□ DGIST (President Kunwoo Lee) has been selected for a new project under the Promising Convergent Technology Pioneer Track (AI Humanoids) of the Future Pioneering Convergence Science and Technology Development Program, jointly administered by the Ministry of Science and ICT and the National Research Foundation of Korea. Through this project, DGIST will undertake the full-scale development of core technologies for “autonomous collaboration,” enabling multiple humanoid robots to autonomously allocate roles and cooperate in performing complex tasks. The project involves Senior Researcher Yongsik Jin of the AX Research Group for Robotics within the Industrial AX Innovation Division, along with Professors Jisoo Mok, Daehee Park, and Sung-Hoon Yoon of the Department of Electrical Engineering and Computer Science.
□ The project will be carried out with a total of KRW 7.75 billion (KRW 1.75 billion for Phase 1 and KRW 6.0 billion in Phase 2) in government research funding from 2026 to 2030. The research team plans to focus on developing “collective humanoid intelligence technologies” that enable different humanoid robots to share task knowledge and behavioral experience, autonomously allocate roles, and collaborate effectively in complex environments.
□ Humanoid robotics is rapidly evolving beyond the mobility and manipulation capabilities of individual robots toward a stage of “collective intelligence,” in which multiple robots can jointly make decisions and act in unstructured, human-centered environments. In particular, in real-world industrial settings such as manufacturing, logistics, equipment operation, and disaster response, there is a pressing need for autonomous collaboration technologies that enable multiple robots to divide and execute long-term, complex tasks and respond immediately to unpredictable situations such as obstacles or robot failures.
□ Through this project, DGIST will focus on developing: 1) Large Language Models (LLM)-based autonomous task allocation technologies that enable robots to understand human language and plan tasks; 2) robot control technologies based on vision-language-action (VLA) models; 3) collective environmental perception technologies that integrate data from diverse sensors to understand the surrounding environments; 4) digital twin technologies for controlling and training robots in virtual environments; and 5) self-organizing technologies that reallocate roles in response to changing conditions.
□ In this project, Pohang University of Science and Technology (POSTECH) and Seoul National University (SNU) will participate as collaborative and subcontracted research institutions, with support from leading Korean robotics and software innovation companies UNIVA, MOTOMIND, and COCELO. Based on a strong industry–academia–research collaboration framework, DGIST plans to establish a collective control platform based on a virtual models (digital twins) and a validation environment for domestically developed humanoid robots, thereby enhancing the practical applicability of the research outcomes and enabling their immediate deployment in real-world industrial settings.
□ “This project aims to move beyond humanoid robots that simply execute human commands and realize advanced autonomous collaborative intelligence in which multiple robots share their experiences and status information and autonomously coordinate their roles,” stated Yongsik Jin, Senior Researcher of the AX Research Group for Robotics at DGIST. “By fully leveraging DGIST’s convergent capabilities in AI, robotics, and control engineering, we will secure core foundational technologies and elevate the technological competitiveness of Korea’s humanoid robotics industry to the next level,” he added.
□ Meanwhile, DGIST is leading the development of foundational technologies that enable the safe operation of AI and robots, with a focus on future strategic fields such as Physical AI, which interacts with real-world physical environments, humanoid robotics, and digital twins. Through these efforts, DGIST plans to realize next-generation industrial and living environments in which humans and robots coexist and to play a pivotal role in strengthening the country’s competitiveness in strategic technologies.
