Why It Matters:
Detecting the presence of an anomaly is one thing. Defining and pinpointing it is another. Your monitoring system should not only warn you about problems, but also provide an accurate diagnosis with specific component identification, location, and indication of the extent of damage. Armed with this information, you can make well-founded decisions about the maintenance procedures you need to take and when you need to take them.
The Best Approach:
There are no shortcuts in developing a system for automated diagnoses. An ac¬curate, detailed understanding of a problem is acquired only through extensive experience in machinery monitoring. The reason is clear: when a monitoring system detects an anomaly, it compares its characteristics with an integrated, experience-based failure mode database. A match of this anomaly with a real-life failure pattern leads to an accurate, reliable diagnosis.
Equally valuable is the message your system communicates to you. In the event of a positive failure pattern match, alarm, or shutdown, you need answers. Modern systems should provide clear communication about the cause for the alarm. For example, the message “Discharge valve leakage, cylinder 2, match 89%” gives the cause of the problem, location of the problem, and confidence in the diagnosis.
Two additional considerations: first, be wary of systems that claim “expert diagnostics routines,” but lack dedicated recip know-how. Precise diagnoses cannot be performed without powerful and intelligent algorithms. Second, think about the capabilities of even the most sophisticated diagnostic technologies and the messages they generate. Is it realistic and reliable for a system to issue an impending damage report that identifies a screw (with its part number) that has to be tightened?
