Details
- Supervisors
- Faculty
- Degree label
- Abstract
- Monitoring the heart rate of firefighters during operations is a critical safety challenge, as physical effort, environmental stress, and rapid body movements can severely degrade the quality of physiological measurements. This thesis addresses this problem by developing and validating a motion-artifact-robust heart-rate tracking method as a step towards an embedded, wearable system for continuous physiological monitoring in firefighting contexts. The proposed approach combines photoplethysmographic sensing with accelerometer-based motion information and relies on a frequency-domain estimation strategy to mitigate motion-induced disturbances while preserving physiological signal content. The work is structured in two complementary stages. First, the algorithm is implemented and evaluated in simulation using both synthetic data and the IEEEPPG benchmark dataset, which contains photoplethysmography and accelerometer recordings under dynamic motion conditions. Second, the method is ported to a real-time prototype built around an STM32f767ZI microcontroller and interfaced with a photoplethysmography sensor and an accelerometer. The prototype is then characterized in terms of estimation accuracy, task scheduling, and resource usage, enabling an assessment of its suitability for real-time operation under constrained embedded conditions. The results show that the proposed method can provide reliable heart rate estimates in the presence of significant motion artifacts, with a mean absolute error of 2.75 BPM in steady-state conditions and 3.92 BPM during controlled physical activity. Estimation cycles complete within the specified timing constraints, while ressource usage remains well below the available microcontroller capacity. Beyond its immediate technical contribution, this thesis demonstrates the feasibility of a real-time physiological monitoring approach that could form the basis of a future wearable system for firefighter safety.