Methodologies for Fielded Repairable System Data Analysis
In addition to extensive capabilities for reliability growth analysis, RGA also provides opportunities to apply the non-homogeneous Poisson process (NHPP) model for fielded repairable system analysis. The analysis options will vary depending on the characteristics of your data set and the questions that you need to answer.
Analyze the Reliability for Fielded Repairable Systems
Distribution analysis techniques (such as "Weibull analysis") are useful for modeling times-to-failure of non-repairable items in order to understand the reliability over time. However, when you are considering repairable systems that might fail and be restored multiple times, different analysis approaches should be applied. RGA's "Repairable" data type facilitates a methodology based on the non-homogeneous Poisson process with a Power Law failure intensity. This analysis helps you to understand the system's reliability behavior over time and calculate metrics of interest without the detailed data sets that normally would be required.
- Optimum overhaul time for a given repair cost and overhaul cost
- Conditional reliability, MTBF or failure Intensity for a given time
- Expected number of failures for a given time
- Time for a given conditional reliability, MTBF or failure intensity
Analyze Data from a Fleet of Systems Operating in the Field
RGA also offers options for analyzing data from a fleet of systems operating in the field. For example, the "Fleet" data type (along with the Crow Extended model) can be employed when you need to evaluate the reliability improvement that you can expect from rolling out fixes for a fleet of units operating in the field.
In addition, if you have warranty data from a fleet of fielded systems, you can use grouped (interval) data analysis in order to calculate the trend in the MTBF (or failure intensity) for the population and then use that information to make estimates about the returns that can be expected in the future.
Design of Reliability Tests for Repairable Systems
Design of Reliability Test (DRT) methods that are based on the parametric binomial, non-parametric binomial or exponential Chi-Squared methods are suitable for non-repairable items. However, when you want to design a reliability demonstration test for a repairable system that may fail and be restored multiple times during operation, another method is required.
With the release of Version 7, RGA now provides a DRT utility based on the NHPP model, which is suitable for repairable systems. This tool will help you to determine the amount of test time (or number of test units) that will be required to demonstrate a specified reliability goal (defined in terms of MTBF or failure intensity at a given time).
