Weibull time-to-fail distributions cannot be correctly estimated from field data when manufacturing populations from different vintages have different failure modes. To investigate the pitfalls of ongoing Weibull parameter estimation, two cases, based upon real events, were analyzed. First, a time-to-fail distribution was generated assuming the same Weibull shape parameter representing an increasing failure rate for each monthly batch or vintage of production. The shape parameter was estimated from simulated field data at regular periods as the population accumulated service time. Estimates of the shape parameter were not constant, but gradually decreased (as had occurred in a real system) with added service time. In the second case, field reliability performance was modeled to match the actual historical data for one product from a disk drive manufacturer. The actual data was proprietary and was not directly available for analysis. A production schedule was modeled with a mix of two failure characteristics. The population reaching the field in the first 12 months had a low, constant failure rate. For the second and third years of production, higher volumes were introduced that had the higher, increasing failure rates of the first case. Assessment of the mixed population at each month of calendar time resulted in an increasing Weibull shape parameter estimate at each assessment. When the two populations were separated and estimated properly, a better fit with more accurate estimates of Weibull shape parameters resulted.
Truth emerges more readily from error than confusion' FRANCIS BACON SUMMARY As a science, reliability has now entered middle age, having achieved almost 40 years of recognized modern practice. As we move into the new decade of the 1990s it is appropriate that we review the status of modern reliability. The history of science has lessons for us to learn concerning the nature of paradigm changes. Clearly some reliability practitioners have had difficulty changing their own world views as scientific knowledge has increased and the conditions of the profession have changed.The very first generalized model for reliability was based upon electron tube life data from the early 1950s. It was with these old types of complex and failure prone products upon which the original reliability model was developed and generalized. This first model dictated that the failure probability density of electronics follows the exponential law which implies that the electronics will show constant failure rates during their useful lives. To get to the constant failure rate period, an infant mortality was traversed and about 10 per cent failures observed. The reliability beliefs of the 1950s when combined with the product successes of the 1960s has created a reliability paradigm problem that first became apparent in the late 1970s. Improved quality and design effort with new technologies and knowledge about effective screening changed the conditions. Modern semiconductor electronic products d o not follow the original electron tube reliability model. The applicability of this original model and the subsequent thinking that it led to must now be questioned. It is time t o create a new and better paradigm to replace the defunct exponential law. KEY WORDS Reliability paradigm Exponential law Constant failure rate Decreasing failure rate Roller coaster curve
scite is a Brooklyn-based startup that helps researchers better discover and understand research articles through Smart Citations–citations that display the context of the citation and describe whether the article provides supporting or contrasting evidence. scite is used by students researchers from around the world and is funded in part by the National Science Foundation and the National Institute on Drug Abuse of the National Institutes of Health.