Data-Driven Predictive Maintenance: A Methodology Primer

Predictive maintenance aims at proactively assessing the current condition of assets and performing maintenance activities if and when needed to preserve them in the optimal operational condition. This in turn may lead to a reduction of unexpected breakdowns and production stoppages as well as maintenance costs, ultimately resulting in reduced production costs. Empowered by recent advances in the fields of information and communication technologies and artificial intelligence, this chapter attempts to define the main operational blocks for predictive maintenance building upon existing standards and discuss key datadriven methodologies for predictive maintenance. In addition, technical information related to potential data models for storing and communicating key information are provided, finally closing the chapter with different deployment strategies for predictive analytics as well as identifying open issues.