Fire Guard-AI: Intelligent IoT-Enabled Fire Prediction and Emergency Response System for Firecracker Industrial Safety

Firework manufacturing plants work under the condition where even a slight change in the parameters of temperature, humidity, or gas concentrations could lead to severe accidents. In such industrial areas like Sivakasi, which houses many firework manufacturing plants, continuous safety monitoring becomes crucial. Explosions can happen because of the leakages of gas or the building up of pressure. Such incidences can also be caused by increased temperatures or any accidental fires. Environmental sensing and predicting can be of much use in avoiding such situations. This research paper talks about an environment sensing and explosion prediction and alerting system based on IoT sensors along with Machine Learning model. MQ-2 gas sensor helps sense combustible gases, while DHT11 sensor measures humidity. BMP280 sensor monitors the temperature and the atmospheric pressure, whereas MLX90640 detects any abnormal heating conditions. All these sensors will transmit their data using ESP8266 board to a XGBoost Classification Machine Learning Model hosted in the cloud, and it would predict whether the current environmental conditions are safe or potentially hazardous. In case it detects explosion-prone conditions, then there would be an immediate generation of alarm via buzzer. The proposed system shows efficient predictive ability and consistent performance.