Predictive Maintenance: A Bibliometric Analysis

Abstract

Predictive maintenance, or PdM for short, has emerged as a significant approach across sectors, employing data-driven methods like artificial intelligence (AI), data mining (ML), and the Web of Things (IoT ) to improve asset performance, decrease downtime, and save maintenance costs.

This review paper explores the significance of predictive maintenance through a bibliometric analysis using VOSviewer, identifying key research trends, influential publications, and thematic clusters in the field. The study synthesizes industry-specific applications, including manufacturing, aerospace, energy, healthcare, and transportation, highlighting successful implementations and their outcomes. Despite significant advancements, research gaps persist in terms of real-world adoption challenges, cost-benefit analyses, and cross-industry standardization. The findings indicate a strong research focus on predictive analytics and condition monitoring but reveal limited studies on implementation barriers, data security, and SME adoption. This review contributes to the existing literature by mapping the intellectual structure of PdM research, identifying emerging themes, and proposing future research directions to enhance scalability and effectiveness. By addressing the existing gaps, this study provides valuable insights for researchers, industry practitioners, and policymakers seeking to advance predictive maintenance strategies.