Korea's system integration (SI) industry is making a renewed push into digital twin technology this year. Major SI companies have already reorganized their digital twin business strategies and begun refining their technical capabilities. The technology is regaining attention as a prerequisite for implementing physical artificial intelligence (AI), which has grown increasingly important.
Digital twin refers to technology that replicates real-world objects or spaces as digital models on computers, creating a virtual "twin" of physical assets, according to SI industry sources on January 1. Korean SI companies jumped into the digital twin business following a global IT industry trend that began five to six years ago. However, manufacturers—the primary sales targets for SI firms—showed little interest, failing to see the necessity of digital twin technology. As a result, digital twin never emerged as a core business for the SI industry.
"Manufacturers have consistently shown interest in adopting digital twin technology, but actual contracts rarely materialized except with large corporations," an executive at a major SI company said.
The atmosphere has shifted recently. The physical AI boom that began last year has brought digital twin back into the spotlight. To deploy AI in traditional industrial settings such as manufacturing, companies need systems that allow AI to monitor and control all equipment lines simultaneously in real time. This process requires transferring various physical spaces and equipment into digital data—precisely where digital twin technology comes in. The synchronization process of digitizing spatial data and updating it in real time to create resources for AI utilization has become critical.
In response, Samsung SDS (018260.KS) transferred its digital twin business unit from its Solutions Division to its Cloud Services Division earlier this year. The move aims to create synergy between its Samsung Cloud Platform (SCP) and digital twin technology. Samsung SDS plans to launch an integrated service that aggregates real-time process data from clients on SCP, links it to digital twin models, and improves simulation accuracy.
LG CNS (064400.KS) is undertaking a major overhaul of its digital twin model. The company is upgrading its digital twin model to a "world model," with the update expected to be completed as early as the third quarter of this year. A world model refers to an internal model that enables AI to learn the structure and causal relationships of the world internally, allowing it to predict and simulate the future before taking actual actions. LG CNS plans to go beyond simply converting physical spaces into three-dimensional images, developing simulation functions that predict process changes over time when clients set various conditions. This would allow companies to use AI to determine optimal equipment placement before building a factory, or predict changes in production volume when altering equipment conditions at currently operating plants.
SK AX is also seeking to strengthen its digital twin business competitiveness this year. The company aims to expand by applying digital twin technology across clients' entire processes. While previous projects were limited to replicating some client equipment as virtual models, the goal now is to scale up by digitizing entire process sites. Hyundai AutoEver (307950.KS) and CJ OliveNetworks also plan to pursue digital twin projects in earnest for automotive manufacturing and logistics operations, respectively.
The resurgence of digital twin is also evident in the global IT industry. At CES 2025 in January, Nvidia CEO Jensen Huang and Siemens CEO Roland Busch announced that the two companies would collaborate to develop digital twin technology applicable across industries, drawing significant attention.
"From manufacturers' perspective, digital twin was optional in the past, but now it's an essential step for implementing physical AI," an SI industry official said.