Reliability Prediction - An Engineering Appraisal

This paper examines the state of health of reliability prediction and in the context of the progress achieved over the past eight to nine years finds it lacking in several important respects, some of these being accuracy as related to the requirements for it, and a lack of precision in the basic failure rate data and its application.

The author focuses attention on six major activity areas wherein effort should be initiated or greatly expanded to correct the existing situation. These include establishment of a standard “benchmark” for failure rate data, a new approach to the prediction problem for mechanical items, evaluation of process variations, and application of quality data.

The author closes with the admonition that in view of the highly projectized thinking attending the growth of military and aerospace activities today, reliability and quality organizations and people must assume an innovative role in new areas if substantive progress is to result.