Usually, insufficient data are
available for valid statistical analysis, and these techniques then must be based on riskmanagement
techniques. Risk management typically involves developing a “matrix” with
the probability of failure along one axis and the consequence of failure (from minor to
major) along another. Devices can be placed in the matrix according to where they sit on
each axis and ranked, allowing resources to be targeted at the highest risk items (i.e., those
where both the probability and consequence of failure are highest). The rationale behind
any technique adopted should always be documented. (See also Chapter 56.)