Using AI-Powered Machine Learning Model to Measure Air Quality Index (AirQ AI)

The research presented a new experiment using automated machine learning algorithms powered by artificial intelligence via the AirQAI application to measure the Air Quality Index (AQI) using images captured by mobile phone cameras or drones. The study was applied to a sample of Iraqi cities (Anah, Haditha, Fallujah, Baghdad, Amara, Samawah, Karbala, and Basra) to assess air quality with an accuracy of 98%, employing advanced artificial intelligence technology, specifically Efficient Net. The application (Android APK) was installed on a mobile device (Infinix XOS v.12.0) from the website (airqai.org). It utilizes image analysis to detect outdoor air pollution levels and categorizes air quality into six levels: Good, Moderate, Unhealthy for Sensitive Groups, Unhealthy, Very Unhealthy, and Hazardous, while monitoring five key gas indicators (PM10, PM2.5, NO2, O3, SO2) and storing the results in the device's memory for future comparison. The study's findings revealed that AI-powered applications provide accurate and user-friendly data, enabling effective measures to reduce pollution and raising public environmental awareness. They also automate air quality detection through image classification. The results indicated variations in air quality indicators ranging between 25-350. Cities like Anah and Haditha recorded an index value of 25, while Fallujah and Baghdad recorded 125, Amara and Samawah recorded 350, Karbala recorded 125, and Basra recorded 75.The variations are attributed to increased vehicle exhaust emissions, industrial pollutants, waste burning, global warming, climate change, traffic congestion, power plants, and dust storms, as well as the absence of green belts, green spaces, or tall trees.