Hitachi Industrial Equipment Systems Co., Ltd. has launched the “Predictive Diagnosis Service” for air compressors used as power sources for factory equipment. Specifically, the service uses machine learning to analyze data obtained through remote monitoring.

Moreover, combines it with the knowledge accumulated by Hitachi Industrial Equipment Systems maintenance staff to detect and prevent problems and abnormalities that could lead to equipment stoppages in advance. This service also uses the expertise of maintenance staff to estimate the effects of factors that reduce operating efficiency. Furthermore, they can propose more efficient operations with less environmental impact.

Background of New Service

To ensure the stable operation of air compressors, it is necessary for specialized technicians to perform maintenance and inspections. However, due to the declining birthrate and aging population, the number of technicians is decreasing. Hence, there is an urgent need for remote maintenance management and more efficient operations.

In addition, there is a growing need to reduce environmental impact by operating equipment more efficiently with lower power consumption.

Since October 2017, Hitachi Industrial Equipment Systems has offered FitLive*, an equipment monitoring service. This reduces the equipment downtime by monitoring the operating status of each product remotely and automatically sends alerts when problems occur.

Moreover, the analysis of FitLive data for air compressors has revealed that temperature-related alarms and malfunctions account for approximately 75% of all alarms and malfunctions. Therefore, we developed a predictive diagnostic service that uses machine learning to estimate the future effects of temperature rise trends detected by sensors and take preventive measures against malfunctions.

Features of the Service

This service systematizes the decision-making know-how accumulated by Hitachi Industrial Equipment Systems’ maintenance staff. Particularly, during their maintenance management work and uses it as a basis for decision making. It is necessary to perform predictive diagnostics by combining machine learning and know-how. As a result, it displays the results and the factors based on the estimation.

Previously, maintenance staff made inferences and provided consulting based on the data they had obtained. Nonetheless, this service enables maintenance staff to make more specific and effective proposals by automatically performing predictive diagnosis.
The data can also be usable to identify cases in which the performance of an air compressor tends to deteriorate. Such as when the ambient temperature of the air compressor is high or the filter needs to be cleaned, and to propose more efficient operation methods.

Currently, this function is only available for the 22/37kW inverter type oil-injected screw air compressors of the NEXT3 series. However, the number of applicable models will be expanded in the future.

“Through the expansion of this service, we will provide more detailed and speedy preventive maintenance proposals to minimize air compressor outages, reduce energy consumption, and propose operations that help reduce environmental impact,” the company said in a statement.

20 June 2024