The Korea Institute of Machinery and Materials said on Tuesday it has registered its autonomous manufacturing technology combining digital twins and artificial intelligence as an official testbed with the U.S. Digital Twin Consortium (DTC).

A research team led by Taek-min Lee (이택민), principal researcher at the Secondary Battery Equipment Research Laboratory under the Nano Convergence Research Division at the Korea Institute of Machinery and Materials, said its digital twin-based autonomous manufacturing technology testbed, MANDATE-R2R Manufacturing (Multi-Agent Network for Digital-Autonomous Twin-Driven Engineering), has been registered with the DTC's Digital Twin Testbed Program.

The testbed is an autonomous manufacturing system that combines digital twins and AI multi-agent technology for roll-to-roll process equipment used to manufacture secondary battery electrodes. The newly registered testbed links physical manufacturing facilities, or a physical twin, with a digital twin in virtual space in real time. It is structured so multiple AI agents cooperate with each other, with each responsible for data collection, preprocessing, modeling, control and maintenance functions.

The research team explained that this implements a closed-loop autonomous manufacturing system that independently recognises and predicts process conditions and autonomously optimises operating and control conditions.

Taek-min Lee said the DTC testbed registration is significant because it officially recognises the technological completeness and international credibility of digital twin-based autonomous manufacturing technology developed in South Korea. He said the team presented autonomous manufacturing in a form that can be implemented at actual manufacturing sites, where processes independently make decisions and optimise by combining AI multi-agents and digital twins. He added that the team plans to continue research with the goal of implementing a software-driven autonomous manufacturing factory that minimises human intervention by expanding the technology beyond individual equipment to manufacturing lines and factory units and advancing it into manufacturing AI foundation technology that can reuse AI agent learning results across various equipment.