摘要
The traditional reliability inference based on the metric of mean life between failures for validating suppliers imposes insufficient deliberations on their supplied variability of unit-to-unit failure times as well as incompetent formation as a standardized metric to verify each supplier’s products reliability. To fulfill this gap in this paper, we develop a standardized lifetime-capability metric based reliability hypothesis in the production verification test, wherein the amount of the mean life between failures outlives the stipulated warranty period is evaluated by the unit-to-unit lifetime deviation. This lifetime-capability based supplier qualification allows the distinct life scales yielded from life-testing applications being standardized into lifetime-capability scores, whereby the accessible reliability–capability tables for suppliers’ validations can be easily built and enacted in practice. Furthermore, with physically accelerated Weibull-life type II testing data collected from supplier submissions, we advance statistical methodologies of the lifetime-capability metric for acquiring its lower confidence bound at a specified level of confidence to reveal the true lifetime-capability performance of the supplier. Moreover, since the lifetime-capability metric is capable of translating the supplier’s reliability capability into financial perspectives of warranty returns and vice versa, this advancement enables frontline managers to select preferred suppliers according to their acceptable warranty return rate, thereby preventing non-economic warranty costs; on the other hand, with the buyer’s warranty reserve set beforehand, the maximum allowable warranty return rate for the reserve and the minimum required lifetime-capability performance for the products are obtained to select the qualified suppliers.