Abstract

Maintenance is defined as the actions that allow machines and equipment to work for an extended period of time by retaining and
restoring equipment to its original state. In Industry 4.0 context, Predictive Maintenance (PdM) is a strategy that utilizes digitized
sensor data and data analytics to continuously monitor the state of machine components or processes to determine when and where
maintenance actions may be required. There are five key types of PdM techniques being used in practice: experience-based,
model-based, physical-based; data-driven; and hybrid. Selecting the most suitable PdM technique for a given setup or scenario is
critical for any successful PdM implementation in industry to optimize cost and time. To help businesses in identifying and
selecting the most appropriate PdM technique for their specific purposes, the authors propose a corresponding decision-making
framework based on several critical factors to be considered in the process. They also discuss how the framework might best be
used in industrial strategic planning processes and elaborate on its limitations and challenges.