SiliciumHex FieldKit

Problem Solving & Quality · Quality Tools

Failure Mode and Effects Analysis

For each function or process step, list how it can fail, the effects and causes, rate them, and act on the highest risks before they happen.

  • Time2 h
  • FormatTeam
  • StageQuality Tools

Failure Mode and Effects Analysis: what it is and why it works

Failure Mode and Effects Analysis is a structured team review of how a product, process or piece of equipment could fail. For each function or process step, the team lists potential failure modes, their effects on the customer or the next operation, their causes, and the prevention and detection controls already in place. Each chain is rated for severity, occurrence and detection, typically on 1 to 10 scales. Priority is then set either with the Risk Priority Number (severity × occurrence × detection) or, in the AIAG-VDA approach, with an Action Priority table that ranks combinations as high, medium or low and gives severity the dominant weight.

FMEA works because it forces risk thinking before the failure happens, when changing a design or a process is cheapest. Its strength is the cross-functional discussion: design, production, maintenance and quality each see different failure paths. Compared with a fault tree, which starts from one top event and works down, FMEA works bottom-up across every element, so it gives broad coverage. It feeds the control plan directly, since high-severity and high-priority lines define what must be checked and how. Potential problem analysis is a lighter cousin for a single implementation plan.

What you need

  • A defined scope: product, process flow or equipment, with drawings or a process flow diagram
  • A cross-functional team of four to eight people who know the design, operation and maintenance
  • Failure history: incidents, complaints, scrap data, maintenance records from similar items
  • Agreed severity, occurrence and detection rating scales and the prioritization method (RPN or Action Priority)
  • Existing prevention and detection controls, such as inspections, interlocks and procedures

What you get

  • A worksheet linking functions, failure modes, effects, causes and current controls
  • Severity, occurrence and detection ratings with a priority for each chain
  • Recommended actions with owners, dates and re-rated risk after implementation
  • The list of special characteristics that must go into the control plan

When to use it

When a new line, a design change or a recurring failure calls for thinking ahead, not after.

How to do it, step by step

  1. Define the scope — a product, a process or an equipment — and gather a cross-functional team.
  2. Break it into functions or process steps; for each, list the potential failure modes.
  3. For each failure mode, write the effects, the causes and the current prevention and detection controls.
  4. Rate severity, occurrence and detection, then prioritize — with RPN or an action-priority table — always treating high severity first.
  5. Assign actions with owners and dates, re-rate after implementation, and keep the FMEA alive after every incident.

Worked example: Process FMEA for a new bulk-bag filling station

Illustrative scenario — figures are realistic but not from a real company.

A specialty chemicals plant was installing a bulk-bag filler for a fine, mildly irritating powder, sold in 2,200 lb bags. The team of production, maintenance, environmental health and safety, quality and one operator had two sessions of two hours before the design freeze.

  1. The team split the process into six steps: hang bag, inflate liner, fill with densification, weigh, close and discharge by forklift. Past incidents from the old manual station were read out first.
  2. Fourteen failure modes emerged, including a bag strap released during filling, dust escaping at the spout seal, overweight bags and a mislabeled lot number.
  3. Using the AIAG-VDA scales, the strap release scored severity 10 (a suspended load falling near a person), and with no control able to detect a loosening strap before release, the Action Priority table rated it high even though occurrence was low.
  4. Actions included a load-cell interlock that stops filling if weight distribution becomes uneven, a keep-out zone marked and guarded during filling, and a spout seal pressure check at start-up.

Result. Nine actions were built into the station before commissioning at little extra cost, compared with retrofits later. The high-priority items dropped to low after re-rating. Dust and overweight issues became lines in the new control plan. The team noted that reading the old incident log first produced half of the useful failure modes.

Open it in the FieldKit Problem Solving app → Get the Problem Solving template pack — join the waitlist

Common pitfalls and how to avoid them

  • Sorting by RPN alone, so a severity 10 failure with low occurrence sits below a cosmetic issue.Treat high severity first, or use the Action Priority logic, which is built to prevent this.
  • Filling the worksheet alone at a desk to meet a deadline.Run it as a team with operators and maintenance, time-boxed in short sessions, and have the facilitator write while others think.
  • Listing causes that are just the failure mode restated, such as "seal fails because seal is defective".Write causes that can be acted on: wrong material, installation error, excessive temperature, missing flush.
  • Archiving the FMEA after launch.Update it after every incident, design change and customer complaint, and review it at a fixed interval.

Frequently asked questions

What is the difference between RPN and Action Priority in FMEA?

RPN multiplies severity, occurrence and detection into a number from 1 to 1,000 and ranks by that number, which can hide a severe failure behind moderate scores. Action Priority, introduced in the AIAG-VDA FMEA handbook, uses a lookup table that assigns high, medium or low priority to each combination, weighting severity first, then occurrence, then detection. It removes the illusion that RPN values can be compared precisely.

What is the difference between a DFMEA and a PFMEA?

A design FMEA looks at how the product can fail to perform its function because of design choices such as materials, geometry and tolerances. A process FMEA looks at how the manufacturing or assembly process can produce a defective product or create a hazard. They are linked: design characteristics that are critical in the DFMEA become characteristics to control in the PFMEA and the control plan.

How long does an FMEA take?

It depends on scope. A focused process FMEA on a small station can take two or three sessions of two hours; a full product design FMEA can take many sessions over weeks. Keeping the scope tight, preparing the structure and failure history beforehand, and limiting sessions to two hours keeps the team engaged and the quality high.

Origin

FMEA — US military procedure MIL-P-1629, 1949; adopted by NASA in the 1960s and by the automotive industry (AIAG, 1993; AIAG–VDA handbook, 2019).

Related methods

More in “Quality Tools”

The statistical and risk tools of industrial quality: Pareto, SPC, capability, measurement, FMEA.