| dc.description.abstract | Forest fires are a serious threat to environmental sustainability and human health, especially in tropical regions like Indonesia. Early detection efforts are crucial to minimize the impacts. This research aims to design and create an Internet of Things (IoT)-based forest fire early detection device that utilizes a MEMS CO sensor to detect carbon monoxide (CO) and an MH-Z19 sensor to detect carbon dioxide (CO₂ ). These two gases are the main indicators of fire. This system is designed using a NodeMCU ESP32 microcontroller integrated with a GPS module for location tracking and sending data in real-time to the ThingSpeak platform. Fire notifications are also sent automatically via the Telegram
application if the gas threshold is exceeded, namely CO> 30 ppm and CO₂ > 600 ppm. Testing on this device was carried out in 3 different locations, namely, in the middle of a forest, a forest near a residential area, and a forest near a highway. Based on testing, the calibrated device showed an accuracy level of 1,41% for the MEMS CO sensor and 1,01% for the MH-Z19 sensor. This shows that the results of the tests carried out are within a technically acceptable tolerance range, so the tool is very suitable for use in collecting research data. Experimental data shows that the system is able to detect fires accurately up to a radius of 3 meters from the
fire point. From the results of the tool testing carried out, the average carbon
dioxide and carbon monoxide were obtained at 3 different locations, namely at the location in the middle of the forest, an average carbon dioxide of 540,2 ppm
and carbon monoxide of 1.6 ppm was produced, while at the forest location near the settlement, an average carbon dioxide of 512,6 ppm and carbon monoxide of 1.6 ppm was produced. At the last location, namely the forest location close to the highway, an average carbon dioxide of 512,9 ppm and carbon monoxide of 4 ppm was produced. So it can be concluded that the difference in CO and CO₂ levels at three different locations is influenced by human avtivities around the test location. | en_US |
