IoT-Based Indoor Air Quality Monitoring and Alert System

Air-Conditioned Room Dataset (Aircond_Room_dataset.csv) This dataset contains ambient temperature, relative humidity, MQ-9 combustible gas readings, and PIR spatial motion data collected within a controlled air-conditioned room. The telemetry was continuously harvested using an ESP32-S3 microcontroller perception node to capture steady-state indoor environmental baselines. It is specifically structured for training and evaluating the localized Hybrid TCN-LSTM edge-AI system's predictive accuracy. 2. Balcony Dataset (Balcony_dataset.csv) This dataset records multivariate atmospheric telemetry, including temperature, humidity, MQ-9 gas concentrations, and PIR motion activity, captured in an open balcony environment. Collected via an ESP32-S3 microcontroller array, it provides a semi-outdoor environmental profile with higher natural variance. This data serves as a validation set for evaluating the domain generalization and robustness of the deployed edge-AI hazard prediction system. 3. Bedroom Stress-Test Dataset (Bedroom_dataset.csv) This dataset captures highly volatile environmental stress-test conditions, logging temperature, humidity, MQ-9 gas readings, and PIR occupancy data during controlled hazard events in a bedroom setting. The raw telemetry was continuously sampled via an ESP32-S3 hardware node to record sudden pollutant spikes. This dataset is critically utilized to evaluate the edge-AI system's proactive response latency and its ability to mitigate alert fatigue using spatial context.