Page 50 - RCM - A practical Guide_V1

RCM - A Practical Guide

Hidden or evident
Arguably, this can be the trickiest of questions to get right. Often, if the question is that difficult to
answer, then it can point to issues within the FMEA and it is worth pausing to double-check that.
If though, the FMEA is robust, then the analysis must default to ‘Hidden’.
Important or not important
Or safety or non-safety. Grey area, again in the FMEA, or difficulty in understanding the effects can
lead to uncertainty regarding how critical the loss of function (or secondary damage) could be.
In this instance the analysis must default to ‘Safety’ – (or whatever the most important factor in their
tailored decision tree).

Historical data
Data is the backbone on which a maintenance schedule is built. It’s a common opinion and, let’s be
st
honest, it’s the 21 century so it should be true but there are problems with its use.
Many organisations collect huge quantities of information regarding the failure of its assets. However,
there are several problems associated with the collection, and subsequent use of the data, which
makes its interpretation difficult and undermines any confidence in its use. The sheer quantity of data
which is presented, usually as a database output, can consist of tens of thousands of lines of
information. The quality of the data collected can vary greatly in its complexity and accuracy. For
instance, often the use of favoured terms such as “failed” or “broken” or “leaking” etc. tells the analyst
little about the actual cause of the failure. The upshot is that although identification of functional
failures is easy, identifying the cause of the specific failure very difficult if not impossible.
And then there’s the big one…any catastrophic failure which does occur or has occurred in the past
should have had preventive measures put in place to prevent a repeat failure. That means that,
provided the mitigation has been successful, an analysis will find no data about what is potentially a
life and death failure and therefore, data analysis cannot be used to identify management policies for
the failures of greatest concern.
This problem associated with the absence of critical data and the availability of a plethora of non-
critical data is referred to as Resnikoff’s conundrum.
The four collectives
To get around data issues an analysis team must look to the collective. The four collectives can inform
decisions made in the absence of explicit data.
They are:
• Collective
• engineering understanding
• experience
• common sense
• agreement (Consensus)
In the absence of clear data, the analysis must apply and agree to its collective judgement to ensure
the best possible answer.

50
© ASPIRE CONSULTING LTD +44 (0) 1827 723820