IHI is pleased to announce that Kobelco Power Kobe No. 2, Inc. has begun full-fledged use of an anomaly diagnosis system that uses the MEDICUS NAVI^*1 operation and maintenance support system at the No. 3 and No. 4 units of its Kobe Power Plant (power output: 650 MW x 2 units).

MEDICUS NAVI is an operation and maintenance support system for power plants developed by IHI. Since FY2020, IHI has been providing Kobe Power Plant’s No. 3 and No. 4 units with remote monitoring system for the real-time monitoring of power plant operating data. This anomaly diagnosis system was added to the existing system in FY2023. Then, in FY2025, after performing verification testing, IHI added a steam leakage detection function and an operation optimization algorithm to the anomaly diagnosis system. This newly expanded system has now been put into full-fledged operation.

The anomaly diagnosis system is based on I-ASTA, an anomaly detection model that uses IHI’s proprietary IHI-MT system^*2. The anomaly diagnosis system’s steam leakage detection function leverages a machine learning model. With these functions, the system is now able to detect anomalies with a high level of accuracy based on operating data from different equipment. Furthermore, by introducing an operation optimization algorithm, the system can not only detect anomalies but also analyze and visualize phenomena that affect boiler efficiency. System users can assess operating conditions and efficiently consider improvements. This contributes to stable power plant operation and operational advances.

IHI will continue to enrich and expand its anomaly diagnosis offerings and improve diagnosis accuracy. It will not stop with providing operation support for thermal power plants, but will contribute to the refinement of customer plant operations by making clear, visible improvements.

*1 For more details on MEDICUS NAVI, please visit:
https://www.ihi.co.jp/technology/techinfo/contents_no/1198960_13491.html

*2 IHI-MT system:
This unique IHI technology utilizes the Mahalanobis-Taguchi System (MT System), a widely used statistical technique, and also takes into consideration power plant output fluctuations, changes in external conditions, and other factors when diagnosing equipment condition. It reduces false positives and makes it possible to identify the signs of facility issues at an early stage.
https://www.ihi.co.jp/technology/techinfo/contents_no/1201201_13491.html

Related press release:
IHI Receives JSAI Annual Conference Award at the 2025 Annual Conference of the Japanese Society for Artificial Intelligence for Technology That Improves Machine Learning Accuracy for Plants and Production Facilities
https://www.ihi.co.jp/all_news/2025/technology/1201763_13760.html

News Top