Automated Hypoxia Prevention and Real Time Human Life Monitoring System In Car

This research advances car cabin environmental monitoring by integrating real time anomaly detection and predictive analytics through artificial intelligence (AI). Existing systems typically operate reactively, identifying threats such as toxic gas accumulation or oxygen depletion only after surpassing critical thresholds, posing significant risks to occupant safety. In contrast, the proposed system leverages a hybrid deep learning architecture combining Long Short Term Memory (LSTM) networks with Generative Adversarial Networks (GANs) to proactively forecast and detect hazardous conditions. Trained on simulated multivariate sensor data, the system continuously monitors key parameters including oxygen, CO2, CO, methane, and temperature. Comparative evaluation against established time series models, including ARIMA, GRU, and XGBoost, demonstrated that the LSTM GAN model achieved the highest performance with 90% accuracy and an 88% F1 score. A correlation heatmap analysis was also conducted to understand interdependencies among environmental variables. This system is especially valuable for safeguarding vulnerable populations, such as children, elderly individuals, and physically challenged passengers, by providing early warnings and enabling timely interventions. The approach holds further potential for deployment in other enclosed environments, including sewage cleaning operations, to prevent exposure to toxic gases.