The world's first robot work AI technology that can easily apply robots to manufacturing work processes has been developed. This technology can be used in various processes such as automobile and machine parts, assembly, and production, and is expected to improve the manufacturing work environment in the future.

The research team led by Kim Chang-hyun, head of the Artificial Intelligence Machinery Research Lab at the Korea Institute of Machinery and Materials (President Ryu Seok-hyun) under the Ministry of Science and ICT, announced on the 16th that it had developed robot work AI technology that can be easily applied to manufacturing work processes.

This technology is currently being applied for the first time in the world to the electronic components manufacturing process in Cheongju, North Chungcheong Province. The research team plans to gradually expand the range of manufacturers to which the process can be applied.

The Korea Institute of Machinery and Materials has developed a robotic task AI technology based on a large language model (LLM) and virtual environment for the manufacturing work process.

This technology understands the user's commands and automatically generates and executes the work commands that the robot must perform.

This technology allows for conversation between workers and robots by translating the worker's speech into the robot's language.

Until now, when introducing robots to manufacturing sites, the sites were modified to fit the robots, not the work environment or objects, and as a result, the work was limited.

In addition, there is a need to develop technologies that integrate various technologies related to work, such as movement and recognition technologies, but only some of the technologies are currently being developed.

The developed technology can automatically generate work sequences and actions based on commands given via voice or text.

Pre-learning in a virtual space can help select the optimal work location for the site, minimize the work process, and automatically recognize objects to avoid collisions.

Recently, there has been a trend to combine AI technologies such as LLM with robots and utilize them for various tasks, but testing has only been conducted in laboratory environments, not in work sites, making it difficult to apply them to actual sites.

The technology developed this time specifies the tasks to be performed by the robot, and enables the robot to easily perform tasks with minimal on-site modifications through pre-learning in a virtual space. In addition, field verification tests are being conducted, so it is expected that it will be able to effectively respond to various situations that may occur in actual work sites in the future.

Kim Chang-hyun, head of the Artificial Intelligence Machinery Research Lab at the Korea Institute of Machinery and Materials, said, “This is the first time in the world that specialized intelligence for robot work has been developed and applied to an actual manufacturing site,” adding, “We are currently testing it in the process of an electric vehicle parts manufacturer, and we will gradually expand the scope of application.”

This study was conducted with the support of the basic project of the Korea Institute of Machinery and Materials, ‘Development of core machine technology for autonomous operation and manufacturing.’