[Paper Review] A Survey on Predictive Maintenance for Industry 4.0

Christian Krupitzer, Tim Wagenhals+7 more|arXiv (Cornell University)|Feb 5, 2020

Reliability and Maintenance Optimization50 references50 citations

TL;DR

This structured literature review classifies predictive maintenance in Industry 4.0 and builds a 10-category framework from 150 papers to guide future research.

ABSTRACT

Production issues at Volkswagen in 2016 lead to dramatic losses in sales of up to 400 million Euros per week. This example shows the huge financial impact of a working production facility for companies. Especially in the data-driven domains of Industry 4.0 and Industrial IoT with intelligent, connected machines, a conventional, static maintenance schedule seems to be old-fashioned. In this paper, we present a survey on the current state of the art in predictive maintenance for Industry 4.0. Based on a structured literate survey, we present a classification of predictive maintenance in the context of Industry 4.0 and discuss recent developments in this area.

Motivation & Objective

• Provide a structured overview of predictive maintenance in Industry 4.0.
• Identify key themes, methods, and gaps across the literature.
• Construct a framework capturing essential attributes of predictive maintenance.
• Discuss how predictive maintenance integrates with e-maintenance and Industry 4.0 concepts.

Proposed method

• Conduct a Structured Literature Review (SLR) using 'Predictive Maintenance' as the primary keyword.
• Extract and tabulate attributes from 150 papers (published 1993–2018) to build a framework.
• Cluster attributes into categories to form a hierarchical framework (tree structure).
• Differentiate conventional maintenance, condition monitoring, prognostics, and e-maintenance within Industry 4.0 context.

Experimental results

Research questions

• RQ1 What are the core goals and motivations of predictive maintenance in Industry 4.0 across the literature?
• RQ2 What attributes and dimensions characterize predictive maintenance, and how can they be organized into a comprehensive framework?
• RQ3 How does predictive maintenance relate to e-maintenance and Industry 4.0 concepts?
• RQ4 How do factors like system size, scheduling, and degradation modeling influence predictive maintenance approaches?

Key findings

• The framework comprises 10 categories: Goals, Condition Monitoring, Maintenance Scope, Degradation Process, Fault Detection, System Size, Scheduling, Prognostic Techniques, Data Handling, and Evaluation.
• 139 of 150 papers mention at least one goal, with cost minimization and downtime reduction being among the most common goals.
• Condition monitoring is typically sensor-based and online/real-time, with continuous monitoring being a key emphasis.
• System size analysis distinguishes single-component versus multi-component systems and highlights dependencies (economic, stochastic, structural).
• Scheduling is addressed in about 62% of papers, emphasizing dynamic scheduling and spare-part availability.
• The survey notes the rise of e-maintenance as an integrated extension of predictive maintenance aligned with Industry 4.0.