Introduction to Six Sigma

A: The boundary condition is whether a process already exists. DMAIC is built for reducing variation in an established system; DMADV is triggered when no viable process or design exists yet. A field failure alone doesn't force a clean-sheet design.

A: The trigger is systematic versus random failure. ISO 26262 is allergic to systematic faults, and high variation is a classic breeding ground for them even if the average looks fine on paper.

A: The hard line is the control limit, not the spec. Control charts are about stability; a single point outside signals the process physics changed, regardless of customer tolerance.

A: The issue is measurability. Six Sigma demands an X you can control or quantify; 'operator error' collapses multiple mechanisms into a catch-all that can’t be improved statistically.

A: Cpk is bounded by the nearest spec. With σ constant, only mean movement erodes it. That boundary condition is what separates capability from centering.

A: The boundary is authority, not skill. Champions carry escalation power and resource control; statistical execution lives with the Belts.

A: The threshold is dominance. Once a single cause exceeds roughly half the population, leverage is obvious; spreading effort dilutes statistical impact.

A: The trigger is sustainment. IATF 16949 is focused on drift after launch; control plans lock in gains so yesterday’s fixes don’t become tomorrow’s warranty claims.

A: The boundary is data integrity. Without knowing the actual as-built distribution, every downstream statistic is noise, and Six Sigma collapses into guesswork.