FOREST FIRE PREDICTION SYSTEM USING FUZZY LOGIC METHOD

Abstract

Forest and land fires are a serious problem that often occurs in Indonesia, especially during
the dry season. The impact is very wide, ranging from ecosystem damage, health problems, to economic
losses. This research aims to design and build  an Internet of Things (IoT)-based forest fire prediction
system using the Fuzzy Logic method. The system is equipped with a temperature sensor (DHT22),
smoke sensor (MQ-4), and flame sensor (Flame Sensor) connected to a ESP8266 NodeMCU
microcontroller. The data collected by the sensors will be processed and analyzed using the Fuzzy Logic
method  to determine the level of fire risk with the categories of "Safe", "Alert", and "Dangerous". The
test results showed that the system was able to detect changes in temperature, humidity, smoke, and the
presence of fire in real-time and provide alerts through the website. The temperature sensor has an
average error of 4.8%–5%, humidity of 4.1%–4.5%, and the fire sensor is able to detect fire up to a
distance of 300 cm. With a good level of accuracy, the system can improve preparedness in dealing with
potential wildfires and help authorities take preventive actions quickly and effectively.