Lean Six Sigma Consultant

Lean Six Sigma Consultant

Overview

This skill provides comprehensive Lean Six Sigma consulting capabilities, integrating the waste elimination focus of Lean with the variation reduction rigor of Six Sigma. It supports practitioners at all belt levels from awareness (White Belt) through strategic deployment (Master Black Belt).

Primary Language: Japanese (default), English supported Knowledge Base: 33 reference files covering methodology, tools, statistics, industries, and training Output Format: Project guidance, analysis reports, templates, training materials

The Integration of Lean and Six Sigma

Six Sigma focuses on reducing process variation and defects using statistical methods (DMAIC methodology). The goal is achieving 3.4 defects per million opportunities (6 sigma level).

Lean focuses on eliminating waste and improving flow using Toyota Production System principles. The goal is maximizing customer value while minimizing waste.

Lean Six Sigma combines both:

• Use Lean tools to identify and eliminate waste, improve flow
• Use Six Sigma tools to reduce variation, solve complex problems
• Use DMAIC as the overarching improvement framework

When to Use This Skill

Use this skill when:

Project Execution:

• Leading or supporting a process improvement project
• Need guidance on which DMAIC phase tools to use
• Want to reduce defects, errors, or variation in a process
• Need to eliminate waste and improve cycle time
• Implementing statistical process control

Analysis Support:

• Conducting root cause analysis (5 Whys, Fishbone)
• Calculating process capability (Cp, Cpk)
• Creating or interpreting control charts
• Performing hypothesis testing or regression analysis
• Analyzing value streams for waste

Training/Education:

• Learning Six Sigma methodology and tools
• Preparing for belt certification exams
• Understanding statistical concepts
• Teaching Six Sigma to team members

Example Requests

1. "I want to reduce the defect rate in our manufacturing line. Guide me through a DMAIC project."
2. "Create a SIPOC diagram for our order fulfillment process."
3. "Calculate the process capability (Cpk) for this data: USL=10.5, LSL=9.5, mean=10.02, std=0.15"
4. "Help me identify the 8 wastes in our customer service process."
5. "What control chart should I use for tracking defect count per batch?"
6. "Explain the difference between Cp and Cpk for my Green Belt exam."
7. "Conduct a root cause analysis using the Fishbone diagram for high customer complaints."

Prerequisites

Before using this skill, ensure the following dependencies are installed:

# Required Python packages for statistical analysis scripts
pip install numpy scipy

Scripts require:

• Python 3.8+
• numpy for numerical calculations
• scipy for statistical functions (normal distribution, hypothesis testing)

Optional dependencies:

• Process measurement data in CSV format
• Specification limits (USL, LSL) for capability analysis

Workflow

This skill supports multiple workflows depending on the user's objective:

1. DMAIC Project Execution (Core Workflow 1)

   • Define → Measure → Analyze → Improve → Control
   • Use for improving existing processes

2. DMADV/DFSS (Core Workflow 2)

   • Define → Measure → Analyze → Design → Verify
   • Use for designing new processes/products

3. Lean Waste Elimination (Core Workflow 3)

   • Value Stream Mapping → Identify 8 Wastes → Design Future State → Kaizen
   • Use for eliminating waste and improving flow

4. Statistical Analysis Support (Core Workflow 4)

   • Process capability, control charts, hypothesis testing
   • Use for data-driven decision making

5. Training/Education (Core Workflow 6)

   • Belt-level curricula (White to Master Black Belt)
   • Use for learning and certification preparation

Quick Start:

# Calculate sigma level from defect data
python scripts/sigma_calculator.py --defects 15 --units 1000 --opportunities 5

# Analyze process capability
python scripts/process_capability.py --demo

# Perform control chart analysis
python scripts/control_chart_analysis.py --demo

Output

This skill produces the following outputs:

Example Output (Sigma Calculator):

================================================================
SIGMA LEVEL ANALYSIS REPORT
================================================================

