Dynamic reliability prediction: how to adjust modeling and reliability growth?

A dynamic reliability prediction technique that has been applied to the DPS 7000, a mainframe computer system, is described. To construct a product reliability prediction that is applicable over the life cycle of the product, both the intrinsic technological characteristics and their evolutionary trends are considered. Improvements in such factors as organization, manufacturing, maintenance support, and customer familiarity can have a substantial influence on product reliability growth. Using observed measure of field performance, trends in reliability growth that are attributable to evolving product maturity can be identified and used to support dynamic adjustments to component reliability and reliability growth models, and thereby create dynamic reliability predictions.<>