IoT-Enabled Indoor Air Quality Monitoring Dataset

This dataset comprises 1,500 time-series observations collected from an IoT-enabled indoor air quality (IAQ) monitoring system built on a Raspberry Pi platform. The system integrates multiple sensors for environmental, gas, and particulate matter measurements, including the BME688, MQ135, PM2.5 sensors, and VOC sensors, with data collected at 5-second intervals through I2C and SPI protocols. Each observation includes temperature (°C), humidity (%), atmospheric pressure (hPa), gas resistance (Ω), PM2.5 (µg/m³), TVOC (ppb), eCO2 (ppm), VOC Index, MQ135 raw values (ADC), sensor voltage (V), estimated gas concentration (PPM), and a machine learning-based air quality classification (Good, Moderate, Unhealthy).

The dataset underwent extensive preprocessing to ensure quality and reliability. Noise reduction was applied using moving averages for the MQ135 sensor, missing values were removed, and outliers were identified and validated using interquartile range (IQR) and Z-score analysis. Continuous features were standardized to support regression and classification tasks. Regression models, including Linear Regression, Decision Tree, Random Forest, Gradient Boosting, XGBoost, and LightGBM, were evaluated to predict PM2.5, TVOC, and eCO2 concentrations, while classification labels were generated using rule-based thresholds and percentile-adaptive methods. Additional analyses include correlation assessment, feature importance ranking via Random Forest, Principal Component Analysis (PCA) for dimensionality reduction, K-Means clustering for pattern discovery, and multi-output regression modeling.

This dataset represents a real-time closed-loop IoT system in which sensor readings are continuously monitored, machine learning models predict air quality, and actuators such as fans, air purifiers, and HVAC systems can be triggered automatically to maintain optimal conditions. Potential applications include smart indoor environment monitoring, industrial air quality management, predictive maintenance of HVAC systems, health-aware building optimization, and edge AI research. The dataset is provided in Excel (.xlsx) format, structured as tabular time-series data, collected under controlled indoor conditions, and suitable for benchmarking machine learning algorithms for IoT-based environmental monitoring.