Advances, Trends and Challenges for Determining the Condition of Railway Rolling Stock Using Automatic Classifiers: A Systematic Review

The use of Machine Learning tools for studying, among others, vibrating signals that enable a comprehensive analysis of the state of the elements under study through Machine Learning techniques has become widespread. Considering the main traditional classification methods of these tools and their associated use of artificial intelligence, this paper thoroughly analyses both current approaches and trends in their use, as well as examining intelligent means for diagnosing faults and monitoring the condition of mechanical systems. These methodologies are becoming increasingly common in Industry 4.0. The objective of this paper is to systematically review the latest trends in research and development for the diagnosis of faults and monitoring the condition of rotating equipment using Artificial Intelligence tools. Therefore, this paper studies Machine Learning techniques applied to the analysis of signals from rotating mechanical elements, particularly bearings and shafts, with a special focus on the classification of the condition of railway rolling stock.