LG CNS discussed ways to apply physical AI technology at industrial sites with robotics experts from South Korea and abroad.

LG CNS said on Wednesday it recently held the "2026 Robotics & Physical AI Forum" at its Future Robotics Lab at its Magok headquarters in Seoul. About 100 people attended, including faculty members from the University of Toronto, Sogang University, Chung-Ang University and Korea University, master's and doctoral students, and LG CNS employees.

Participants discussed the research status of key physical AI technologies such as robot foundation models (RFM), intelligent robot hardware and simulation, and technical tasks that need to be addressed in the process of industrial application.

LG CNS shared its experience from more than 1,000 smart factory and smart logistics projects carried out over 20 years. It also introduced its robot learning platform, "PhysicalWorks Forge," released this year, and its AI agent-based robot operation and control platform, "PhysicalWorks Baton."

Yi Chigen (이치근), a professor at the University of Toronto, presented the development process of RFM and technologies for deploying it in real robots. He explained ways to integrate and use data from different robots, and techniques such as flow matching and action chunking to control robot movements. He also cited data 확보 and differences in movements and latency across robots as key challenges.

Lee Jun-ho (이준호), head of LG CNS' Smart Logistics City Business Division, said, "We will continue to exchange with academia to strengthen technological competitiveness and enhance business capabilities so that research results can lead to innovation at actual industrial sites."