Design and Development of a Wireless Sensor Network-Based Landslide Early Warning System (LEWS) Using ESP32 and LoRaWAN
Bayu Satrio (a), Agustina Rachmawardani (a), Agustya Adi Martha (b), Dwi Indra Prasetyo (a)

a) Program Studi Instrumentasi MKG Sekolah Tinggi Meteorologi Klimatologi dan Geofisika, Jalan Meteorologi Nomor 5, Tanah Tinggi, Kota Tangerang, Banten, 15119 Indonesia
b) Badan Riset dan Inovasi, Bogor 16911, Indonesia

Abstract

Landslides are geological disasters with significant impacts on human safety and infrastructure, requiring a reliable early warning system. This study develops a prototype of a Web-Based Landslide Early Warning System (LEWS) using a Wireless Sensor Network (WSN) to monitor soil parameters in real time. The system consists of three sensor nodes, each equipped with an MPU9250 accelerometer and a capacitive soil moisture sensor. Node-to-base station communication utilizes the LoRaWAN protocol, while data transmission to the server is performed via Wi-Fi using the MQTT protocol. Data are displayed through a Laravel-based web interface and sent as alerts via Telegram. The system was tested through two simulations. First, a LoRa communication test showed coverage up to 100 meters with RSSI values ranging from -46 dBm to -74 dBm. Second, a landslide simulation showed a simultaneous spike in Peak Ground Acceleration (PGA) reaching 0.4-0.5 g across all three nodes before returning to a normal level below 0.1 g. Soil moisture monitoring showed stable readings between 45%-55%. The results demonstrate that the WSN-LEWS is capable of reliable data monitoring and transmission, and can provide early detection of landslide indicators through acceleration and soil moisture changes.

Keywords: LEWS, WSN, LoRaWAN, MPU9250, Internet of Things (IoT)

Topic: Instrumentation and Computational Physics