Heartbeat sensor fault detection in IoT using machine learning

By allowing real-time monitoring of critical data, the Internet of Things (IoT) has revolutionized healthcare; nevertheless, its dependability depends on precise sensors. With an emphasis on the MAX-30100 sensor, a pulse oximeter and heart rate monitor, this work uses machine learning to detect faults in heartbeat sensors. K-Nearest Neighbours (KNN) and Recurrent Neural Networks (RNN) were combined to create a hybrid system that analyses sensor data and finds anomalies. Whereas the KNN model compares actual data to historical data, the RNN model finds patterns across time. High fault identification accuracy was shown in the experimental data, and the hybrid technique outperformed solo models. This method reduces false positives and ensures accurate monitoring, which improves IoT-based healthcare by reducing the risks of malfunctioning sensors and offering trustworthy real-time health monitoring.