Reliability metrics and the REMM model

Abstract

This paper discusses differences in the understanding and application of current reliability metrics across engineering disciplines within an aerospace environment. Differences in the interpretation of these metrics by designers, project managers and product support engineers within an organization may compromise the effectiveness of any technical or commercial decision-making that is based upon reliability prediction and field return data analysis. The same concern is true for customer and supplier interfaces with the organization, where reliability requirements and contractual guarantees, such as spares provision, are defined. The main issues appear to be related to an inadequate understanding of reliability requirement objectives. The process of negotiating and agreeing to requirements between a supplier and customer appears to be flawed in many cases by a lack of knowledge regarding the assumptions associated with the chosen metric. Holistic approaches to reliability such as REMM (reliability enhancement methodology & modeling) can provide alternative metrics from those traditionally used in the engineering community. This paper outlines the metrics available from such techniques and compares them with the more commonly used engineering metrics such as removal rate, mean time between failure (MTBF) and mean time between removal (MTBR). In conclusion, without a shared and agreed understanding of the reasons for specific reliability requirements, products may be designed to meet a customer specification without actually meeting the real reliability objectives.