Process · Reliability & Quality
Measurement system analysis
Studies that check whether a gauge, metrology tool or tester measures well enough for its purpose, by quantifying bias, linearity, stability, repeatability and reproducibility.
It is one of the automotive core quality tools, alongside APQP, PPAP, FMEA and SPC.
- Also known as
- Gauge R&R, Gage R&R, MSA
- Type
- Process
- Domain
- Reliability & Quality, Quality systems
Overview
Measurement system analysis (MSA) is a set of studies that checks whether a measurement process, meaning the instrument, method, operators and environment together, is good enough for the decisions made with its data. It quantifies how far readings sit from the true value and how much they scatter when the same item is measured again. In semiconductor work it applies to metrology tools in the fab, testers in production and gauges in assembly.
How it works
MSA separates measurement error into two families. Accuracy concerns bias, the offset from a reference value, together with how that bias changes across the measuring range (linearity) and over time (stability). Precision concerns spread: repeatability is the variation when the same instrument measures the same item several times, and reproducibility is the extra variation between operators, tools or setups. A gauge repeatability and reproducibility (gauge R&R) study estimates both by measuring several items several times on each tool or by each operator, then splitting the variation with analysis of variance.
The results are judged against the job the measurement has to do. One common measure is the precision to tolerance (P/T) ratio, comparing measurement error with the specification width; another compares it with the real spread of the process. A gauge that cannot tell good parts from bad, or cannot see the variation that SPC is meant to control, produces misleading data. Similar studies exist for pass or fail judgements such as visual inspection.
In fabs, where automated tools make most measurements, operator effects are usually small. The focus moves to load and unload effects, changes over time and matching between tools of the same type, so that a wafer measured on any tool gives the same answer.
Examples
- A gauge R&R study on a film thickness tool before its readings are used to set control limits for a deposition step.
- A matching study across several CD-SEMs so that lots can be measured on any of them without shifting the results.
- A correlation study showing that two testers and load boards give the same leakage current readings on the same parts.
- An agreement study for operators who classify visual defects on packaged parts.
Why it matters
Every control chart, capability index, test limit and yield analysis assumes the measurements can be trusted. Metrology engineers, statistical process control engineers, test engineers and quality engineers run MSA, and automotive customers expect it as one of the core quality tools that AIAG describes alongside APQP, PPAP, FMEA and SPC.
Related terms
Part of the Reliability & Quality glossary, with every term in the topic and the lifecycles, tools and standards that use them.
Where it appears on the Hub
Used on 23 pages of the Hub. In each group, the pages that mention it most come first.
Lifecycle stages and gates
Skills
Suggest an improvement
Suggest an improvement to this entry
A clearer definition, an example, a correction or a source an editor can check. Editors review every suggestion before the entry changes.
Last reviewed . Written by The Semiconductor Hub editorial team. The full glossary