How do you conduct MSA for attribute data in PPAP?

1 Define the attribute

The first step is to define the attribute that you want to measure and analyze. You need to have a clear and objective definition of what constitutes a pass or a fail for your product. For example, if you are measuring the color of a paint coating, you need to have a standard color chart or a spectrophotometer to compare your product with. You also need to specify the tolerance or acceptance limits for your attribute, such as the maximum deviation from the target color.

2 Select the measurement system

The next step is to select the measurement system that you will use to assess your attribute. The measurement system consists of the instrument, the method, and the operator that perform the measurement. For attribute data, the instrument can be a simple device, such as a go/no-go gauge, a visual inspection, or a digital sensor. The method can be a written procedure, a checklist, or a software program. The operator can be a human or a machine that follows the method and uses the instrument. You need to select a measurement system that is suitable, reliable, and consistent for your attribute.

3 Plan the MSA study

The third step is to plan the MSA study that you will conduct to evaluate your measurement system. You need to determine the sample size, the number of trials, the number of operators, and the data collection method for your study. A common approach for attribute data is to use the attribute agreement analysis (AAA), which measures the agreement among the operators, the instruments, and the standards for your attribute. For AAA, you need to select a representative sample of products that cover the range of variation for your attribute. You also need to have at least two operators and two instruments to perform the measurement on each product. You need to repeat the measurement at least two times for each operator-instrument combination. You need to record the results of each measurement as pass or fail.

4 Analyze the MSA data

The fourth step is to analyze the MSA data that you collected from your study. You need to use statistical tools, such as Excel, Minitab, or R, to perform the AAA and calculate the agreement metrics for your measurement system. The main metrics are the percent agreement, the kappa statistic, and the Kendall's coefficient of concordance. The percent agreement measures the proportion of measurements that agree with each other, the kappa statistic measures the agreement beyond chance, and the Kendall's coefficient measures the agreement among multiple raters. You need to compare these metrics with the acceptance criteria, such as 80% for percent agreement, 0.75 for kappa, and 0.8 for Kendall's coefficient, to determine if your measurement system is acceptable or not.

5 Improve the measurement system

The fifth step is to improve the measurement system if you find any issues or problems in your MSA data analysis. You need to identify the sources of variation and error in your measurement system, such as the instrument, the method, or the operator, and take corrective actions to eliminate or reduce them. For example, you can calibrate your instrument, standardize your method, or train your operator to improve the accuracy and precision of your measurement system. You can also use other tools, such as control charts, histograms, or Pareto charts, to monitor and control your measurement process.

6 Report the MSA results

The final step is to report the MSA results to your customer as part of your PPAP documentation. You need to prepare a summary report that includes the following information: the attribute definition, the measurement system description, the MSA study plan, the MSA data analysis, and the improvement actions. You also need to provide the supporting data, such as the raw data, the calculations, and the charts, as evidence of your MSA study. You need to present your report in a clear and concise manner that shows that your measurement system is capable and reliable for your attribute data.