An IoT-based forest fire detection system

Abstract

The objective of this project was to design an IoT based system that detects fire as quickly as possible and alert authorities before it spreads over an extensive area. The proposed system consists of an Arduino board integrated with a temperature, humidity and pressure sensor which will communicate via long-range loT protocols such as LoRaWAN to relay information about their environment to a central server. Readings are measured and sent wirelessly every couple of seconds, saved in a specified text file from which graphs for the three respective weather variables (temperature, pressure and humidity) against time can be generated to detect the change of rate of that particular variable being monitored. This is done since setting a threshold would not be enough to detect whether there is fire or simply a change in climate. The intention of using a multisensorial approach is to reduce false alarms and provide more accurate monitoring of the area under surveillance. Such a system would reduce the time needed to detect outdoor fire, resulting in less damage to the area. A key outcome observed from the results obtained was that the sensor module could detect changes in all three weather variables even up to 7.5 meters away from the heat source, potentially farther away, obeying the laws of Physics and the ‘Inverse Square Law’; temperature as well as pressure increasing whilst humidity decreasing. The simulations were performed with a successful wireless transmission from one microcontroller to the other, both connected to different power supplies and located in different areas from each other to test the LoRaWAN protocol connectivity distance. In conclusion, it can be said that from all the simulations carried out, the system performed and produced positive outcomes as intended, that is; Measurements from the sensor module were gathered and saved into a text file for record purposes and the generation of graphs, and sent via the Internet of Things to a microcontroller acting as a base station situated farther away from the area under observation.