Reliability growth planning curves based on multi-phase projections

Abstract

The planned reliability growth curve gives the interim milestone reliability targets in which the system is judged as the reliability growth program progresses across single or multiple test phases. At the end of each test phase the system reliability is assess and compared to the corresponding value on the planned reliability growth curve. If the assessed reliability is on or above the milestone value the reliability program continues. If the assessed reliability is below the milestone value changes in the reliability program may be necessary. It is therefore important that the planned reliability growth curve be realistic and based on input parameters and assumptions that adequately reflect the configuration and characteristics of the system. In many cases some of actual parameter values during testing are different than the values inputted into the planning model. These differences may or may not affect meeting the reliability goals set by the initial planning curve. However, if there is a negative impact regarding meeting the objectives in the next and future test phases then addressing this by management and engineering in a timely manner is very important. The focus of this paper is to provide a methodology to progressively update the planned reliability growth curve across all future test phases based on actual test data and give projections for the expected reliability over multiple future test phases. This methodology is currently being used to provide confidence that a program is on track or to identify risks.