Calculation of Measurement Uncertainty Using Prior Information

Area of Science:

• Metrology and Measurement Science
• Statistical Inference
• Quality Control

Background:

• Traditional measurement uncertainty calculations often do not leverage available prior information.
• Incorporating prior knowledge can lead to more accurate and efficient uncertainty assessments.
• International standards like ISO 14253-1 provide frameworks for assessing conformance to specifications.

Purpose of the Study:

• To present a Bayesian inference approach for integrating prior information into measurement uncertainty calculations.
• To demonstrate the potential reduction in expanded uncertainty using this Bayesian method.
• To apply the Bayesian approach to workpiece conformance testing under ISO 14253-1 and explore modifications to uncertainty guard bands.

Main Methods:

• Utilizing Bayesian inference to combine prior knowledge with measurement data.
• Calculating measurement uncertainty with the Bayesian approach.
• Applying the method to assess workpiece conformance to specifications.
• Investigating adjustments to expanded uncertainty guard bands for enhanced conformance zones.

Main Results:

• The Bayesian approach can effectively reduce expanded uncertainty by up to 85% in typical examples.
• The method provides a viable framework for proving workpiece conformance according to ISO 14253-1.
• A procedure for modifying guard bands was discussed to increase the conformance zone.

Conclusions:

• Bayesian inference offers a powerful tool for refining measurement uncertainty and improving conformance assessment.
• The integration of prior information leads to more precise and potentially larger conformance zones.
• This approach has significant implications for quality control and metrology practices.