[Edaily Reporter Kim Hyun-ah ] SKTelecom(017670)is embarking on global technology standardization in the field of Physical AI based on digital twin and artificial intelligence (AI) technologies. The company’s strategy is to establish international standards for the production and management of data required for robot learning, thereby securing a leading position in the robotics industry ecosystem.

SKTelecom announced on the 23rd that its “Robot Data Factory (RDF) Interoperability Structure and Methodology” was adopted as a new standardization project at the International Telecommunication Union Telecommunication Standardization Sector (ITU-T) international standardization conference held in Geneva, Switzerland, from the 14th to the 22nd.

ITU-T is the telecommunications standardization body of the International Telecommunication Union (ITU), a United Nations agency, and develops international standards in the information and communications technology (ICT) sector. Currently, approximately 190 member states and over 1,000 organizations, companies, and research institutes participate in its activities.

This standardization project defines the structure and methods for systematically producing and utilizing data necessary for AI training of intelligent robots, such as humanoids.

In the robotics industry today, securing large-scale, high-quality motion data has emerged as a key challenge for improving AI performance. However, limitations in data sharing and utilization have existed due to differences in data formats and interfaces across robot manufacturers and platforms.

To address this, SKT proposed a plan to standardize the functional roles within RDF, data exchange methods, and system integration structures. This project was officially adopted as a new standardization project following review by ITU-T Study Group 11 (SG11).

The draft standard divides the RDF structure into five layers: ▲ Services ▲ Robot Models and Data ▲ Infrastructure ▲ DEVICES ▲ Management. SKT plans to establish the roles and interoperability specifications for each layer to enable easy data integration between different robots and platforms, as well as facilitate functional expansion and replacement.

SKT cited its accumulated expertise in digital twin and AI technologies as the foundation for this standardization initiative. Building on its experience in developing a digital twin platform for manufacturing AI in collaboration with SK hynix, the company is expanding its scope to include a robot learning platform that generates robot operation data and utilizes it for AI training.

In particular, the digital twin-based robot learning platform—which generates and validates various task data occurring in real-world environments within a virtual space to enhance the efficiency of robot AI model training—is gaining attention as a core infrastructure for the widespread adoption of physical AI.

Taking this adoption of the international standardization project as an opportunity, SKT plans to enhance the usability of its digital twin-based robot learning platform—scheduled for release in the second half of the year—and expand its collaboration with domestic and international robotics and AI companies and research institutions.

Choi Dong-hee, Head of SKT’s AI Strategy Planning Division, stated, “The adoption of this new standardization project signifies that SKT has laid the groundwork to lead international standardization discussions in the field of physical AI, building on the digital twin and AI capabilities we have accumulated.” He added, “Going forward, we will secure technological competitiveness encompassing both robot learning platforms and AI models to continue leading international standardization in the field of robotics.”