Failure mode and effects analysis (FMEA) is a bottom-up method that lists the ways each part of a system or process could fail, rates the severity, likelihood and detectability of each, and prioritizes fixes.

Failure mode and effects analysis (FMEA) is a structured, bottom-up method for finding how each component, step or function in a system could fail, what the effects of each failure would be, and which failures need action first. A failure mode is the way something fails, such as a valve sticking open, a sensor reading low or an operator skipping a step.
FMEA was first formalized by the US military in the late 1940s and later spread to aerospace, automotive, medical devices, manufacturing and healthcare. In safety, it is used to study equipment, processes and tasks before problems cause harm, and it is one of the methods OSHA accepts for a process hazard analysis.
Most failures are predictable if someone takes the time to look. FMEA gives teams a disciplined way to walk through every part or step, ask how it could go wrong and decide which problems deserve attention. That is far cheaper than finding the same failures through injuries, recalls or downtime.
FMEA also creates a living record of known risks. When equipment is modified, staff change or an incident occurs, the FMEA shows what was already considered and what controls were relied on, making future decisions faster and better informed.
The traditional risk priority number (RPN) is S x O x D, giving a score from 1 to 1,000. Because very different risks can share the same RPN, the 2019 AIAG and VDA FMEA Handbook used in the automotive industry replaced RPN with an action priority rating of high, medium or low, which puts more weight on severity. Many safety teams also treat any failure with a high severity score as a priority, whatever its RPN.
A packaging plant ran a process FMEA on a new palletizer cell. One failure mode was the light curtain muting function staying active after a pallet left the cell. The effect would be that a worker could walk in while the robot was moving. Severity was rated 10, occurrence 3 and detection 7 because there was no indication to operators that muting was stuck.
The team added monitoring that faults the cell if muting lasts longer than expected, a muting status lamp at the entry and a weekly function test. The new detection score dropped to 2.
The risk priority number is severity times occurrence times detection. It is used to rank failure modes, though many teams now pair it with severity-first rules or action priority ratings.
FMECA adds a criticality analysis to FMEA, ranking failure modes by both severity and probability, often with quantitative data.
FMEA is bottom-up: it starts with each component's failure modes and asks what happens. Fault tree analysis is top-down: it starts with one unwanted event and works back to its causes.
Ideally during design, before equipment or processes are built, and again when they change. It can also be used on existing operations to find weaknesses.
Severity, which rates how bad the effect would be; occurrence, which rates how likely the cause is; and detection, which rates how likely current controls are to catch the failure before it causes harm.
People who know the design or process from different angles, such as engineers, operators, maintenance technicians, quality staff and safety professionals. A facilitator keeps the session on track.
Part of SafetyIQ's EHS glossary: plain-English definitions of workplace health and safety terms.
> Keep learning
> See it in action
Capture hazards, run inspections, track training and close out corrective actions in one platform built for high-risk, field-heavy work.
Book my free demo