Description

Problem

Rolling bearing failures are among the main causes of electric motor breakdowns in Electric Vehicles (EVs). Existing maintenance strategies are mostly reactive or preventive, leading to unplanned downtime or unnecessary part replacements. Current predictive models often lack adaptability and generalization across varying EV conditions.

Goal

To develop and evaluate a hybrid predictive maintenance approach for rolling bearings in EV drivetrains by combining classical reliability models with machine learning techniques. The goal is to enable early fault detection, accurate lifetime prediction, and improved reliability.

Required Skills and Knowledge

• Understanding of machine learning and signal processing
• Experience with Python and data analysis libraries (NumPy, Pandas, Scikit-learn, PyTorch/TensorFlow)
• Knowledge of vibration analysis and bearing fault diagnosis
• Basic understanding of electric vehicle drivetrain systems
• Familiarity with wavelet transforms and time-frequency feature extraction