Gage Repeatability and Reproducibility (Gage R&R) is a methodology used across manufacturing and process improvement initiatives, such as Six Sigma, to evaluate the quality of a measurement system. Gage R&R quantifies the amount of variation present in a measurement system, helping organizations determine if their instruments and operators are reliable enough to make informed decisions about product acceptance and process control. Understanding the sources of this measurement variation is necessary for improving quality and reducing waste.
Understanding Measurement System Analysis
Measurement System Analysis (MSA) serves as the overarching discipline that encompasses the Gage R&R study, focusing on the properties and performance of the entire system used to obtain measurements. Every measurement taken in a manufacturing or quality control environment contains some level of error or uncertainty, meaning the recorded value is not the exact, true value of the characteristic being measured. If the error within the measurement system is too large, it becomes impossible to accurately distinguish between acceptable and unacceptable parts or to effectively monitor process shifts.
The purpose of MSA is to quantify the total measurement error and separate it from the actual variation in the manufacturing process. Organizations rely on measurement data to verify product specifications and maintain process stability. When the measurement system itself is unstable or inaccurate, the resulting data is unreliable, leading to poor decisions like accepting bad products or rejecting good ones.
Defining Gage Repeatability and Reproducibility
Gage R&R is the specific tool used within MSA to quantify the total variation within a measurement system that is attributable to the measuring instrument and the individuals who use it. This statistical technique breaks down the total observed variation into components caused by the equipment and components caused by the operators. The resulting analysis provides a single metric representing the overall measurement system error relative to either the process variation or the engineering tolerance.
The objective of conducting a Gage R&R study is to verify that the measurement system possesses sufficient precision and accuracy for its intended application. A system with a low R&R score is considered stable and capable of distinguishing between different parts. Conversely, a high R&R score indicates that a significant portion of the total variation observed in the data is due to measurement error, which requires immediate attention and improvement.
The Two Components of Variation
The total measurement system variation quantified by a Gage R&R study is systematically divided into two distinct components: repeatability and reproducibility. Differentiating between these two sources of error is necessary because the required corrective actions differ depending on which component contributes more significantly to the overall variation. This separation allows quality engineers to target improvements precisely, either by adjusting the equipment or by training the personnel.
Repeatability (Equipment Variation)
Repeatability refers to the variation observed when the same operator measures the same part multiple times using the same measuring equipment. Also known as Equipment Variation (EV), this component measures the inherent precision of the measuring instrument itself. Poor repeatability often points to issues with the gage, such as inadequate mechanical design, internal wear, faulty calibration, or fixture instability.
When the repeatability value is high, it suggests the instrument is incapable of consistently providing the same reading even under identical conditions. For instance, if a single technician measures the diameter of a shaft three times and gets three vastly different results, the problem likely lies with the precision of the caliper being used. Addressing high repeatability requires focusing on the equipment, potentially through recalibration, maintenance, repair, or replacement with a more precise device.
Reproducibility (Appraiser Variation)
Reproducibility is the variation observed when different operators measure the same part using the same measuring equipment. Commonly referred to as Appraiser Variation (AV), this component measures the differences in the average measurements obtained by the people using the gage. High reproducibility indicates that the measurement results are heavily influenced by the individual taking the reading.
Differences in operator technique, training, reading method, part alignment, or fatigue can all contribute to high reproducibility. For example, if three different technicians measure the same part and consistently record different average values, the variation is likely due to inconsistencies in their application of the measurement procedure. Reducing high reproducibility typically involves standardizing the measurement procedure, developing detailed work instructions, and providing additional training to the operators to ensure a consistent application of the gage.
The Process of Conducting a Gage R&R Study
A Gage R&R study follows a standardized, controlled experimental design to collect the necessary data for statistical analysis. The most common experimental setup, known as a crossed study, involves selecting a minimum of three operators, ten parts representing the full range of process variation, and three repeated trials per part for each operator. This setup requires 90 individual measurements, which are then analyzed to separate the measurement error from the actual part-to-part variation.
Before starting data collection, the selected parts must be uniquely marked but in a way that is invisible to the operators taking the measurements. The operators must be trained on the correct measurement procedure and should be those who routinely perform the measurement in the production environment. Measurements must be taken in a completely random order to prevent operators from remembering their previous readings or anticipating the next part, which helps minimize bias in the results.
The data collection involves recording the part number, the operator identifier, and the measurement value for each trial. This data is then organized into a table format suitable for statistical analysis using methods like the Average and Range method or the Analysis of Variance (ANOVA) method.
Interpreting the Results and Acceptance Criteria
The final step of a Gage R&R study involves analyzing the collected data to derive key metrics and comparing them against industry-standard acceptance criteria. The most commonly reported metric is the percentage of the total study variation (%SV) or the percentage of the tolerance (%Tolerance) attributed to the Gage R&R. These metrics determine the capability of the measurement system to accurately reflect product variation or meet specification requirements.
Industry guidelines suggest that a total Gage R&R value of less than 10% is generally considered an acceptable measurement system. A system falling in the 10% to 30% range may be conditionally acceptable, depending on the application, the cost of the measurement, and the process capability. Any Gage R&R result exceeding 30% indicates an unacceptable measurement system that requires improvement before the data can be reliably used for quality control.
When the percentage is too high, the analysis of the two components—repeatability and reproducibility—guides the corrective action. If repeatability (Equipment Variation) is the larger contributor, the focus shifts to the measuring instrument, perhaps requiring better maintenance or a more precise gage. If reproducibility (Appraiser Variation) dominates, the organization must address the differences between people, typically by retraining operators or clarifying the standard measurement procedure.