Verify the evidence source
Use Gage R&R to estimate repeatability, reproducibility, and the system’s ability to distinguish parts.
Topic / Quality Engineering
Measure first.Decide with evidence.
Quality engineering connects three different questions: can the measurement be trusted, is the process stable, and can that stable process meet specification? This hub organizes every existing Mechatrovich article, tool, product, and service around that decision sequence.
INTRODUCTION
One discipline. Three connected decisions.
Measurement System Analysis identifies how much observed variation comes from the measurement process. Statistical Process Control examines behavior over time. Process capability compares the stable distribution with engineering specification limits. The pages below connect the explanation, worked evidence, calculation tools, complete products, and higher-touch review options already available on Mechatrovich.
01 / MSA
Measurement Systems Analysis
Determine whether the measurement system can distinguish real part-to-part variation before using the data to judge the process.
OVERVIEW
Gage R&R Basics
Repeatability, reproducibility, %GRR, ndc, and the reasoning behind acceptance.
Read the MSA overview →WORKED STUDYGage R&R Example
A complete crossed study using 90 measurements, ANOVA, diagnostic charts, and an acceptability decision.
Review the worked evidence →DIAGNOSTICNDC Below 5
Why ndc can remain low even when %GRR appears acceptable—and what to check first.
Diagnose a low ndc result →INTERPRETATIONAcceptable Gage R&R, Incapable Process
Why a trustworthy gage can correctly report that the process does not meet specification.
Separate measurement from capability →02 / SPC
Statistical Process Control
After measurement adequacy is understood, use process order and rational subgroups to separate routine variation from special-cause signals.
OVERVIEW & EXAMPLE
Xbar-R Chart Basics
Subgroup logic, A2/D3/D4 calculations, control limits, and a worked 25-subgroup example.
Learn the Xbar-R method →INTERPRETATIONIn Control but Cpk Still Low
A clean control chart means stable behavior—not automatic specification fit.
Separate stability from capability →03 / CAPABILITY
Process Capability
Once the data is trustworthy and the process is stable, compare process spread and centering with the engineering specification.
OVERVIEW & EXAMPLE
Cp vs Cpk
Potential capability, actual centering, formulas, graphs, assumptions, and practical interpretation.
Understand Cp and Cpk →CROSS-METHOD DECISIONGood Gage, Poor Capability
A passing measurement study and a low Cpk answer different questions and can both be correct.
Read the combined interpretation →CROSS-METHOD DECISIONStable Process, Poor Capability
Why statistical control does not prove that the distribution fits specification limits.
Read the stability comparison →Engineering workflow
Measurement → Stability → Capability
Move through the workflow in order. Each step reduces a different uncertainty before the next conclusion is made.
→
Verify process behavior
Use control charts to determine whether the process is behaving consistently over time.
→
Evaluate specification fit
Use Cp and Cpk to compare the stable process distribution with the specification limits.
Do not collapse the questions. An acceptable Gage R&R does not make a process capable. A stable control chart does not guarantee a good Cpk. Each result must be interpreted within the question it actually answers.
Featured tools
Apply each method with the right working surface.
Use a focused calculator for a quick check, a complete tool for repeated analysis, or the combined suite for the full quality-engineering sequence.
MSAMSA Studio
Crossed Gage R&R analysis with ANOVA, Average & Range, diagnostic charts, Excel, and local HTML.
View MSA Studio →GRR Calculator
Calculate %GRR, %Tolerance, repeatability, reproducibility, and ndc using the Average & Range method.
Open the calculator →
SPCSPC Studio
Build Xbar and R charts, calculate control limits, and review stability signals.
View SPC Studio →
CAPABILITYCpk Dashboard
Analyze process spread, centering, and capability against specification limits.
View Cpk Dashboard →
Featured services
When the output exists but the decision is unclear.
Use an interpretation review for one existing result or the full analysis sprint when the dataset still needs to be analyzed.
24-HOUR REVIEW
View serviceResults Interpretation Check
Submit one existing MSA, SPC, or capability output for focused interpretation, risks, limitations, and prioritized next checks.
48-HOUR ANALYSIS
View serviceQuality Analysis Sprint
Submit one manufacturing dataset for a fixed-scope MSA, SPC, capability, or focused combined analysis.
Recommended learning path
Build the reasoning in sequence.
Follow the core pages below before choosing the tool or service that matches the engineering task.
01
Understand Gage R&R
Learn what repeatability, reproducibility, %GRR, and ndc actually measure.
02
Inspect a complete measurement study
Follow the evidence through ANOVA, diagnostic charts, and the acceptability decision.
03
Learn process stability
Understand rational subgrouping, Xbar and R limits, and signal interpretation.
04
Understand specification fit
Compare potential capability, centering, and actual process capability.
05
Connect the decisions
See why acceptable measurement, stability, and capability remain separate conclusions.
Frequently asked questions
Keep the questions separate.
The sequence is simple, but each method has a different engineering purpose.
Where should I start with quality engineering?
Start with Measurement System Analysis, then check process stability with SPC, and only then interpret process capability. This prevents measurement noise or unstable behavior from being mistaken for a capability conclusion.
Why does MSA come before SPC and Cpk?
SPC and capability calculations use measured data. Gage R&R helps determine how much observed variation comes from repeatability and reproducibility before that variation is attributed to the manufacturing process.
Can a process be stable and still have a low Cpk?
Yes. A control chart evaluates whether process behavior is consistent over time. Cpk evaluates how the stable distribution fits the specification limits. A process can be consistently too wide or off-center.
Can Gage R&R be acceptable while the process is not capable?
Yes. An acceptable Gage R&R result supports confidence in the measurement. It does not guarantee that the process mean and spread meet the specification.
Which free tool should I use first?
Use the GRR Calculator when you already have a crossed Average & Range study and need a quick measurement-system check. Use the limited working previews on the SPC Studio and Cpk Dashboard pages for process-stability and capability examples.
When should I use the Results Interpretation Check instead of the Quality Analysis Sprint?
Use the Results Interpretation Check when an MSA, SPC, or capability output already exists and the main need is interpretation. Use the Quality Analysis Sprint when the manufacturing dataset still needs the analysis, charts, and engineering conclusions to be produced.