Developer Tutorial: State-of-the Art Use of Repeated Measures Degradation to Model Product or Part Failures

This session is for JMP users who have a basic background in reliability principles and approaches. 

Repeated measures degradation analysis models degradation processes that generally contain more information than failure-time data for assessing reliability and predicting the life span of parts and systems. This type of analysis can model products that do not have hard failures or breakdown, as well as products that have hard failures that are extremely difficult to observe.

The new JMP 17 Repeated Measures Degradation platform implements the Bayesian hierarchical modeling method discussed in Meeker, Escobar and Pascual (2022). This advanced statistical approach streamlines the process of inducing failure-time distribution from degradation data with a rigorous Bayesian inference framework. The advantages of the approach include being unrestricted by large numbers theorem, reducing modeling uncertainties, and having better statistical properties that bring more confidence to decision making.

The key JMP Reliability Developer, whose team was advised by with Dr. Meeker, explains how and why JMP uses the new approach.  The session includes time for Q&A.

This JMP Developer Tutorial covers: repeated measures degradation, understanding specifications for Bayesian estimation, using or modifying the built-in specifications, and evaluating results.