📥 INPUT DATA:
   Defects: 15
   Units: 1000
   Opportunities Per Unit: 5
   Total Opportunities: 5000
   Dpu: 0.015

📊 CALCULATED METRICS:
   DPMO: 3,000.00
   Sigma Level: 4.16σ
   Yield: 99.7%
   Defect Rate Percent: 0.3%

📈 INTERPRETATION:
   Good (Industry Average)
   3,000 defects per million opportunities
   Target: 6σ = 3.4 DPMO
================================================================

Lean Six Sigma Framework

DMAIC Overview

DMAIC is the core improvement methodology for existing processes:

DMADV Overview (Design for Six Sigma)

DMADV is used for designing new processes or products:

Belt Hierarchy and Competencies

Core Workflow 1: DMAIC Project Execution

Use this workflow when improving an existing process.

Step 1: Define Phase

Objective: Clearly define the problem, scope, and goals.

1.1 Create Project Charter

Load assets/project_charter_template.md and complete:

Business Case: Why is this project important?

• Financial impact (cost of poor quality, revenue loss)
• Customer impact (complaints, satisfaction scores)
• Strategic alignment

Problem Statement (Specific, factual, quantified):

• What: What is wrong or not performing?
• When: When does it occur?
• Where: Where does it occur?
• Extent: How big is the problem? (Quantify)

Goal Statement (SMART):

• Specific: What will be achieved?
• Measurable: What is the target metric?
• Achievable: Is the target realistic?
• Relevant: Does it align with business objectives?
• Time-bound: When will it be achieved?

Scope:

• In scope: Process boundaries, included areas
• Out of scope: Excluded areas, constraints

Team:

• Champion: Executive sponsor
• Process Owner: Accountable for process
• Team Leader: Belt leading the project
• Team Members: Subject matter experts

1.2 Create SIPOC Diagram

Load references/tools-by-phase/define/sipoc_guide.md and assets/sipoc_template.md.

SIPOC provides high-level process view:

• Suppliers: Who provides inputs?
• Inputs: What materials, information, resources?
• Process: 5-7 high-level steps
• Outputs: What does the process produce?
• Customers: Who receives outputs?

1.3 Capture Voice of Customer (VOC) and CTQs

Load references/tools-by-phase/define/voc_ctq_guide.md.

VOC Collection Methods:

• Customer interviews
• Surveys
• Complaint data analysis
• Focus groups
• Social media analysis

CTQ (Critical to Quality) Tree:

Customer Need → Driver → CTQ (Measurable)
"Fast delivery" → "Delivery time" → "95% orders delivered within 2 days"

1.4 Define Phase Tollgate Questions

Before moving to Measure:

• Is the problem clearly defined and quantified?
• Is the project scope appropriate (not too big/small)?
• Are the goals SMART and achievable?
• Is the team assembled with right skills?
• Is champion engaged and supportive?
• Are baseline metrics identified?

Step 2: Measure Phase

Objective: Establish baseline performance and data collection system.

2.1 Develop Data Collection Plan

Load references/tools-by-phase/measure/data_collection_plan.md.

Operational Definitions: Clear, unambiguous definitions

• What exactly is a "defect"?
• How is cycle time measured (start/end points)?
• What units? What precision?

Data Types:

• Continuous: Measured on a scale (time, weight, temperature)
• Discrete/Attribute: Counted (defects, errors, pass/fail)

Sampling Strategy:

• Sample size: Use statistical calculations or rules of thumb
• Frequency: How often to collect
• Stratification: Collect by shift, machine, operator to enable analysis

2.2 Validate Measurement System (MSA)

Load references/tools-by-phase/measure/msa_guide.md.

Before collecting data, validate the measurement system:

• Gage R&R (Repeatability & Reproducibility): Is measurement variation acceptable?
• Accuracy: Does the measurement match the true value?
• Stability: Is measurement consistent over time?

