Thesis – Energy-Efficient on-sensor Signal Processing for Wireless Bearing Monitoring in Railway Ap…

About Televic Rail:

With over 30 years of experience in designing, manufacturing and maintaining on-board communication and control systems, Televic Rail is a leading, trusted partner for railway operators and train builders worldwide. Its Passenger Information Systems and Control Systems are high quality, tailor-made solutions that offer the flexibility, user-friendliness and stability that our clients ask for. Our various types of on board control systems such as our bogie monitoring systems are innovative, yet reliable products designed specifically for the railway business. Trains and trams all around the world are equipped with Televic Rail solutions, from New Zealand to Canada, from China to the United States, from India to Belgium, England and France.

Your project:

This thesis focuses on the development of an ultra-low-power signal processing strategy for a new wireless accelerometer sensor designed for bearing monitoring within the COSAMIRA Edge ecosystem. The sensor is based on a Cortex-M4 microcontroller and is battery-powered, requiring a very long operational lifetime (15+ years). In wireless systems, energy consumption is dominated not only by sensing and processing, but especially by data transmission. Continuously streaming raw vibration data significantly reduces battery life and increases communication cost. The application for this thesis is railway bearing condition monitoring, where characteristic defect frequencies must be detected reliably under variable speed and load conditions.

The objective is to design and implement an embedded signal processing pipeline that transforms raw acceleration data into compact, meaningful health indicators directly on the sensor. Candidate approaches include multiscale FFT around expected defect frequencies, FFT combined with peak detection, band-energy indicators, or other lightweight spectral features.

Main challenges include:

• Designing algorithms compatible with Cortex-M4 computational constraints

• Minimizing RAM and CPU usage

• Quantifying energy consumption of processing vs. wireless transmission

• Optimizing the trade-off between diagnostic performance and battery lifetime

• Providing a clear battery consumption model linked to algorithm complexity and data rate

The outcome will be a validated low-power embedded architecture that maximizes battery life while preserving reliable bearing fault detection performance.