Rotating machines are essential in numerous sectors, including railways, energy, and robotics. These machines exhibit unique degradation patterns and critical components that require monitoring. Despite the existence of various fault detection and diagnostic methods in current literature, only a few of them effectively consider different data sources and variable operating conditions. Furthermore, a generalized approach for consistent monitoring across different systems remains challenging. Therefore, this special issue aims at enhancing the generalization and application of these methods to diverse systems, emphasizing their robustness.     

This special issue is dedicated to addressing the challenge of developing robust and generic Prognostics and Health Management (PHM) algorithms for rotating machines. It particularly focuses on condition monitoring, fault detection and diagnostics under various operating conditions and with heterogeneous data sources.

To promote innovative methods in the field of rotating machines fault detection and diagnostics, this special issue centers around several practical and scientific aspects, each reflecting a key challenge in this field. These challenges are into specific highlights, which are outlined below:

1. Advanced fault detection and diagnostic methods: emphasizing a range of applications for rotating machines.
2. Multi sensors and faults: exploring the integration of data from varied sensors for the monitoring of multiple faults.
3. Condition variability: addressing the complexities arising from different operational environments in systems.

To facilitate research and exploration, we provide three comprehensive data sets, each representing different case studies of rotating machines in various operating conditions, monitored using multiple sensors. These data sets include:

1. Detection and diagnostics of electrical and mechanical faults (AMPERE data sets).
2. Detection and diagnostics of gearbox faults (LASPI data sets).
3. Detection and diagnostics of multi-axis robot faults (METALLICADOUR data set).

A comprehensive description of these data sets, including experimental and acquisition details, is already available online in the published paper accessible via the DOI link here. Also, each data set is provided with a code for loading and structuring the data for training and testing models, and can be found in the following GitHub link here. Both documentation and codes enable researchers to easily integrate and analyze the data in their studies. They streamline the initial stages of research, ensuring that contributors can focus more on developing innovative PHM solutions.