Contents I. INTRODUCTION
Supervisory Control and Data Acquisition (SCADA) technology allows facility operators and managers to understand the overall operation status and health of critical equipment. However, facilities and machines relying only on monitoring and SCADA technology usually operate in a reactionary mode – responding to failures and catastrophic power losses, rather than predicting and preventing them from occurring in the first place. This is reactive maintenance. It is easy to understand that with reactive maintenance, operation efficiency could be seriously affected if critical facilities fail at critical periods. The other type of maintenance is called scheduled or preventive maintenance, which shifts plant downtime to non-critical periods. Since the preventive maintenance does not take the actual machine health into consideration, unnecessary machine shutdown might be inevitable, which may incur unnecessary cost. In view of the above, predictive maintenance has been actively pursued in the manufacturing industry in recent years, where equipment outages are predicted and maintenance is carried out only when necessary. Benefits of predictive maintenance include better utilization of the service life of equipment and lower backup inventory. In an era of intensive competition, where asset usage and plant operating efficiency must be maximized, predictive maintenance becomes very important.
