a5c-ai Root Cause Analyzer
Overview
The Root Cause Analyzer skill provides comprehensive capabilities for systematic root cause identification and verification. It supports multiple analysis methods including 5 Whys, Ishikawa diagrams, fault tree analysis, and statistical hypothesis testing.
Capabilities
- 5 Whys facilitation
- Ishikawa (fishbone) diagram creation
- Fault tree analysis (FTA)
- Hypothesis formulation
- Chi-square testing
- Correlation analysis
- Pareto chart generation
- Root cause verification
Used By Processes
- SIX-005: Root Cause Analysis
- QMS-005: FMEA Facilitation
- CI-002: A3 Problem Solving
Tools and Libraries
- Cause analysis tools
- Statistical testing libraries
- Visualization platforms
- Collaboration tools
Usage
skill: root-cause-analyzer
inputs:
problem_statement: "25% increase in customer complaints for product X"
analysis_method: "fishbone" # 5_whys | fishbone | fta | is_is_not
data:
defect_counts_by_category:
materials: 45
methods: 23
machines: 67
manpower: 12
measurement: 8
environment: 5
hypothesis_to_test: "Machine wear is causing increased defects"
outputs:
- root_cause_diagram
- potential_causes
- verified_root_causes
- pareto_analysis
- statistical_verification
- recommendations
Analysis Methods
5 Whys
Ask "Why?" iteratively to drill down to root cause:
- Why did the problem occur?
- Why did that happen?
- Why did that happen?
- Why did that happen?
- Why did that happen?
Ishikawa (Fishbone) Categories
- Man (Manpower) - Training, skills, fatigue
- Machine - Equipment, tools, technology
- Method - Procedures, processes, policies
- Material - Raw materials, consumables
- Measurement - Inspection, calibration
- Mother Nature (Environment) - Temperature, humidity, conditions
Fault Tree Analysis
- Top event (undesired outcome)
- AND gates (all inputs required)
- OR gates (any input sufficient)
- Basic events (root causes)
Verification Techniques
| Technique | Use Case |
|---|---|
| Chi-square test | Categorical data relationships |
| Correlation analysis | Continuous variable relationships |
| Regression | Predictive relationships |
| Designed experiments | Controlled cause-effect testing |
| Pareto analysis | Vital few vs. trivial many |
Integration Points
- Quality Management Systems
- Corrective action systems
- Statistical analysis tools
- Knowledge management systems