Continuous Gage R&R concepts

Learning objectives

• Design a crossed continuous Gage R&R study.
• Interpret variance components and % study variation.
• Improve the measurement system when it dominates variation.

What the study estimates

A crossed Gage R&R separates observed variation into part-to-part, repeatability and reproducibility components. The goal is to know how much of what you ‘see’ is really the process versus the gauge.

Free spreadsheets can compute the standard variance components; the guideline is <10% of study variation is acceptable, 10–30% marginal, >30% unacceptable.

Garment-factory example

Three technicians measure ten polos twice each for chest width using the approved fixture, with prior results concealed and order randomised.

Method

1. Select 10 parts that span the process range.
2. Use 2–3 trained appraisers and 2–3 randomised trials.
3. Blind each trial from previous results.
4. Compute components and % study variation and act on the largest.

Common mistakes

• Using consecutive identical parts.
• Letting appraisers see prior readings.