Learn how to pinpoint defects in materials or processes for repairable systems and identify design vulnerabilities so you can understand the best way to correct them.
Repairable: observations depend on the previous failure time. They are only independent when the underlying distribution is Weibull and beta = 1.
Nonrepairable: observations are independent (and identically distributed). When the underlying distribution is Weibull, this relationship holds regardless of the value of beta.
See how to:
- Perform Recurrence Analysis for repairable components
- Perform Recurrence Analysis with a grouping variable
- Useful for extrapolation of system performance
- Determine spare parts inventory needs, warranty risk etc.
- Analyze Reliability Growth for concurrent or parallel repairable systems
- Good for understanding how reliability has changed as a function of design changes
- Use JMP Pro to create Reliability Block Diagrams to show reliability relationships among a system's components and illustrate how component reliability contributes to system success or failure
Q: Can I get confidence intervals/bounds for Recurrence Analysis?
A: No, not visually. However, you can specify a time or number of evetns and get confidence intervals on that using Specific Time for Cumulative and Specific Intensity and Cumulative options. See below: