Comment mettez-vous à jour et entretenez-vous votre système MSA au fil du temps?

O MSA (Measurement System Analysis) é uma ferramenta que permite a análise do sistema de medição de forma completa levando em conta características associadas tanto ao equipamento quanto ao processo de medição composto pelo operador e pelas demais variáveis. A sua aplicação nos permite entender restrições do sistema como em qual "espaço de inferência" (variantes e intervalo da variável resposta) atuamos e quais melhorias precisam ser implementadas para reduzirmos o erro de medição considerando futuros resultados.

Precise measurements are the foundation on which informed decisions are made, driving efficiency, quality, and ultimately, organizational excellence. Whether assessing product dimensions, analyzing process performance, or identifying areas for improvement, the reliability of measurements directly impacts the effectiveness of strategies used. Maintaining a robust MSA system is critical, ensuring that data-driven insights are not only reliable but also actionable.

Measurement System Analysis (MSA) is a crucial tool utilized across various industries, including automotive, to ensure the reliability of measurement systems. It is mandated by standards like IATF16949:2016 and is integral to processes like PPAP and DMAIC in Six Sigma methodology. It involves studies to check precision and accuracy of the measurement system.

A Análise do Sistema de Medição é uma ferramenta capaz de avaliar o procedimento e o instrumento que serão utilizados durante uma coleta de dados. Ele gera um nível de aprendizado relevante diante das variações do processo gerador da entrega e do modo como operamos um equipamento de medição. Através dele podemos observar 5 características principais: acuracidade, repetibilidade/precisão, discriminação, reprodutibilidade e estabilidade. Elas são verificadas a partir de gráficos e cartas de controle, onde determinados comportamentos indicam maior variação de medida a medida ou entre amostras, fazendo com que possamos determinar diferenças e instabilidades. Com o estudo, evitam-se desperdícios, principalmente em coletas de dados futuras.

Measurement System Analysis (MSA) is a comprehensive evaluation method used to determine the adequacy of a measurement system. It focuses on the sources of variation in measurements, including repeatability, reproducibility, bias, stability, linearity, and precision to tolerance. MSA is pivotal in assessing the capability of your measurement system to produce consistent, accurate data over time. Enhancing Understanding Depth of Analysis: MSA goes beyond surface-level assessments to deeply understand measurement errors and their origins. System Components: Evaluates every component of the measurement system, including instruments, operators, methods, and the environment, ensuring a holistic analysis.

As said by Peter Drucker "If you can't measure it, you can't manage it". And the only way to make sure you are measuring it correctly is through a Measurement Systems Analysis. MSA is an important step, otherwise, all your project assumptions can be wrong.

MSA helps us to check the accuracy and precision of our measurement system. Unless these parameters are correct, our measurements won't give correct results and the same results every time. As a result, there will always be a risk of passing a defect to the customer, rejecting good products (unless our measurement system is accurate and precise).

Some people say the data never lies but that is not always true. Sometimes you can have bad misleading data. The data usually doesn't lie but if you have garbage data it can mislead you. Make sure you can trust the data. Even data can lead you astray if it is not accurate.

MSA’s importance cannot be overstated in Six Sigma and other quality management systems. It validates the accuracy and reliability of the data collected, which forms the basis for all further analysis and decision-making processes. Impact on Quality and Decision-Making Data Integrity: Ensures that decisions are based on reliable data, safeguarding against costly mistakes or misinterpretations. Process Improvements: Identifies areas where the measurement process itself can be enhanced, leading to overall process improvements

MSA evaluates the quality of the measurement system or instrument. Organizations typically consider measurement data when making decisions about the process or business. Therefore, the data must be accurate. The accuracy of the data is one of the fundamental requirements in the DMAIC process (Define, Measure, Analyze, Improve, Control), which is a central methodology of Six Sigma. Monitoring and measuring resources are an important requirement in quality management systems, such as ISO 9001, AS9100, etc. By evaluating bias, linearity, stability, and repeatability, MSA tools help manufacturers identify and correct measurement errors, improving process control and product quality.

