaining attention not only in smart factories but also in smart city fields such as transportation, safety, environment, and disaster management.

A representative example of applying cyber-physical systems in smart cities is Singapore’s ‘Virtual Singapore’ project, which began in 2014 and was completed in 2019. The entire city was virtualized in 3D to analyze and manage urban resources and services, and to simulate changes in urban facilities virtually. When new buildings are constructed, airflow is analyzed to adjust building placement and design the area for optimal ventilation. Shadow changes of buildings are analyzed to select optimal locations, and energy production can be predicted based on the installation direction of solar panels. South Korea is also building a three-dimensional spatial information system of Seoul city, but its simulation functions are still limited.

Due to cyber-physical systems, simulation accuracy is advancing to digital twin technology at a level comparable to reality. In 2014, General Electric (GE) in the United States introduced digital twin technology that digitally replicated airplane engines and connected them with data to monitor engine performance and effectively identify functional issues. A digital twin refers to a computer modeling program that digitally replicates physical objects (systems and processes). Digital twins are being developed to depict the movements and the numerical and shape results measured from actions and reactions of physical objects identically. Digital twins are used in areas such as predicting possible situations occurring in the object or informing optimal operating conditions through simulations reflecting information generated by the physical object.

Digital twin technology is also used in automobile safety testing. When crash testing cars with dummies, a new car must be used each time, but with digital twins, simulation programs can crash virtual cars into virtual walls, allowing dozens or hundreds of tests without damaging a single car. When changing car designs, virtual safety tests can be conducted to select the safest design.

As digital twin technology evolves, it becomes possible to create virtual objects identical to real ones in software and conduct simulations. This provides information on the state, functions, and operation scenarios of real objects, which can be used to improve the design or structure of the objects. It is about creating physical objects that depict digital models derived from physical objects.

Recently, the U.S. Air Force started developing a next-generation fighter jet for the first time in 20 years and conducted test flights, achieving the first flight within one year of starting the design. Although it was originally expected to take several years, the development period was shortened by utilizing digital twin and digital engineering technologies. They designed digitally, built a virtual aircraft for testing, and then produced a full-scale prototype that succeeded in test flights. Building a digital model and creating a new physical object saved development time and budget.

Attempts to build digital twins of the human body are also increasing. Recently, pharmaceutical companies have been using digital twin technology in clinical trials. They are developing systems that simulate drug reactions in the human body and conduct clinical trials on digital humans. It is expected that pharmaceutical companies will be able to develop new drugs without clinical trials in the future. If virtual humans are digitally replicated, the potential applications are limitless. If it is possible to simulate how specific diseases manifest in the body and how the body reacts to certain drugs virtually, disease prevention and treatment could be revolutionarily improved in advance.

The Korean government also announced the ‘Virtual Convergence Economy Development Strategy’ at the end of last year. The government plans to create digital twins identical to three domestic manufacturing sites in chemicals, automobiles, and shipbuilding and marine industries by next year, and implement a virtual convergence platform by linking virtual reality technology to all processes. The problem is that digital twin technology cannot advance with digital technology alone. To create digital numerical models of the physical properties and actions of real objects, numerical data of physical properties and actions must first exist. Mathematical formulas that can model physical laws must also be discovered. This is the domain of natural sciences and basic sciences. Digital twins cannot be built with digital technology alone. This is why countries strong in basic sciences lead digital twin technology. It has become an era requiring cooperation between digital technology and basic sciences.