Acceptance Criteria:

• Gage R&R < 10%: Excellent
• Gage R&R 10-30%: Acceptable with caution
• Gage R&R > 30%: Unacceptable, fix measurement system first

2.3 Create Detailed Process Map

Map the process in detail:

• Swim lane diagram: Show handoffs between departments
• Value-added analysis: Mark each step as VA, NVA, or BNVA
• Cycle time: Record time for each step
• Identify pain points: Bottlenecks, rework loops, wait times

2.4 Establish Baseline Metrics

Load references/tools-by-phase/measure/baseline_metrics.md.

Primary Metric (Y): The main output measure

• Baseline value: Current performance
• Target: Goal to achieve
• Entitlement: Best possible (benchmark)

Sigma Level Calculation:

DPMO = (Defects / (Units × Opportunities)) × 1,000,000
Sigma Level = Look up in conversion table or calculate

2.5 Measure Phase Tollgate Questions

Before moving to Analyze:

• Is the measurement system validated (MSA)?
• Is baseline data collected and reliable?
• Is the process mapped in sufficient detail?
• Is the sigma level or capability calculated?
• Are key process inputs (Xs) identified?

Step 3: Analyze Phase

Objective: Identify and verify root causes.

3.1 Generate Potential Causes

Load references/tools-by-phase/analyze/root_cause_analysis.md.

Fishbone (Ishikawa) Diagram: Organize potential causes into 6 categories (6M):

• Man (People): Skills, training, fatigue
• Machine (Equipment): Age, maintenance, calibration
• Material: Quality, specifications, suppliers
• Method: Procedures, work instructions
• Measurement: Accuracy, calibration
• Mother Nature (Environment): Temperature, humidity, lighting

5 Whys Analysis: Ask "Why?" repeatedly to drill down to root cause:

Problem: Machine stopped
Why 1: Overloaded → Why 2: Bearing failed → Why 3: Insufficient lubrication
→ Why 4: No PM schedule → Why 5: No maintenance program
Root Cause: Lack of preventive maintenance program

3.2 Prioritize with Pareto Analysis

Load references/tools-by-phase/analyze/pareto_analysis.md.

Pareto Principle: 80% of effects come from 20% of causes

Steps:

1. Categorize defects/problems
2. Count frequency of each category
3. Sort in descending order
4. Calculate cumulative percentage
5. Identify the "vital few" (typically top 20% of categories causing 80% of problems)

3.3 Verify Causes with Data

Load references/tools-by-phase/analyze/statistical_analysis.md.

Statistical Tools by Situation:

Hypothesis Testing Framework:

• H₀ (Null): No effect/difference
• H₁ (Alternative): Effect/difference exists
• α (Alpha): Significance level (typically 0.05)
• p-value < α → Reject null hypothesis → Effect is statistically significant

3.4 Analyze Phase Tollgate Questions

Before moving to Improve:

• Are root causes identified with data evidence?
• Are the "vital few" causes prioritized?
• Is there statistical verification of cause-effect?
• Can the team explain why the problem occurs?
• Are potential solutions starting to emerge?

Step 4: Improve Phase

Objective: Develop, test, and implement solutions.

4.1 Generate Solutions

Load references/tools-by-phase/improve/solution_generation.md.

Brainstorming Guidelines:

• Quantity over quality initially
• No criticism during generation
• Build on others' ideas
• Encourage wild ideas

Benchmarking: Learn from best practices elsewhere

TRIZ Principles: Structured inventive problem solving

4.2 Select Best Solutions

Load references/tools-by-phase/improve/solution_selection.md.

Pugh Matrix (Concept Selection):

1. Define evaluation criteria
2. Select baseline concept
3. Compare each alternative against baseline (+, -, S)
4. Sum scores and select winner

Decision Matrix (Weighted Scoring):

4.3 Conduct FMEA

Load references/tools-by-phase/improve/fmea_guide.md and assets/fmea_template.md.

FMEA (Failure Mode and Effects Analysis): Proactively identify and mitigate risks in the solution.

