Introduction:

Reliability prediction involves estimating the reliability of equipment or products prior to their production or modification.  The Center for System Reliability has conducted several reliability prediction studies for both industry and government customers.  Successful reliability prediction generally requires developing a reliability model of the system.  The level of detail of the model will depend on the level of design detail available at the time.  Data required to quantify the model is obtained from sources such as company warranty records, customer maintenance records, component suppliers,   or expert elicitation from design or field service engineers.  Reliability prediction combines rigorous analysis procedures with expert judgment to develop a realistic estimate of product performance.

                         Figure 1. Comparison of Predicted Vs Observed MTBF                  

Benefits:

• Provides an early indication whether a design is likely to meet reliability goals,
• Points to potential reliability problem areas in a new design or design modifications,
• Identifies areas where additional data are needed, and
• Identifies components needing further testing.

 Capabilities:

• Fault tree analysis and failure modes and effects analysis,
• Reliability data collection and expert elicitation,
• Design trade-off studies,
• Reliability optimization, and
• Cost-benefit analysis.