That Day, After Submitting the PPAP, the Client Replied with Just Four Words: "MSA Resubmit"
That day, we finally managed to prepare a PPAP package and eagerly submitted it to the client. This report took nearly a month to compile, from FMEA to Control Plan, with every single item checked until our eyes were blurry. And what happened? The client's email came back with the subject "PPAP Review," and the body contained just four short words: "MSA Resubmit." Damn! I almost spat out my coffee right then. My internal monologue was, "I've been doing this for over a decade, how could I still mess up MSA?" But the fact was, the client wasn't buying it!
Where Did the Problem Lie? It's Not That You Measured Inaccurately; It's That the Client Believes You Did
Simply put, MSA (Measurement System Analysis) is about demonstrating that your measurement system is reliable. Today, when you ship products to clients, you use various instruments to measure product dimensions, electrical properties, and appearance. But how does the client know if your instruments are working correctly? Are the measured data truly accurate? What if you measure 100, but when they take it back and measure, it turns out to be 80? This is the core problem MSA aims to solve.
So, the key point is: The purpose of MSA is to build client confidence in your measurement data. Whether the data is accurate or not isn't up to you to say; it needs to be proven with data.
How is it Actually Done? It Depends on Whether It's R&R or Bias & Linearity
Frankly, the most common MSA reports required for PPAP submission are Gauge R&R (Gauge Repeatability and Reproducibility) and Bias & Linearity.
- Gauge R&R: This checks whether your measurement system produces significantly different data when different people measure the same product repeatedly at different times. Typically, a %GRR less than 10% is considered excellent, 10% to 30% is acceptable, and over 30% is problematic. For example, if you measure a product with a resistance of 100 ohms, and three engineers each measure it three times, if every measurement falls between 99.8 and 100.2, the %GRR will be very good. But if someone measures 95 and another measures 105, then there's definitely an issue with the system.
- Bias & Linearity: Bias is essentially an offset. Does your measurement system consistently measure slightly more or slightly less each time? For instance, if the true value is 100, and you consistently measure 102, that's a Bias. Linearity, on the other hand, examines whether this Bias changes across different measurement ranges. If the Bias is small at lower points but large at higher points, then Linearity is poor.
In other words: R&R looks at "precision," while Bias & Linearity looks at "accuracy." Both are crucial, and it depends on what the client requires.
The Most Common Pitfalls: Wrong Gage Selection, Insufficient Samples, or Simply Not Understanding
The most outrageous case I encountered involved a novice engineer. The client requested an MSA for a critical dimension. He simply took a caliper from Quality Assurance to measure it and submitted a caliper R&R report. The client immediately rejected it. Why? Because the tolerance for that dimension was only ±0.01mm, and the caliper's measurement precision was simply insufficient! Using an instrument with inadequate precision to measure, even if the R&R data looks perfect, is futile.
Another time, a new supplier submitted a PPAP, and its R&R report had a pitifully small sample size: only three operators measuring two parts. The DPMO came out as 6210, and the Cpk was only 1.08, but the client required a Cpk greater than 1.33. Submitting such data was a complete waste of time.
So, the key point is: Before conducting MSA, first understand your measurement specifications, tolerances, and the client's requirements for instrument precision. Then, select the correct gage and ensure a sufficient sample size.
One Thing You Can Do Today
Go back and check your measurement instruments to see if they have the latest calibration reports. This is a fundamental requirement for MSA!