This chapter investigates the role of multi-sensor data fusion in enhancing the accuracy and robustness of heart rate estimation, particularly under conditions involving significant noise interference on the signal. By combining signals from multiple sensor channels, data fusion techniques can mitigate the impact of noise and improve the reliability of the estimated heart rate. The chapter introduces and compares a range of state-of-the-art and novel fusion strategies, with a focus on their performance using the MIT-BIH Arrhythmia Database supplemented by noise from the MIT Noise Stress Test Database. Special attention is given to scenarios with low signal-to-noise ratios (SNR), where two approaches—Kalman fusion and \(\alpha \) -trimmed mean filtering—demonstrate superior performance. Kalman fusion excels when both input channels are noisy, whereas \(\alpha \) -trimmed mean filtering is more effective when at least one clean channel is available. To leverage the strengths of both methods, a hybrid algorithm is presented that dynamically switches between them based on a signal quality indicator (SQI) serving as a proxy for SNR. This adaptive fusion strategy achieves improved heart rate estimation accuracy and outperforms several leading techniques, offering valuable insights for designing resilient, real-time health monitoring systems.