Design of Landslide Early Warning System Using Fuzzy Method Based on Android

Abstract

In Indonesia, landslides are one of the many natural disasters that often occur during the rainy season. Especially in mountainous areas, cliffs, hills, which cause many losses. Therefore, it is necessary to create a landslide Early Warning System. Slope, vibration, and excessive water content in the soil are the leading causes of landslides. To measure these parameters, an Internet of Things (IoT) based system is used that is connected to various sensors. In this study, the fuzzy value obtained from the measurement of the MPU6050 Accelerometer and Gyroscope sensor, also Soil Moisture sensor sent to the Antares server using LoRa. In research, Fuzzy algorithm is used to analyze the sensor detection results in the form of three final decision rules based on the knowledge of a landslide expert, namely Safe, Alert, and Watch out, which can be seen on an android device with 90% accuracy value and 10% error.