Capability confirmation after improvement

Learning objectives

• Explain the purpose of capability confirmation after improvement in a garment-factory improvement project.
• Apply the described method to a representative shop-floor situation.
• Recognize the common mistakes and how to avoid them.

Concept

Post-improvement capability confirms a stable process can meet specifications with enough margin under sustained conditions.

Cp/Pp assess spread; Cpk/Ppk include centering.

Use correct specifications, representative data, distribution review, confidence/uncertainty, and practical defect evidence.

Variables, units, and assumptions

Y in engineering units; USL/LSL; sigma_within (short-term) vs sigma_overall (long-term).

Process stable, roughly normal (or transformed), representative sample ≥ 30-50 subgroups.

Free-spreadsheet workflow

Free spreadsheet: Cp=(USL-LSL)/(6σ_within); Cpk=min((USL-μ),(μ-LSL))/(3σ_within); Pp/Ppk use σ_overall.

Interpretation and limitations

Cpk ≥ 1.33 typical minimum; ≥ 1.67 for critical CTQs.

Non-normal data need transform or non-parametric capability; small n gives unstable estimates.

Garment-factory example

Confirm chest measurement capability across sizes, lines, shifts, and material lots after pattern and method changes.

Method

1. Confirm gauge and chart stability.
2. Collect sustained data.
3. Stratify.
4. Calculate Cp/Cpk/Pp/Ppk in a free spreadsheet.
5. Inspect tails.
6. Compare with charter goal.

Common mistakes

• Computing capability on unstable pilot data.
• Pooling distinct size tolerances into one calculation.