Zone Chart Pattern Recognition
Advanced pattern recognition using Western Electric rules and zone analysis for comprehensive statistical process control monitoring.
Overview & Key Features
Zone Boundaries & Formula Focus
Zone boundaries partition the control chart into bands of one standard deviation each, forming the basis for all eight Western Electric pattern-detection rules.
Zone boundaries relative to the centerline
= centerline (process mean)
= process standard deviation
Reading the Zone Chart
Pattern Interpretation Guidelines
Signs of Control
- • Points randomly distributed around centerline
- • Approximately equal points above and below centerline
- • Most points (≈68%) within Zone C
- • No systematic patterns or trends
- • Variation appears random and stable
Signs of Special Causes
- • Any Western Electric rule violation
- • Obvious trends or cycles
- • Clustering of points
- • Sudden shifts in level
- • Too many or too few points near centerline
When to Investigate
- • Immediately: Any rule violation occurs
- • High Priority: Multiple rule violations
- • Medium Priority: Approaching violation thresholds
- • Low Priority: Unusual but non-violating patterns
Common Causes of Patterns
- • Trends: Tool wear, operator fatigue
- • Shifts: New material batch, setup change
- • Cycles: Environmental conditions, maintenance
- • Clustering: Measurement system issues
Best Practices
- • Use automated pattern detection software when possible
- • Train operators to recognize common patterns visually
- • Document all investigations and corrective actions
- • Review patterns periodically for systematic improvements
- • Consider multiple rules simultaneously for comprehensive monitoring