RPN = Severity × Occurrence × Detection

Prioritize actions for high RPN items (typically > 100).

4.4 Pilot Testing

Load references/tools-by-phase/improve/pilot_testing.md.

Pilot Plan:

• Scope: Limited area/time for testing
• Success criteria: What determines success?
• Data collection: How to measure pilot results?
• Contingency: What if pilot fails?

Before-After Comparison:

• Compare pilot metrics to baseline
• Use statistical tests to verify improvement
• Document lessons learned

4.5 Improve Phase Tollgate Questions

Before moving to Control:

• Are solutions addressing verified root causes?
• Is solution selection documented and justified?
• Were risks identified and mitigated (FMEA)?
• Was a pilot conducted successfully?
• Is improvement statistically significant?
• Is implementation plan ready?

Step 5: Control Phase

Objective: Sustain improvements and prevent regression.

5.1 Develop Control Plan

Load references/tools-by-phase/control/control_plan_guide.md and assets/control_plan_template.md.

Control Plan Elements:

• What to control (CTQs, key parameters)
• How to measure
• Sample size and frequency
• Control method (control chart type)
• Specification limits
• Reaction plan (what to do if out of control)
• Responsible person

5.2 Implement Statistical Process Control (SPC)

Load references/tools-by-phase/control/control_charts_guide.md.

Control Chart Selection:

Out-of-Control Rules (Western Electric):

1. One point beyond 3σ
2. Two of three consecutive points beyond 2σ (same side)
3. Four of five consecutive points beyond 1σ (same side)
4. Eight consecutive points on same side of centerline
5. Six consecutive points trending up or down

5.3 Create Standard Work

Load references/tools-by-phase/control/standard_work.md.

Standard Work Documentation:

• Step-by-step work instructions
• Visual aids and photos
• Key quality checkpoints
• Safety considerations
• Required tools and materials

Training:

• Train all operators on new process
• Verify competency
• Post visual instructions at workstation

5.4 Handoff to Process Owner

Handoff Checklist:

• Control plan documented and implemented
• Control charts established and being used
• Standard work documented and trained
• Response plan for out-of-control conditions
• Monitoring schedule and responsibilities clear
• Project documentation archived

5.5 Control Phase Tollgate Questions

Before closing project:

• Is the control plan implemented?
• Are control charts showing stable process?
• Are standard work documents completed?
• Is the process owner trained and accepting ownership?
• Are results meeting the goal?
• Is project documentation complete?

Core Workflow 2: DMADV/DFSS for New Processes

Use this workflow when designing a new process or product (not improving existing).

Load references/methodology/02_dmadv_dfss.md for detailed guidance.

When to Use DMADV Instead of DMAIC

Use DMADV when:

• Creating entirely new process/product
• Existing process is beyond repair (needs redesign)
• No baseline exists to improve from
• Customer requirements are significantly changing

DMADV Phases Summary

Define: Identify the project goals aligned with customer demands and enterprise strategy.

Measure: Measure and determine customer needs and specifications.

• Extensive VOC analysis
• Translate needs to measurable CTQs
• Benchmark competitors

Analyze: Analyze process/product options to meet customer needs.

• Generate design concepts
• Use Pugh Matrix for selection
• Predictive analysis

Design: Design the process/product to meet customer needs.

• Detailed design development
• FMEA for design risks
• Simulations and modeling
• Design reviews

Verify: Verify the design performance and ability to meet customer needs.

• Pilot testing
• Full-scale validation
• Capability studies
• Handoff to operations

Core Workflow 3: Lean Waste Elimination (VSM-based)

Use this workflow to identify and eliminate waste in processes.

Step 1: Understand Value

Load references/methodology/03_lean_principles.md.

Value Definition: What the customer is willing to pay for

• Does this activity transform the product/service?
• Does the customer care about this activity?
• Is it done right the first time?

Activity Categories:

• Value-Added (VA): Transforms product, customer pays for it
• Non-Value-Added (NVA): Pure waste, eliminate
• Business Non-Value-Added (BNVA): Required but no customer value (compliance, etc.)

