"Well-done" is the other part of the equation. To work well, predictive condition monitoring has to be a comprehensive program, with firm scheduling and dedicated personnel. It can't just be something that gets done if nothing else gets in the way. Also, the schedule has to include time for analysis and follow-up, at least as much time as is allocated for data collection. Otherwise, it's like having a Hummer in the garage and never driving it. It's just a waste of money.
Where is the opportunity for the most improvement?
Computerized condition monitoring has been around for 25 years, and people have learned the importance of a proper setup and collection of good data regularly. But what then? In many cases, the still-to-be-realized benefit is appropriate analysis of the data and making use of the huge advantage of good history over months or years.
Too many analysts refuse to commit themselves, choosing to rely on other people or "industry standards" rather than on their own analysis of their own data. An expert who comes in when a problem has arisen and who has to make a decision without benefit of context or history has a very difficult job. Even when you need to call in an outside expert, you should insist on due consideration being given to the quality history you have developed. Compare the two jobs in the table on the right.
Always remember that the job does not end with collecting data. It does not even end with analyzing data. The point is to make a justifiable decision about each machine, act on it, and follow up. One such decision might be thatou do need to hire a consultant to help solve the problem, because the problem is difficult, unusual, or potentially very expensive. If so, make sure you choose one with a proven track record, and who will add extra skills, not just duplicate yours. OWN your own machinery; don't cop out.
Val Zacharias earned a Master's Degree in Educational Curriculum and Instruction with a specialty in Communications. She spent 20 years working for Beta Machinery Analysis and Beta Monitors & Controls, where she participated in the development of the first computer-aided condition monitoring instruments and software, beginning before the existence of IBM PC's. Since 2000, she has served as Executive Director of the Canadian Machinery Vibration Association. www.cmva.com