Design and Deployment of an IoT-based Landslide Early Warning System

Abstract

The occurrence of rainfall-induced landslides is observed worldwide. National and local government authorities have implemented monitoring systems to mitigate landslide disasters. However, most of these systems use complex devices that are high-cost and require a constant power source to transmit data to the servers. This research designed and deployed an IoT-based landslide early warning system (EWS). The architecture is divided into three layers― Data collection, Data transmission and Data display and analysis. For the tier of data collection, an off-the-grid solar energy-powered integrated platform with various portable and low-cost sensors was built; For the tier of data transmission, mobile routers were used to support the sensor platform; In terms of data display and analysis, research provided an open architecture to analysis data and proposed a method for predicting landslides with multiple indicators. The framework is tested, extended and tuned by the model test. Results presented accurate predictions, all warning times are before the occurrence of the landslide. Then the proposed EWS was put into an embankment slope in Fukuoka, Japan by considering geological conditions and sensor suitability. This research qualitatively guarantees the accuracy of landslide predictions while operating in a low-cost, sustainable manner. It presents how IoT-based EWS can generate a positive impact on disaster mitigation work.