Step 2: Map the Current State Value Stream

Load references/lean-tools/value_stream_mapping.md.

Value Stream Map Elements:

• Process boxes (with data: cycle time, changeover, uptime)
• Inventory triangles (with quantities)
• Information flows (orders, schedules)
• Timeline (processing time vs. lead time)

Key Metrics:

• Lead Time: Total time from order to delivery
• Processing Time: Sum of value-added time
• Value-Added Ratio: Processing Time / Lead Time (often < 5%!)

Step 3: Identify the 8 Wastes (DOWNTIME)

Load references/lean-tools/eight_wastes_downtime.md.

Step 4: Design Future State

Future State Principles:

• Produce to takt time (customer demand rate)
• Create continuous flow where possible
• Use pull systems (Kanban) where flow isn't possible
• Level the production mix (Heijunka)
• Build in quality (Jidoka)

Step 5: Implement with Kaizen Events

Load references/lean-tools/kaizen_events.md.

Kaizen Event (Rapid Improvement):

• Focused 3-5 day event
• Cross-functional team
• Make changes during the event
• Measure before/after
• Sustain with visual management

5S Implementation: Load references/lean-tools/five_s_guide.md.

Core Workflow 4: Statistical Analysis Support

Use this workflow when needing statistical guidance.

Process Capability Analysis

Load references/statistics/process_capability.md.

Cp (Process Capability):

Cp = (USL - LSL) / (6σ)

• Measures process potential if perfectly centered
• Cp = 1.0 means process spread = specification spread
• Cp = 1.33 is typical minimum requirement
• Cp = 2.0 is Six Sigma target

Cpk (Process Capability Index):

Cpk = min[(USL - μ)/(3σ), (μ - LSL)/(3σ)]

• Measures actual capability considering centering
• Cpk < Cp indicates process is not centered
• Cpk = Cp when perfectly centered

Interpretation Guide:

Control Chart Implementation

Load references/statistics/control_chart_types.md.

Control Limit Calculation (X-bar/R Chart):

UCL_X = X̄̄ + A₂R̄
LCL_X = X̄̄ - A₂R̄
UCL_R = D₄R̄
LCL_R = D₃R̄

Constants Table:

Hypothesis Testing Guide

Load references/statistics/hypothesis_testing.md.

Decision Flowchart:

Q: Comparing what?
├── Two means → 2-sample t-test
├── Multiple means → ANOVA
├── Proportions → Chi-square / Z-test
├── Relationship → Correlation / Regression
└── Factors/Interactions → DOE

Core Workflow 5: Tool Selection Guide

Use this matrix to select appropriate tools based on DMAIC phase and purpose.

Define Phase Tools

Measure Phase Tools

Analyze Phase Tools

Improve Phase Tools

Control Phase Tools

Core Workflow 6: Training/Education Mode

Use this workflow for learning and certification preparation.

Belt-Level Learning Paths

Load appropriate curriculum from references/training/.

White Belt (4-8 hours):

• What is Lean Six Sigma?
• Basic concepts: waste, variation, DMAIC overview
• Role in improvement projects
• Load: references/training/white_yellow_belt.md

Yellow Belt (1-2 days):

• DMAIC phases in more detail
• Basic tools: Fishbone, 5 Whys, Pareto, Process Mapping
• Data collection support
• Team participation skills
• Load: references/training/white_yellow_belt.md

Green Belt (2-4 weeks):

• Complete DMAIC methodology
• Statistical tools: capability, hypothesis testing, control charts
• Project leadership
• Load: references/training/green_belt_curriculum.md

Black Belt (4-8 weeks):

• Advanced statistics: DOE, regression, ANOVA
• Complex project management
• Mentoring Green Belts
• Organizational change management
• Load: references/training/black_belt_curriculum.md

Sample Certification Questions

Green Belt Level:

1. What are the 5 phases of DMAIC?
2. What is the difference between Cp and Cpk?
3. When would you use a P-chart vs. C-chart?
4. What is the purpose of MSA?
5. How do you interpret a Pareto chart?

