Reliability Growth Prediction
Abstract
A reliability growth data base on nine avionic systems and thirty equipment items was compiled for the study. Several growth models were investigated and three were selected for application to the data: The Duane, AMSAA and IBM models. Goodness-of-fit tests for the models were applied and relationships between growth model parameters and equipment characters/program attributes were investigated. Both the Duane and AMSAA models were found to yield reasonably good fits to the data sets. Both models, however, were found to have limited utility as predictive tools. Anomolies in the growth data, such as burn-in effects and time delays in incorporating corrective actions tended to mask out any relationships between the growth parameters and equipment characteristics or program attributes. The IBM model was found to provide a more workable methodology for growth prediction. A prediction procedure was developed, based upon IBM model, which uses engineering information available at the start of the growth program. Guidelines for conducting reliability growth tests and examples of the reliability growth prediction procedures are provided in the report.
Document Details
- Document Type
- Technical Report
- Publication Date
- Sep 01, 1986
- Accession Number
- ADA176128
Entities
People
- G. J. Gibson
- K. K. Mclain
- R. K. Gates
Organizations
- TASC, Inc