Meaning
Measurement analysis method evaluates the total variation in a testing system by isolating the differences caused by the measuring instrument itself from the differences introduced by the human operators. This study is a fundamental part of the Six Sigma methodology and is used to determine if a measurement system is capable of accurately judging the quality of a product. In a typical gage r&r test, several operators measure the same set of parts multiple times using the same equipment.
The results are analyzed to calculate the repeatability, which is the equipment variation, and the reproducibility, which is the operator variation. A high percentage of variation relative to the total process spread indicates that the measurement system is unreliable. Ideally, the measurement error should account for less than ten percent of the allowable tolerance.
Measurement Stability
Reliability of the test equipment is the first component analyzed in the study. During the gage r&r procedure, the term repeatability refers to the ability of the same person to get the same result when measuring the same part several times. If the instrument is sensitive to vibration, temperature or electrical noise, the readings will fluctuate even when nothing else has changed.
A lack of repeatability suggests that the equipment might be worn out, poorly calibrated or simply not precise enough for the task. This variation is purely mechanical or electronic in nature and cannot be fixed by training the staff. Identifying this base level of noise is the first step in improving the testing process.
Operator Reproducibility
Human element of the measurement process often introduces significant errors that are harder to control than equipment drift. The reproducibility part of the gage r&r study focuses on the differences between the averages of measurements made by different people. If one technician uses a caliper differently than another, the results will shift.
These differences can be caused by varied levels of experience, different interpretations of the test procedure or physical factors like eyesight. A high reproducibility error indicates that the standard operating procedure is not clear enough or that the operators need more training. In automated systems, this value should be near zero, as the machine performs the task the same way every time.
Variance Contribution
Total error in the system is the sum of the squares of the repeatability and reproducibility components. This combined value is then compared to the total variation seen in the manufacturing process to see if the gage r&r results are acceptable. If the measurement system is too noisy, it will reject good parts or, worse, accept bad ones.
This leads to wasted money and potentially dangerous product failures in the field. Managers use the results of this study to decide whether to invest in new measuring tools or to refine the existing test methods. A successful study provides the confidence that the data being collected on the factory floor is a true reflection of the product quality.
This verification is a requirement for any supplier entering a high volume production contract.
