💨Zephyr Sense
Air Quality Monitoring and Automatic Air Purifying in Houses Near Cement Industrial Areas
💡Project Description
This project develops a system for real-time Air Quality Monitoring and Automatic Air Purifying using the core concepts of Hardware Engineering and Artificial Intelligence. It detects the usual air quality parameters like humidity, temperature, and dust with additional parameters like carbon monoxide (CO) and nitrogen dioxide (NO2) which are known to be produced by cement factories.
The entire system is managed by an ESP32 microcontroller as its core. All the sensor data sent to the microcontroller are then processed on-device into certain classes using a pre-trained Support Vector Machine (SVM) model. The air qualities are divided into four classes: Good, Moderate, Poor, and Hazardous. All those data are then sent to an AWS Cloud server using HTTP for further monitoring on the web dashboard.
🤝Meet Our Team
🧱Problem Statement
- 🌫️ High air pollution in industrial factory areas especially cement factories
- 📉 Lack of indoor air quality monitoring
- ⚠️ No responsive automatic air purification
🎯Goals
- 🔍 To monitor indoor air quality in real-time
- 🤖 To automate the air purification process
- 🛡️ To improve health and comfort for residents
💡 Solution Statements
🌬️ Use air quality sensors (e.g., PM2.5, CO, NO₂) to monitor indoor pollutants.
📶 Use ESP32 with Wi-Fi to process and transmit sensor data.
🧠 Apply a pre-trained SVM model to classify air quality levels in real time.
⚙️ Activate the air purifier automatically when poor air quality is detected.
🌐 Display real-time data on a web dashboard hosted on AWS.
🔔 Notify users when air quality becomes poor or hazardous
🧰 Prerequisites – Component Preparation
🔩 Hardware
🔌 ESP32 Dev Board – for processing, control logic, and Wi-Fi communication.
🌫️ PM2.5 sensor – for particulate matter detection.
🧪 MQ-7 gas sensor – to detect CO
- 🧪 MiCS-2714 gas sensor – to detect NO2
🌡️ DHT22 – for temperature and humidity sensing.
- 🌀 DIY Air Purifier – for purifying surrounding air.
🔋 Power supply & wiring components – including breadboard, jumper wires, USB cable, etc.
💻 Software / Platform
🧠 Pre-trained SVM model – compiled or converted to run on ESP32 for local inference.
🌐 AWS cloud services – to host the web dashboard and receive/send device data.
🛠️ Arduino IDE – for coding and flashing the ESP32 firmware.
🧾 Web technologies – such as Laravel PHP and HTML/CSS/JS for dashboard front-end and backend integration, REST API for data transmission endpoints, and HTTP for the application communication protocol.
ESP32 pin datasheet:
🧩 Schematic
🔲Block Diagram
🎬Demo
Note: If you want to jump immediately to the demo, please proceed to minute 2:29
✅ Conclusion
This project demonstrates a smart and responsive indoor air quality monitoring system using embedded AI. By combining environmental sensors with an ESP32 and a pre-trained SVM model, the system can detect harmful air conditions and automatically activate a purifier. Real-time data is displayed through a web dashboard hosted on AWS, allowing users to stay informed and safe. The solution is practical, scalable, and well-suited for residential areas near industrial environments.
🗃️Github Repository
Web Dashboard: https://github.com/tangerineshirt/zephyr-sense
Arduino IDE Code: https://github.com/tangerineshirt/Pollution-Monitor
