ISE548 Chap.6 FMEA and Risk Priority Numbers
FMEA and Risk Priority Numbers
Failure analysis with score discipline
FMEA records a function, failure mode, effect, cause, current prevention or detection control and recommended action. Severity, occurrence and detection scores produce a risk priority number for ranking work inside an agreed scale. This chapter makes score direction explicit, separates action mechanisms and requires a second score after implementation.
It also protects high-severity rows from disappearing behind a convenient product.
The worksheet should reveal a causal chain rather than a list of labels. State what function is required, how it fails, what effect reaches the customer or system and which cause creates that mode. Existing controls then support separate occurrence and detection judgements.
Completed actions are tested before the revised factors replace the baseline.
What this chapter covers
- 01
Functions and failure modes
- 02
Effects and system levels
- 03
Causes and existing controls
- 04
Severity, occurrence and detection
- 05
RPN calculation and ranking
- 06
Action completion and rescoring
Worked example · free
Rescore a chilled package error
- 2Calculate the original RPN.
- 4Explain factor movement.
- 3Calculate and interpret the new RPN.
Key terms
- Failure Mode
- The specific way a product, service or process can fail to perform its function.
- Effects Analysis
- The examination of what a failure does at local, system and customer levels.
- Detection Score
- An FMEA ranking of how likely current controls are to find a failure before its effect, with poor detection scored higher.
- Risk Priority Number
- The product of severity, occurrence and detection scores used to order attention within one analysis.
FMEA and Risk Priority Numbers FAQ
Why does severity often stay unchanged after action?
Many actions prevent a cause or improve detection without limiting the ultimate consequence. Severity should fall only when containment, redundancy, recovery or another mechanism genuinely reduces the effect if failure reaches the customer.
Can RPN alone decide which row comes first?
It is useful for ranking, but high severity deserves a separate review because multiplication can hide a rare catastrophic effect. State the product rule and any severity override used by the decision maker.
What proves that an FMEA action is complete?
Record the owner, target date, implemented change, test evidence and new scores. Purchasing equipment or editing a procedure is an input; completion requires evidence that the failure path or detection performance changed.
Why can equal RPNs require different action?
A high-severity rare failure with strong detection can equal a moderate frequent failure that is difficult to find. The first may require a severity override, design assurance or contingency; the second may call for process prevention and better detection. Keep all three factors and the failure chain visible. The product ranks work within a scale but does not erase the pattern that produced it.
How do I decide which factor an action changes?
Trace the mechanism. Removing a cause or reducing its frequency changes occurrence. Finding the mode before the effect reaches the customer changes detection. Containment, redundancy or recovery that limits the final effect may change severity. A good action can affect several factors, but each reduction needs its own route and evidence; do not lower every score as a reward for implementation.
What does a high detection score mean?
Under the scale used here, a high detection number means the current controls are unlikely to find the failure before its effect. This direction is easy to reverse because ordinary language treats high detection as good. Write the scale beside the worksheet, describe the actual method and ask whether it operates under every relevant mode, shift and demand condition.
How should failure modes be selected for analysis?
Choose modes that represent distinct mechanisms and meaningful effects, not several phrasings of one fault. For a dispatch process, selection error, equipment failure and information delay may expose different controls and owners. Cover severe tails even when their product is not highest. State the process boundary so upstream causes and downstream consequences are not mistakenly entered as competing modes.
How should an FMEA ranking be reviewed after action?
Keep the baseline row, record the implemented action and rescore only after verification. Then sort the complete set again because removing one leading failure changes the relative priority of the rest. Check severity overrides separately and examine whether the new control created another mode or common dependency. A lower product is useful only when the causal pattern and operating boundary remain visible.
Exam move
Build three FMEA rows for one process using distinct causes. Calculate every product by hand, then propose one preventive, one detection and one consequence-limiting action. Rescore only the factors affected and compare the score pattern rather than reporting only the percentage reduction.
Working through FMEA and Risk Priority Numbers in ISE548? Sia is AskSia’s AI Management tutor — ask any ISE548 FMEA and Risk Priority Numbers question and get a clear, step-by-step explanation grounded in how ISE548 is taught and assessed. Read this chapter free, then take your hardest questions to Sia.