The Electronics and Telecommunications Research Institute (ETRI) has developed an AI-based Factory Energy Management System (FEMS) tailored to the process characteristics of energy-intensive industries such as bio-pharmaceuticals, food, metal and glass melting, and paper manufacturing. The research team announced that through demonstrations at 15 factories, they confirmed an energy saving rate of over 12–15%, a cumulative energy saving of 5,800 TOE, and a reduction of 18,600 tons of carbon emissions.

ETRI announced the development of a packaged FEMS platform that combines an Energy Optimization System (EOS) and an Energy Information System (EIS). The core of the platform is its ability to reduce energy consumption by collecting and analyzing manufacturing process and equipment operation data, and to coordinate the operation of processes and utility facilities.

The adoption of digital-based energy management technologies has been difficult for energy-intensive industries. This is because the bio, pharmaceutical, and food sectors must meet strict manufacturing environment standards such as GMP and HACCP, while the metal and glass melting and paper industries have maintained production structures centered on large-scale facilities and operational methods relying on on-site experience. The research team explained that as a result, improvements in energy efficiency were also limited.

The research team designed control technologies tailored to the specific process characteristics of each industry. For the bio and pharmaceutical sectors, they applied control technology combining operational data and physical models for cleanroom cooling environments; for the food sector, they developed technology to optimize steam supply for boilers and sterilization processes based on thermal energy demand forecasts. For metal and glass melting processes, they introduced technology to determine the melting state in real time using vibration data from electric induction melting furnaces, while for the paper manufacturing process, they implemented technology to simultaneously optimize steam supply and pneumatic system operation for the drying process.

The platform also includes functions for energy consumption forecasting and abnormal condition diagnosis. It supports production process operations through a large-scale language model (LLM)-based chatbot and enables stable data collection even in densely populated equipment environments by applying a low-power wide-area (LPWA) wireless sensor network. Furthermore, it has established a process-specific energy saving measurement and verification (M&V) system based on the International Performance Management Program (IPMVP), an international standard for energy performance measurement and verification.

As a result of this demonstration, ETRI presented a cumulative energy saving of 5,800 TOE and a reduction in carbon emissions of over 18,600 tons. This is equivalent to the amount of energy used by approximately 19,000 four-person households in one year. The research team plans to further advance the technology so that it can be adopted by small and medium-sized enterprises and promote its expansion to energy-intensive factories within industrial complexes nationwide.