Two Decades of Progress in Deep Learning for Drive Failure Prediction (2004-2024): A Bibliometric Analysis

This paper systematically analyzes research trends, hotspots, and collaboration patterns in the application of Deep Learning (DL) for Drive Failure Prediction (DFP) through a scientometric approach. By examining relevant literature from major academic databases over the past two decades, it identifies leading authors, institutions, and countries/region contributing to the field. Furthermore, the paper explores the evolution of research themes and academic interests using co-occurrence network analysis. The findings reveal a significant increase in DLdriven predictive maintenance research, with growing interest in enhancing prediction accuracy and operational efficiency. This study provides valuable insights for researchers and practitioners aiming to advance the field of drive failure prediction through DL technologies.