Black Belt Level:

1. Explain the 1.5 sigma shift in capability analysis.
2. Design a 2³ factorial experiment for this scenario.
3. What are the assumptions for ANOVA?
4. How do you calculate control limits for an X-bar/R chart?
5. When is DMADV preferred over DMAIC?

Industry Applications

Manufacturing

Load references/industries/manufacturing.md.

Common Applications:

• Defect reduction in production lines
• Cycle time improvement
• OEE (Overall Equipment Effectiveness) optimization
• Setup time reduction (SMED)
• Machine capability studies

Specific Tools Emphasis:

• SPC and control charts
• FMEA for process/equipment
• 5S for workplace organization
• TPM for equipment reliability

Services/Transactional

Load references/industries/services_transactional.md.

Common Applications:

• Transaction error reduction
• Cycle time for approvals/processing
• Customer wait time reduction
• Call center optimization
• Invoice processing accuracy

Adaptation Notes:

• "Defect" = error, rework, exception
• Harder to see waste (information flow vs. physical)
• Process mapping is critical for visibility

Healthcare

Load references/industries/healthcare.md.

Common Applications:

• Patient wait time reduction
• Medication error prevention
• Readmission reduction
• Operating room turnaround
• Clinical pathway optimization

Special Considerations:

• Patient safety is paramount
• Regulatory compliance (HIPAA, Joint Commission)
• Evidence-based medicine integration

IT/Software

Load references/industries/it_software.md.

Common Applications:

• Defect density reduction
• Deployment frequency improvement
• Incident resolution time
• Change failure rate reduction
• Sprint velocity optimization

Integration with Agile/DevOps:

• Use Lean principles in Kanban
• Apply Six Sigma for defect reduction
• Combine with DORA metrics

Resources

Reference Files

Methodology (references/methodology/):

• 01_dmaic_overview.md - Detailed DMAIC guidance
• 02_dmadv_dfss.md - Design for Six Sigma
• 03_lean_principles.md - Lean fundamentals

Tools by Phase (references/tools-by-phase/):

• Define, Measure, Analyze, Improve, Control subdirectories
• Each contains detailed tool guides

Lean Tools (references/lean-tools/):

• VSM, 8 Wastes, 5S, Kaizen guides

Statistics (references/statistics/):

• Process capability, control charts, hypothesis testing, sigma calculation

Industries (references/industries/):

• Manufacturing, services, healthcare, IT applications

Training (references/training/):

• Belt-level curricula

Templates (assets/)

• project_charter_template.md
• sipoc_template.md
• control_plan_template.md
• a3_report_template.md
• fmea_template.md
• tollgate_review_checklist.md

Calculation Scripts (scripts/)

• sigma_calculator.py - DPMO, sigma level, yield
• process_capability.py - Cp, Cpk, Pp, Ppk
• control_chart_analysis.py - Control limits, out-of-control detection

Best Practices

Project Success Factors

1. Clear problem definition: Spend adequate time in Define
2. Data-driven decisions: Verify with data, not opinions
3. Management support: Active champion engagement
4. Cross-functional team: Include process experts
5. Realistic scope: Avoid boiling the ocean
6. Sustain improvements: Control phase is critical

Common Pitfalls to Avoid

1. Jumping to solutions: Skip analysis, implement pet solutions
2. Scope creep: Project grows beyond original charter
3. Poor data quality: Garbage in, garbage out
4. Ignoring resistance: Change management neglected
5. Weak control phase: Improvements fade over time
6. Over-reliance on tools: Tools serve the process, not vice versa

Tips for Success

• Use tollgate reviews to ensure phase completion
• Communicate progress regularly to stakeholders
• Document lessons learned for future projects
• Celebrate successes to build momentum
• Build internal capability through training