Start with measurement process mapping. If you are not an expert on the measurement process, go to the Gemba and learn with the process experts. With process steps clearly mapped, process factors identified, and well classified (controllable, noises, standard operations - naming conventions may vary from company to company) you can start thinking about your sampling strategy. But don't forget to keep in mind your objectives! Are you looking for measure-to-measure error estimation? Do you want to compare multiple measurement processes like different labs, equipment, and operators? Are you just trying to identify a bad sample in a batch? Is it a consistency over time issue? Those questions may lead to different experimentation approaches!

- MSA involves studies such as Bias, Linearity, Stability, Repeatability, and Reproducibility for variable measurement systems. Bias, Linearity and Stability studies are used to measure/check the accuracy of the measurement system. Repeatability and reproducibility studies are used to check precision. -For attribute measurement systems, a Kappa study assesses agreement between appraisers and reference values, along with measures like effectiveness, miss rate, and false alarm. These evaluations collectively determine the acceptability of the measurement system. Effectiveness, miss rate and false alarm measures are used to check the accuracy of the attribute measurement system. And Kappa is used to measure precision.

Conducting MSA involves a series of structured steps, starting from defining the measurement system to collecting data, analyzing it, and interpreting the results. Key components include gauge repeatability and reproducibility (GRR) studies, bias, linearity, and stability assessments. Approach Preparation: Clearly define what is being measured and identify the measurement equipment, operators, and the process for measurement. Data Collection: Use a structured approach to collect data, ensuring a broad representation of the variation you expect in the actual process. Analysis: Employ statistical methods to analyze the data, focusing on identifying the sources and magnitudes of measurement system variation. Interpretation

I prioritize regular audits to refine my MSA system. I stay updated through continuous learning and industry research, ensuring my methods align with the latest standards. By seeking feedback from frontline staff, I gain valuable insights. Documenting changes maintains transparency and ensures everyone's on board. This approach not only sharpens my skills but also fosters a culture of excellence in measurement accuracy.

Improving MSA involves addressing identified sources of variation and error. This can include recalibrating equipment, retraining operators, or modifying measurement procedures to enhance reliability and accuracy. Continuous Improvement Strategies Equipment Maintenance: Regularly check and calibrate measurement equipment to ensure its accuracy. Operator Training: Provide continuous training and certification for operators to minimize human error. Procedure Standardization: Standardize measurement procedures to reduce variability caused by different methods or conditions.

There is no standard operation here: the strategy for improving your measurement system must be based on your MSA results. Different actions will be required according to MSA results. If you identified that the biggest source of variations comes from operator to operator, maybe training sessions will be enough. If you identified that your measurement process is not capable of identifying differences between good parts and bad parts, maybe you need more accurate equipment.

Instrumentation is often calibrated on a one year cycle. By monitoring the drift and analyzing with Weibull Analysis, calibration cycles may be adjusted. Maybe shorter if demonstrated, or maybe longer if instrumentation remains stable.

- Frequently MSA studies will have to be repeated to check consistency of the measurement system and its performance - The changes/repairs to measurement systems will have to be tracked and studied, which might require repeating MSA study

Where possible, you can measure a "golden sample" or a known standard periodically and put the results on a Control Chart. This helps assure your system is still calibrated and the precision and accuracy haven't changed since your last MSA.

Maintaining an MSA system requires ongoing attention and updates to adapt to changes in measurement processes, equipment, or the items being measured. Regularly conducting MSA studies helps in early detection of issues and ensures the measurement system remains reliable over time. Maintenance Best Practices Periodic Reviews: Schedule and conduct MSA studies regularly or when significant changes occur in the measurement process or system. Control Charts: Use CC to monitor the measurement system’s performance continuously, allowing for timely adjustments. Audits and Feedback loops: Implement audits to ensure adherence to MSA standards and procedures. Establish FL to incorporate insights from MSA studies into continuous improvement efforts.

You can repeat and check consistency (as mentioned previously) which honestly will require time and people allocation. Or... you can try to automate the data collection of a few measurement process factors you have previously identified as critical ones. If you are capable to collect and plot this data in control chars, you will increase your confidence in your measurement process. Consider your measurement processes as any other processes (identify their inputs, factors, outputs, ...), and I'm pretty sure you will be able to control the important factors instead of only repeating MSA studies year over year.