Data Quality Requirements Methodology for an Adapted PHM Implementation

The technologies of digitization allow organizations to rely on data mining for performance improvement. In this context, data-driven Prognostics and Health Management (PHM) is being introduced as a new framework for data management and knowledge extraction. However, the collected data are generally accompanied by quality issues that influence PHM results. Metrics are therefore needed to quantify data suitability for PHM application. The majority of existing works propose to improve PHM tools without taking into account the adequacy of the used data to the fixed objectives. This paper aims to propose a set of data quality requirements for PHM applications and in particular for the fault detection task.