Biomedical Engineering Reference
In-Depth Information
TABLE 7.5 Example of an FMEA Template
Process
Step/Unit
Operation
Failure SEV Cause OCC Current
Control
DET RPN Risk
Accepted
system function, speed, lighting levels, etc. The qualification would verify that
the system will function in a consistent manner.
7.3.10 Risk Ranking Level of Detail
Risk assessment and risk ranking tools provide the information needed to help
make an informed risk-based decision. However, they should not necessarily
make the decision. The focus should be on evaluating the outcome of the risk
assessment, rather than evaluating strictly the risk ranking value.
Consider how an FMEA (failure mode effects analysis) model is used to rank
risk. Table 7.5 presents a typical FMEA spreadsheet template for ranking risk.
The level of numeration detail or granularity of the ranking choices, shown in
Table 7.6, is the number or quantity of choices one has in ranking risk elements,
including severity, occurrence, and detection. Some FMEA models utilize ranking
value choices ranging from 1 to 10. This would represent a relatively high detail
or granularity method. Some models use less numbers or qualitative choices, such
as high, medium, and low as illustrated in Table 7.8. This represents a relatively
low detail or granularity method.
Where quantitative data is available, such as number or rate of defects, pro-
duction yields, or failure rates, higher granularity methods should be objective
and useful. However, where quantitative data is not readily available and qualita-
tive ranking is used, such as quality of, confidence in, or experience of a vendor;
a lower granularity method may be more useful and effective. In these cases,
lower granularity, such as high, medium, and low may reduce the potential for
bias and ranking value “debate.”
In the FMEA assessment example presented in Tables 7.6 through 7.8, the
difference between a “marginal” and a “critical” risk is one point, 500 versus
501. Taking into account the descriptions presented in Table 7.7, this may mean
that the difference between an acceptable and unacceptable risk is one point. In
this case, the assessment team may be “tempted” to select individual scores in
order to achieve a lower risk ranking. For instance, the team might spend much
of the time debating whether an occurrence value, is a four, a five, or a six.
If, on the other hand, the choices were high, medium, and low, as presented in
Table 7.8, then occurrence might be often (high) or seldom (low), or anything in
between (medium). The team would ask questions such as: Does the event occur
often? If so, then the value is high. Does it occur seldom? If so, then the value
is low. Otherwise it is medium. This might lead to productive discussion over
Search WWH ::




Custom Search