Track 2 Tutorial 1

3:30 to 5:00 p.m. Tuesday, June 15, 2010

Methods for Planning and Continually Assessing a Multi-Phase Reliability Growth Test Program

For most reliability growth testing programs, corrective actions would be expected to be incorporated into the system at several predetermined times during the testing. Generally, the testing is stopped at these times and several corrective actions are incorporated into the system as a group. These stopping points often define the test phases and there are usually interim reliability goals to be met for each test phase.

This tutorial discusses the Extended Reliability Growth Planning model for determining these interim goals and the test time necessary in order to meet the final reliability goal or requirement. Once a reliability growth testing program has started, we want a model that will utilize all the data across the test phases to best use the entire data set for valid assessments.

This tutorial also discusses the Extended Continuous Evaluation reliability growth model for conducting reliability assessments multiple times during the test program and, in particular, at the end of each test phase. Each time an analysis is conducted, the Extended Continuous Evaluation model allows for several key metrics to be calculated, including growth potential and projections that support the intent of the 2008 Defense Science Board Task Force recommendations. At the end of each test phase, the Extended Continuous Evaluation model uses all test data to date for these analyses and consequently provides timely and valid metrics to better manage the reliability growth test program. The practical, real-world applications of these models and concepts are illustrated by numerical examples.

Key Words: Extended Reliability Growth Planning Model, Extended-Continuous Evaluation Model, Multi-Phase Reliability Growth Testing, Growth Potential, Defense Science Board Task Force Recommendations

Larry H. Crow

Crow Reliability Resources, Inc.

Madison, Alabama