Nouhan KOUROUMA

- Working with African and Moroccan companies on AI projects. - Assisting professors with lectures and practical sessions, particularly in electronics, robotics, and artificial intelligence. - Managing students in academic projects.

The aim of this project is to develop a diagnostic system for bearing faults in the High-Speed Trains (HST) in Morocco. Our approach is based on a research and development methodology aimed at conducting a comprehensive review of scientific work in this field, while also contributing through our own practical work. Initially, we undertake in-depth scientific research on bearing faults in HST, then proceed to implement our research on a dedicated test bench, where we collect acoustic and vibration data. These data are then processed using tools such as LabVIEW, MATLAB, and Python. Subsequently, we extract relevant features, which we use to train machine learning models to detect bearing faults in real-time. Finally, we implement these models for inference and create dashboards to track the evolution of bearing faults and generate alerts when issues are detected. To validate our theoretical and experimental study, we acquire real data from Moroccan HST, following the same procedure as in the theoretical and experimental study up to the final deployment of models for real-time diagnostics. Keywords: High-Speed Trains, Bearing Faults, Signal Processing, Machine Learning