Hmcomm Begins Equipment Monitoring for Shikoku Electric Power Using Anomaly Sound Detection AI Solution "FAST-D Monitoring Edition"

Hmcomm Co., Ltd. (Headquarters: Minato-ku, Tokyo; President and CEO: Koji Mihon, hereinafter "Hmcomm") announces the provision of its AI-powered anomaly sound detection solution "FAST-D" to Shikoku Electric Power Company, Inc. (Takamatsu, Kagawa; President: Yoshihiro Miyamoto).

"FAST-D" is an application that continuously monitors equipment operating sounds using AI, automatically creates outlier detection models, and captures early signs of anomalies. This initiative follows a joint verification conducted in fiscal year 2025, which confirmed the potential and effectiveness of detecting abnormal sounds caused by equipment issues without relying on human hearing or experience. Building on these verification results, the project aims to apply FAST-D to equipment monitoring operations while accumulating practical knowledge. Through this solution, Hmcomm will contribute to the safety and reliability of social infrastructure.

[Features of the Sound-Based Equipment Monitoring Application "FAST-D"]

FAST-D is an anomaly detection application developed independently by Hmcomm based on acoustic analysis technology. By pre-learning the normal operating sounds of equipment, the system can detect any deviations in real time during operation, enabling early identification of potential anomalies. Thanks to its AI-based self-learning capabilities, FAST-D flexibly adapts to the unique noise characteristics of each installation site, detecting subtle changes that are difficult to capture with conventional sensors or periodic inspections.

FAST-D can be operated and monitored via PCs and tablets, helping reduce the workload of on-site personnel and accelerating maintenance operations. The system also supports cloud integration and remote monitoring, making it suitable for future multi-site deployments.

[Overview of the Initiative]

This initiative aims to continuously monitor equipment related to power generation. In such equipment, early signs of failure often manifest as unusual sounds, making acoustic AI monitoring a highly effective approach. By introducing FAST-D, the burden of regular patrol inspections can be reduced, and equipment conditions can be monitored even during periods outside scheduled inspections. This advancement is expected to facilitate the realization of Condition-Based Maintenance (CBM).

[Expected Benefits]

The expected benefits of this initiative include:

• Reduced risk of equipment shutdown through early detection of anomalies

• Realization of trend monitoring and predictive maintenance through continuous data accumulation

Through this initiative, Hmcomm aims to advance infrastructure maintenance using AI, while contributing to optimal allocation of human resources, improved productivity, enhanced safety, and the establishment of smart operational systems required in the decarbonization and energy-saving era.

[Supplementary Material] Provision of Anomaly Sound Detection AI Solution to Shikoku Electric Power Company, Inc.

[About Hmcomm Co., Ltd.]

• URL: https://hmcom.co.jp

• Address: 4F, PMO Hamamatsucho III, 2-10-6 Hamamatsu-cho, Minato-ku, Tokyo

• Business: Research and development of core technologies based on artificial intelligence (AI) and speech processing, along with provision of solutions and services

[About Shikoku Electric Power Company, Inc.]

• URL: https://www.yonden.co.jp/

• Address: 2-5 Marunouchi, Takamatsu, Kagawa

• Business: Electric power services

<Contact for Press Inquiries>

Media: Hmcomm Co., Ltd. IR Department, hm_ir@hmcom.co.jp

Corporate Inquiries: Hmcomm Co., Ltd. Sales Division, sales_team@hmcom.co.jp

TEL: 03-6550-9830 FAX: 03-6550-9831