Proper toothbrushing is essential for preventing oral diseases such as periodontitis, yet many individuals struggle to maintain effective brushing habits due to the lack of visual feedback and awareness of brushing coverage. While high-end electric toothbrushes address this issue by tracking brushing regions and duration, these solutions are often costly and rely on proprietary hardware.

Earables—ear-worn devices equipped with multiple sensors—offer a promising alternative for unobtrusive and low-cost activity monitoring. Positioned close to the oral cavity and facial bones, earables are well suited for sensing brushing-related signals without requiring additional instrumentation. OpenEarable 2.0, an open-source earable platform, is equipped with a rich sensor suite including inertial measurement units (IMUs) and magnetometers, enabling a wide range of context-aware and health-related sensing applications.

Electric toothbrushes emit characteristic electromagnetic fields during operation, which can be captured by the magnetometers embedded in OpenEarable 2.0. By wearing an earable in each ear, it becomes possible to infer the relative position and orientation of the toothbrush with respect to the mouth. Prior research has demonstrated the feasibility of detecting brushing activity using microphones, cameras, and various wearable sensors; however, many of these approaches suffer from privacy concerns, complex setups, or additional hardware requirements.

This thesis builds on existing work by investigating whether magnetometer data from OpenEarable 2.0 can be used to reliably detect and classify toothbrushing regions. By leveraging an open, extensible, and ear-worn sensing platform, the proposed approach aims to provide a practical, privacy-preserving, and accessible solution for real-time brushing feedback without the need for specialized toothbrush hardware.

• Designing and conducting a controlled data collection study using OpenEarable 2.0 to capture magnetometer signals during electric toothbrushing with multiple participants

• Developing and training a machine learning model to classify toothbrushing regions based on magnetometer data, accounting for both static and dynamic brushing movements

• Evaluating the accuracy, robustness, and cross-user generalization of the proposed detection approach under realistic usage conditions

• A paragraph explaining your motivation.
• Your study program (Bachelor/Master), current semester, and field of study.
• A transcript of records (courses and grades).
• Your programming experience.
• Any areas of interest relevant to the topic.
• Your CV (if available)