Design for Six Sigma and failure mode analysis

When a defect appears in an existing process, first ask whether the process can be improved within its design. If so, DMAIC defines the problem, measures it, analyzes causes, improves the process and controls the result. If the product or process has not yet been designed, or the present design fundamentally cannot meet a customer requirement, use Design for Six Sigma. DMADV defines and measures needs, analyzes alternatives, designs and verifies. IDOV identifies needs, designs, optimizes and verifies. The different labels preserve a common decision: translate customer needs into testable requirements and test the design before release. Verification asks whether the design was built to its specification. Validation asks whether it meets its intended use; a medicine label can pass a font-size check and still confuse real users.

Consider a fictional new medicine label. The team defines severity S on a 1–10 scale, with 10 a severe safety effect; occurrence O runs from 1 rare to 10 frequent; detection D runs from 1 easy to catch before reaching the user to 10 hard to catch. These are hypothetical working scales that the organization must set before rating; they are neither percentages nor universal measurements. Mode one: the label resembles that of a different strength, so a user could select the wrong strength. The team rates S=9, O=3 and D=7, giving a risk priority number RPN = 9×3×7 = 189. Mode two: a wrinkle delays scanning, rated S=3, O=8, D=8, giving 192. The second mode is not automatically more important because 192 exceeds 189. The first mode's high severity may require a different safety decision. Examine the rationale for each rating and address severe effects even when their product is lower.

A design action for mode one is to make strengths visually distinct and test their use with representative people. A process action is a scanner that matches package code to the packing order and blocks release on a mismatch. After adding a check, do not lower S merely because a check exists; the consequence of a missed error remains. D might fall if the scanner is shown effective. If S stays 9, O stays 3 and D falls to 2, RPN becomes 54. Do not call that success until scanner failure modes have been challenged and residual error documented. If a requirement cannot be met at the line, revisit product design.

The result depends on the definition of the mode and the severity judgment, not multiplication alone. Two teams can assign different figures under different scales. Retain the scale version, rating participants, rationale, evidence, action owner and follow-up. FMEA does not replace an actual safety or compliance test. A project may use similar reasoning for operating risk, but a design FMEA is more than a generic risk list without functions, effects and controls.