The impact of variable hazard rates on in-service reliability

Abstract

Abstract

This paper uses various models to assess the impact of variable hazard rates. It shows that the best interests of safety, operational effectiveness and cost are met by the identification and elimination of early failure subgroups, regardless of their apparently benign mean times to failure. It supports the observation that hazard rates that increase with usage reflect a design feature, while those that decrease with usage are associated with quality shortfalls. A high-level model is then introduced that identifies priorities for detailed analysis. It then recommends a more vigorous approach to demonstrating that certification requirements are being met and upheld in the civil aviation industry. This would require a more formal analysis of the current mandatory records of maintenance actions and operations. Although this paper has been written principally around civil aviation, the analytical techniques are equally valid to other applications.