Masterthesis in the field Sound Based Predictive Maintenance (m/f/d)
About us
This master thesis is conducted within the Mechatronics, Sensors & VCA Power department and focuses on the development of a Sound-Based Predictive Maintenance (SBPM) approach for commercial vehicles. The objective is to investigate how vibration and structure-borne sound signals can be used to automatically detect mechanical defects and anomalies before they become critical failures. The work combines vehicle engineering, signal processing, machine learning, and practical validation on test vehicles.
You can look forward to these tasks
- Investigation of state-of-the-art methods in predictive maintenance, anomaly detection, and vibration analysis
- Analysis of vehicle vibration and structure-borne sound data
- Identification of mechanical failure modes and defect signatures
- Development of signal processing algorithms for feature extraction
- Evaluation of machine learning and deep learning approaches
- Creation of datasets and benchmark scenarios
- Validation of algorithms using real-world vehicle test data
- Assessment of detection performance under different operating conditions
We are looking forward to
- Master student in Computer Science, AI, Mechatronics, Electrical or Automotive Engineering
- Knowledge of Python, NumPy, Pandas and SciPy
- Experience with PyTorch, TensorFlow or Scikit-Learn
- Understanding of signal processing, vibration analysis or acoustics
- Strong analytical mindset and independent working style
- Good English communication skills
What we offer
Opportunity to work on an innovative AI-based diagnostic solution with direct industrial relevance, access to real vehicle measurement data and test environments, collaboration with experienced engineers and data scientists, and the possibility to contribute to future predictive maintenance concepts within TRATON.