Staying on Track in a Complex Environment

Introduction

The “low hanging fruit” has been harvested in most places. Now comes the challenge of gathering the most bountiful harvest, that which is amongst the leaves and branches, without harming the tree. This will require practical expertise. This will require computational models that narrow the scatter band and are more accurate that are asking the right questions (which requires practical knowledge, technical knowledge and experience = expertise). In this editorial, I will point out some of the pitfalls I see in the inspection and reliability arenas and present some insight and solutions that will help “IJ” readers stay on track and emerge more successful as a result.

There are times when using a model that is very coarse (what I call flying at the 50,000' level) is appropriate and may serve as a screening method to justify further analysis of the higher risk equipment. Usually the coarser models are used because they take less time. Although I have seen, with the creative use of computing technology coupled with practical experience, more quantitative models, that are more accurate, achieve much better results in comparable timeframes, at no additional cost/time and less subjectivity.

I really want to stress the importance of good analytical models in this editorial. There are basic considerations that you would do well to think about, seriously,

1) We often abandon tools with great computational models for a software tool that is “all things to all people” with weaker, more quantitative models, and losing metrics in the process. Here someone is looking for synergies of stored data and creating relationships that lead to the one grand master program and queriable reporting capabilities. It is a noble goal that does not need to suffer losing the effective computational models. Here are some proposed solutions:

• Keep the original computational program/s, that makes more accurate predictions, and feed the output to an overall notification and scheduling platform that is mainly concerned with a unique equipment ID and a date with a description of the recommended course of action. You may have a multitude of such analysis programs, e.g. for fixed equipment, for rotating equipment, for instrumentation and controls, etc. whose inputs would be managed by the overall-scheduling platform. Some of these exist in CMMS programs and independently. Since most companies still have budget/cost centers for these different areas the output from the overall scheduling program would tag each item according to type.