The purpose of this research is to explore the relationship among multivariate process capability indices, expected losses and design reliability for the nominal-the-better case so the likelihood and consequence resulting from the failure of a multi-characteristic engineering design can be evaluated simultaneously. This new approach provides reliability practitioners a decision making tool in performing the quantitative evaluation of reliability improvement for a multi-characteristic engineering design. Both the impact of process capability indices on the expected loss and its associated reliability can be clearly understood. Finally, a realistic example in geotechnical application is given to demonstrate how the expected losses can be estimated by the multivariate process capability indices. Then, practicing reliability managers and engineers will be able to accurately evaluate the reliability improvement results of a multi-characteristic engineering design by conducting a decision tree analysis.
All Science Journal Classification (ASJC) codes
- Computer Science Applications
- Artificial Intelligence