Sensor-based slope stability prediction using a digital twin and AI-driven stability forecasting

Abstract. The need for better geological risk management techniques has increased due to the frequency and severity of natural disasters like floods and landslides, which are being caused by urbanization and climate change. Such management has always depended on limited simulations and static models derived from historical data. But more dynamic methods for modelling physical situations in real time and predicting future events are now available thanks to recent developments in digital technology, especially Digital Twins (DT). The use of DT in landslide prediction is examined in this work, with an emphasis on the use of inexpensive sensors in real-time monitoring of vital environmental factors such ground movement, pore water pressure, and volumetric water content. The research was conducted on a test site located on the Feo di Vito hill within the University of Reggio Calabria, a geologically vulnerable area. The proposed system integrates real-time environmental monitoring with advanced modeling and predictive techniques, ultimately supporting early risk detection and response. Results highlight the potential of this approach to enhance forecasting accuracy and responsiveness, offering an effective, scalable, and low-cost decision-support tool for mitigating landslide risk in vulnerable areas.