Fires are a major safety concern in homes and offices. Early detection of fires is critical for preventing property damage and saving lives. This project aims to build a fire detection system using gas and temperature sensors to predict when there is a fire or not.

The system will use a gas sensor to detect the presence of liquified propane gas, carbon monoxide and smoke that typically hint fire hazard. Additionally, a temperature sensor will be used to detect abnormal temperature increases and humidity decreases.

The readings from the sensors will be processed by an Arduino board that will be connected to a computer. A Python script will analyze the data, and machine learning algorithms will be used to develop a model that can predict the likelihood of a fire.

The system includes a dashboard-like interface where the end-user can get alerts for detected fires and monitor the physical and chemical quantities around the environment.

• Arduino
• MQ2 Gas Sensor
• DHT11 Temperature Sensor
• sklearn
• Pynecone

• Node.js 12.22.0+

• Setup your arduino, breadboard and sensors as shown here
• Upload this sketch to the microcontroller using Arduino IDE or VS Code
• pip install -r requirements.txt
• pc init
• pc run