Lean Six Sigma for Quality

DMAIC phases

1. 1.Define— Frame the problem, customer, scope and measurable goal.
   Key activities: Draft project charter with problem and goal statement; Identify customer and CTQ; Map high-level process (SIPOC)

   Exit criteria: Signed charter; Approved scope and Y metric
2. 2.Measure— Establish a validated baseline of the process performance.
   Key activities: Operationally define Y and Xs; Run measurement system analysis; Collect representative baseline data

   Exit criteria: Validated measurement system; Baseline capability documented
3. 3.Analyse— Identify and confirm root causes driving variation in Y.
   Key activities: Stratify data and generate hypotheses; Run graphical and statistical analysis; Confirm root causes with evidence

   Exit criteria: Verified root causes; Prioritised improvement targets
4. 4.Improve— Design, pilot and implement counter-measures.
   Key activities: Generate and evaluate solutions; Pilot with defined success criteria; Deploy validated solution

   Exit criteria: Piloted solution with measured impact; Post-improvement capability
5. 5.Control— Sustain the gain and hand back to the process owner.
   Key activities: Publish control plan and standard work; Set monitoring, response plan and audit cadence; Transfer ownership and close project

   Exit criteria: Active control plan; Sustained performance for defined horizon

Six Sigma tools

• SIPOC
  Phase: define

  High-level view of Suppliers, Inputs, Process, Outputs and Customers.

  Prerequisites: Agreed project scope

  Output: SIPOC diagram anchoring scope boundaries

  Misuse warning: Not a substitute for a detailed process map in Analyse.

• Critical-to-Quality Tree
  Phase: define

  Translate customer needs into measurable requirements.

  Prerequisites: Voice of the customer captured

  Output: CTQ tree with metrics and targets

  Misuse warning: Author-driven CTQs without customer input mislead the project.

• Measurement System Analysis / Gauge R&R
  Phase: measure

  Confirm the measurement system is capable of detecting real variation.

  Prerequisites: Operationally defined metric; Trained operators

  Output: MSA report with repeatability and reproducibility indices

  Misuse warning: Never draw improvement conclusions before the measurement system is proven capable.

• Process Capability (Cp, Cpk, Pp, Ppk)
  Phase: measure

  Quantify how a process performs against specification limits.

  Prerequisites: Stable process (for Cp/Cpk); Validated measurement system

  Output: Capability indices with the assumptions recorded

  Misuse warning: Reporting Cp/Cpk on an unstable process, or comparing indices across different sampling plans, misleads decisions.

• Statistical Process Control
  Phase: control

  Detect special-cause variation using control charts.

  Prerequisites: Stable data collection; Correct chart type for data

  Output: Live control chart with response rules

  Misuse warning: Reacting to every point ignores the purpose of control limits.

• Pareto Analysis
  Phase: analyse

  Prioritise the vital few causes contributing most to the effect.

  Prerequisites: Categorised defect or cause data

  Output: Ranked bar chart with cumulative line

  Misuse warning: Categories chosen for convenience hide the true drivers.

• Cause-and-Effect (Fishbone) Diagram
  Phase: analyse

  Structure brainstorming of possible causes across categories.

  Prerequisites: Clearly stated problem

  Output: Populated fishbone ready for data-based verification

  Misuse warning: Brainstormed causes are hypotheses, not confirmed root causes.

• 5 Whys
  Phase: analyse

  Drill from symptom toward systemic cause by asking why.

  Prerequisites: Access to people who do the work; Verified problem statement

  Output: Chain of causes ending at a systemic cause

  Misuse warning: Stopping at a convenient blame level misses the real cause.

• Failure Mode and Effects Analysis
  Phase: analyse

  Rank failure modes by severity, occurrence and detection.

  Prerequisites: Process or design map; Cross-functional team

  Output: Ranked risks with mitigation actions

  Misuse warning: RPN scoring without action ownership converts risk analysis into paperwork.

• Hypothesis Testing
  Phase: analyse

  Test whether an observed effect is statistically significant.

  Prerequisites: Defined hypotheses; Adequate sample size

  Output: Test result with p-value and practical interpretation

  Misuse warning: Statistical significance without practical significance drives the wrong decision.

• Cost of Poor Quality
  Phase: define

  Quantify the financial impact of failure, appraisal and prevention.

  Prerequisites: Access to financial and quality data

  Output: COPQ estimate with categories

  Misuse warning: Ignoring hidden costs understates the case for action.

• Standard Work
  Phase: control

  Codify the current best-known method for a task.

  Prerequisites: Stable process; Trained operators

  Output: Standard work documents with cycle and takt data

  Misuse warning: Freezing standard work prevents the improvement it should enable.

• Mistake-Proofing (Poka-Yoke)
  Phase: improve

  Design the process so errors are prevented or made obvious.

  Prerequisites: Understood failure mode

  Output: Poka-yoke device or design change

  Misuse warning: Adding checks without changing the design is inspection, not prevention.

• Control Plans
  Phase: control

  Define what is monitored, by whom, with what response when out of control.

  Prerequisites: Confirmed improvement; Trained owners

  Output: Signed control plan with response actions

  Misuse warning: Control plans without response actions leave gains unprotected.

Related calculators

• Line Efficiency Calculator
• Production Capacity Calculator