The language of FMEA: on the effective use and reuse of FMEA data

Abstract Practical uses of Failure Modes and Effects Analysis (FMEA) range from the identification of potential design defects and safety hazards, to maintenance planning, diagnostics and Prognostics and Health Management (PHM). According to the broadly accepted standard for FMEA, MIL-STD-1629A, a successful FMEA is one that conducted in a timely manner, so that the results can be used to identify potential design flaws and thus mitigate the risk or criticality of a system in its early stages. This paper reviews and compares the current usage of terminology to define functions and failures in accordance with those defined in MIL-STD1629A. It is argued that inconsistencies in the interpretation and application of such terms has reduced the effectiveness of FMEA, and that a standardised functional and failure taxonomy can alleviate this problem. The functional and failure taxonomies proposed in this paper were created for use in an automated FMEA generation software product. The taxonomy is used to build a functional system model upon which Fuzzy Cognitive Maps are based to predict the system-wide effects of failure. The challenge was to develop a taxonomy that would achieve consistent, repeatable results in automated functional FMEA and enable a wider application and sharing of the FMEA results. The intention was to create a software package that could be used in a concurrent engineering environment, and provide reusable, shareable models for use across Design, Reliability and Maintenance (R&M) and Prognostics and Health Management (PHM) departments within an organisation.