LANDSLIDE PREDICT: MACHINE LEARNING AND SENSOR-BASED EARLY WARNING

Abstract

Landslides are natural disasters that pose significant threats to human life and infrastructure. This project presents a machine learning (ML) and sensor-based early warning system to predict landslides effectively. The system integrates real-time sensor data with ML algorithms to enhance prediction accuracy and provide timely alerts. Key hardware components include soil moisture sensors, rain gauges, vibration sensors, and tilt sensors, all interfaced with an ESP32 micro- controller for data acquisition. The collected data is analyzed and processed using a trained machine learning model, which predicts landslide-prone areas based on historical patterns and real-time conditions. The trained model has achieved 98By combining sensorbased monitoring and machine learning-based predictions, this system enhances early warning capabilities, thereby reducing the risk of landslide- related damages.