FMEA

FMEA is the preventive cousin of root cause analysis. Most quality work happens after a defect has shown up, the customer complains, the team investigates, the fix gets installed. FMEA flips the order. It assumes you can list how a process or product is likely to fail before it has failed, score the risk, and put the most important preventive actions in place during design or process planning. The discipline is harder than it sounds because it requires a team to imagine failure carefully rather than wait for it to happen.

"Failure that surprises you is the failure that costs the most. Imagine it before it imagines you."

How FMEA works

An FMEA is built one row at a time. The team walks through a process step by step, or a product part by part, and asks for each one: what could fail here, and how. Each plausible failure becomes a row in the FMEA worksheet. Then, for each row, three scores from one to ten are assigned:

• Severity (S), how bad the failure would be for the customer or downstream process if it occurred.
• Occurrence (O), how often the failure is expected to occur given the current process.
• Detection (D), how likely the current process is to catch the failure before it reaches the customer. A low score means good detection.

The three scores multiply into the Risk Priority Number (RPN = S x O x D), which gives the team a single number for ranking. The highest RPNs become the targets for preventive action. After actions are implemented, the team rescores the same rows. A meaningful FMEA shows RPNs dropping over time as countermeasures land. An FMEA that sits in a binder unchanged for two years was never an FMEA, just a form.

Two flavors are common. Design FMEA is done during product design, on the product itself: how could this housing crack, how could this connector mis-mate. Process FMEA is done on the manufacturing process: how could this assembly step skip a fastener, how could this paint step produce the wrong color. Most small shops use process FMEA because the design is set by the customer.

Where FMEA fits on the shop floor of a small manufacturer

Imagine a 25-person precision parts shop about to launch a new product family of stainless brackets for a food-equipment OEM. The bracket has three machined features, a weld, a passivation step, and a heat-treat cycle. The customer is strict on weld profile and surface finish. Mistakes after launch will be expensive.

A two-day FMEA session reorders the work. The team walks the routing step by step with two operators, the welder, and the quality lead. They list 38 plausible failure modes. Five score high RPNs. The weld has a porosity risk with a severity of nine and detection of seven. The passivation has a contamination risk if the prior step is rushed. The heat-treat has a load-density risk that has bitten other shops.

The team installs countermeasures for the top five before the launch run. A pre-weld surface check. A poka-yoke fixture in passivation that physically blocks parts from the wrong rack. A heat-treat load standard with a visual cue at the oven. After two months of production, the rescored RPNs on those five modes have all dropped to manageable levels. Zero customer rejects in the first 1,200 parts.

That is FMEA at small scale. Not a forty-tab spreadsheet, just a one-pager per launch with disciplined scoring and real follow-up.

Common mistakes with FMEA

• Treating it as paperwork. An FMEA done to satisfy a customer audit never produces preventive action. If no RPNs ever drop, the document is theatre.
• Inconsistent scoring scales. If severity 7 means one thing this month and another next month, the rankings are noise. Calibrate scales with examples before scoring.
• No rescoring after action. The whole point of FMEA is risk reduction over time. A worksheet without follow-up scores is documenting failure modes, not managing them.
• Doing it without the operators. Engineers and managers who do not run the process daily will under-score occurrence and over-score detection. Bring the people who do the work.
• Stopping after the first session. FMEA is a living document. Process changes, material changes, and customer-impact changes all change the rankings. Review at least once a year.

FMEA and related Lean tools

FMEA produces the priority list; poka-yoke is often the countermeasure installed against the highest-RPN modes, since a well-built poka-yoke drops the occurrence or detection score directly. When FMEA highlights a mode that has already produced a defect, the team usually pivots into 8D problem solving for the formal containment and corrective action. FMEA fits inside the broader practice of early equipment management and feeds countermeasure decisions on the floor.