ITU-T adopts 'RDF interoperability structure and method' task

Move expected to improve data format and interface compatibility

SK Telecom's Euljiro headquarters. [SK Telecom]
SK Telecom's Euljiro headquarters. [SK Telecom]

SK Telecom is pushing into global technology standardization in the physical AI sector, with plans to extend its digital twin capabilities — built through manufacturing AI — into robotics.

SK Telecom said Thursday that a new standardization task on robot data production and utilization systems was adopted at an International Telecommunication Union Telecommunication Standardization Sector (ITU-T) international standardization meeting held in Geneva, Switzerland, from July 14 to Wednesday.

The adopted task covers the "Robot Data Factory (RDF) interoperability structure and method." SK Telecom said it proposed the task to establish a systematic framework for producing and using the data needed to train AI in intelligent robots, including humanoid robots.

ITU-T is the telecommunication standardization arm of the International Telecommunication Union (ITU), a specialized UN agency, and develops international standards in the information and communications technology field. More than 1,000 organizations, companies and research institutes from over 190 member countries participate in the ITU.

According to SK Telecom, robot motion data functions much like practice problems used to train robot AI — the smarter a robot needs to become, the more high-quality data it requires. The company said, however, that the methods for creating and exchanging such data vary by robot model and installed software, making compatibility across different systems difficult.

SK Telecom also projected that the broader robot industry's growth will drive surging demand for large-scale, high-quality motion data needed to train robot AI to pick up and move objects and learn to use tools. At the same time, it said, institutions and companies collect and manage data in different ways, making it hard to secure sufficient scale and consistent quality.

In response, SK Telecom proposed a new standardization task to ITU-T Study Group 11 (SG11) to define the roles of the key functions that make up the RDF, along with its interoperability structure and data exchange methods. The task was adopted as a new standardization item at the July meeting after review and discussion by ITU-T members.

Under the task, the RDF structure is divided into five layers: service, robot model and data, infrastructure, device, and management.

SK Telecom plans to establish the roles of each layer and the interoperability specifications between them, enabling functions to be easily connected, replaced and expanded across different robots and platforms.

The company also said it has worked with SK Hynix to build and advance a digital twin platform in the manufacturing AI sector. Building on that foundation, it plans to expand into a robot learning platform that generates robot task data for use in AI training.

SK Telecom intends to use the standardization task to strengthen the utility of its digital twin-based robot learning platform, which is being prepared for launch in the second half of the year. The company said it aims to broaden cooperation with domestic and international robotics and AI companies and research institutions through the international standardization discussions, and to contribute to the growth of the robot data ecosystem.

"The adoption of this new standardization task has laid the groundwork for SK Telecom to lead international standards discussions, drawing on the digital twin and AI capabilities we have built up in the physical AI sector," said Choi Dong-hee, head of SK Telecom's AI Strategy Planning Division. "We will continue to drive international standardization in robotics, backed by our technological competitiveness spanning the robot learning platform and AI models."


chami@heraldcorp.com