Lean Six Sigma Matters
⚠️ Risk Management

FMEA — Prevent Failures
Before They Happen

Failure Mode and Effects Analysis is a structured, proactive technique for identifying potential failure modes in a process or product, assessing their impact, and prioritising actions to prevent them — before they reach the customer.

What is FMEA?

FMEA was originally developed by the US military in the 1940s and later adopted by the aerospace and automotive industries. Today it is a standard risk management tool used across manufacturing, healthcare, software, and service industries — and a core tool in both Six Sigma and ISO 9001 quality management systems.

The fundamental question FMEA answers is: "What could go wrong, how bad would it be, and what can we do about it?" It converts subjective risk assessment into a quantified priority score — the Risk Priority Number (RPN) — so teams can focus their effort where it matters most.

Which FMEA Do You Need?

There are three main types of FMEA — each focused on a different stage or aspect of product and process development.

Design FMEA (DFMEA)

Analyses potential failure modes in a product design before it is released to manufacturing.

  • Conducted during the design phase
  • Focuses on component and system failures
  • Identifies design improvements early
  • Used in automotive, aerospace, medical devices
  • Drives design for reliability and safety

Process FMEA (PFMEA)

Analyses potential failure modes in a manufacturing or service process — the most common type in Lean Six Sigma.

  • Conducted before process launch or during improvement
  • Focuses on process steps and operator actions
  • Identifies control gaps and error-proofing opportunities
  • Directly linked to control plans and SOPs
  • Standard tool in DMAIC Improve and Control phases

System FMEA (SFMEA)

Analyses interactions and interfaces between subsystems and components at the highest level.

  • Conducted early in system architecture phase
  • Focuses on system-level interfaces and integration
  • Identifies risks from subsystem interactions
  • Common in complex engineering systems
  • Forms the basis for Design and Process FMEAs

The RPN — Quantifying Risk

Each potential failure mode is scored on three dimensions. The product of these three scores gives the Risk Priority Number (RPN), which prioritises which failures require action first.

S

Severity

How serious is the effect of the failure on the customer or process? Rated 1–10.

1–2Minor — no effect or slight annoyance
3–5Moderate — process disruption or rework
6–8High — loss of primary function
9–10Critical — safety or regulatory impact
O

Occurrence

How likely is the failure mode to occur? Rated 1–10 based on frequency or probability.

1–2Remote — unlikely to occur
3–5Occasional — occurs sometimes
6–8Frequent — occurs regularly
9–10Almost certain — very high frequency
D

Detection

How likely is the current control to detect the failure before it reaches the customer? Rated 1–10.

1–2Almost certain detection — very reliable controls
3–5Moderate — some chance of escape
6–8Low — likely to escape detection
9–10No detection — no controls in place

Risk Priority Number

RPN = Severity × Occurrence × Detection

Maximum RPN = 1,000  |  Minimum RPN = 1  |  Typical action threshold: RPN ≥ 100, or any Severity = 9–10

How to Conduct an FMEA

FMEA is a team exercise — it should involve people with direct knowledge of the process or design. Follow these steps in sequence.

1

Define the Scope

Clearly define what process, product, or system is being analysed. Break it into its component functions or process steps.

2

Identify Failure Modes

For each function or step, ask: "In what ways could this fail?" List every potential failure mode — do not filter at this stage.

3

Identify Effects

For each failure mode, identify the effect on the customer, process, or next operation downstream.

4

Identify Causes

Determine the root causes of each failure mode. Use tools like the 5 Whys or fishbone diagram to go beyond symptoms.

5

Score S, O, D and Calculate RPN

Rate Severity, Occurrence, and Detection (1–10 each) for each failure mode. Calculate RPN = S × O × D.

6

Prioritise and Take Action

Focus on the highest RPN scores and any Severity 9–10 items. Assign owners, actions, and due dates.

7

Implement Actions

Carry out the recommended actions — design changes, process controls, error-proofing, or additional testing.

8

Re-score and Review

After actions are implemented, re-score the RPN to verify improvement. FMEA is a living document — update it regularly.

9

Link to Control Plan

For Process FMEA, ensure all critical controls identified are reflected in the process control plan and SOPs.

FMEA Table — Invoice Processing Example

This simplified example shows how an FMEA table is structured for a transactional process. In practice, a full FMEA may contain many more rows and columns.

Process Step Potential Failure Mode Effect of Failure S Potential Cause O Current Controls D RPN Recommended Action
Data Entry Incorrect invoice amount entered Overpayment or underpayment to supplier 7 Manual entry error, no validation 6 Manager review before payment 5 210 Implement automated validation against PO
Approval Invoice approved without authorisation Fraudulent or duplicate payment 9 Approval workflow bypassed 4 System access controls 4 144 Mandatory dual authorisation for amounts above threshold
Payment Run Duplicate payment issued Financial loss, supplier reconciliation issues 8 No duplicate check before payment 3 Manual check by accounts team 5 120 Automate duplicate invoice detection in ERP
Filing Invoice not archived correctly Unable to retrieve for audit — compliance risk 6 No standard filing procedure 5 None 7 210 Create SOP for electronic filing and naming convention

FMEA Resources & Further Reading

Need Help Running an FMEA Workshop?

I facilitate FMEA workshops for manufacturing, healthcare, and service organisations — helping teams identify and eliminate risk before it reaches the customer.