I Signed MOU to advance Physical AI for industrial sites by combining robot operation technology and AI models
I Promoting joint development of a robot foundation model capable of performing high-risk, high-intensity tasks
POSCO DX is embarking on the joint development of Physical AI-based robot intelligence technology applicable to various robots for performing high-risk and high-intensity tasks in industrial sites, together with NC AI, an AI specialist company.
POSCO DX (CEO Shim Min-suk) signed a Memorandum of Understanding (MOU) for the implementation of an AI-based autonomous robot work system on the 29th at NC AI (CEO Lee Youn-soo)’s Pangyo headquarters, with the attendance of officials from both companies, including Yoon Suk-june, Head of POSCO DX’s Robot Automation Center, and Kim Min-jae, CTO of NC AI. We plan to proceed with the implementation of a robot foundation model that combines POSCO DX’s robot simulation and control technologies with NC AI’s AI models.
Under this agreement, POSCO DX plans to perform motion planning and control simulation verification for robots to be applied to industrial sites, and to configure and provide a digital twin- based virtual test environment to ensure stable operation of the robots.
NC AI is responsible for developing AI models for the Robot Foundation. By collecting and analyzing industrial site data to enable robots to understand various work situations, the company plans to enhance robots’ decision-making capabilities and operational accuracy by developing and applying Vision Language Action(VLA) models based on this data. Furthermore, it will collaborate with POSCO DX on establishing and stabilizing a digital twin-based VLA simulation environment.
It is expected that when the Robot Foundation model jointly developed by the two companies is applied, industrial robots will be able to simultaneously understand visual information and verbal instructions to interpret and judge work situations on their own. This means that by enabling flexible responses to various variable factors that frequently occur in industrial settings, such as deviations in workpiece position or differences in interfaces between equipment, robots can handle even non-standard tasks that were previously limited by the rule-based control of existing industrial robots.
Through this, VLA- based robots are expected to not only replace hazardous or repetitive tasks previously handled by human workers but also naturally establish a collaborative system with other industrial robots on-site, thereby significantly enhancing safety, precision, and efficiency in manufacturing environments.
POSCO DX is continuously advancing industrial robot automation technology, centered around its Robot Automation Center. By implementing simulation-based verification and establishing standardized models, the company is enhancing the stability and deployment speed of industrial robots. Furthermore, by developing a heterogeneous robot operation platform, it is strengthening operational efficiency through the integrated management of various robots and real-time data analysis. This collaboration is significant as it represents an attempt to advance physical AI technology in industrial robot automation to the next level by combining POSCO DX’s foundational technologies with NC AI’s advanced AI capabilities.
Yoon Suk-june, Head of the Robot Automation Center at POSCO DX, stated, “Through continuous collaboration with companies possessing specialized technology, we are internalizing core solutions such as robot control and operation platforms and elevating the level of automation technology for high-risk and high-intensity sites.” He added, “We expect this collaboration to confirm the potential for utilizing general-purpose robots in industrial settings.”
Kim Min-jae, CTO of NC AI, said, “General-purpose robot technology is rapidly evolving into next-generation AI technology applicable in various industrial environments.” He continued, “Through our collaboration with POSCO DX, we will strengthen our competitiveness in robot AI technology and jointly lead the general-purpose physical AI ecosystem targeting the global market.”
