DEVELOPMENT OF WIRELESS SENSOR NETWORK SYSTEM FOR LANDSLIDE MONITORING & PREDICT EARLY WARNING

Abstract

Landslides are regarded as hazardous natural calamities, which can cause loss of lives, properties, and immense impact on the environment, therefore, it is crucial to monitor and issue early warnings. This thesis provides the design and deployment of a Wireless Sensor Network (WSN), which is an ad hoc system designed specifically for use in monitoring and predicting landslides. The architecture of WSN has been proposed with the implementation of multiple sensor nodes spread in the landslide susceptible regions. The Sensor is 1-meter-long with a 0.1-meter diameter with its end shaped like a nail. Every sensor node contains the necessary measuring instruments for the selected parameters like moisture, vibrations, and atmosphere in the ground. The components are Arduino, Gyroscope, accelerometer, moisture sensor and battery. They integrate the sensors in a way that gives adequate coverage with the possibility of differences in the data depending on the region. The system's design emphasizes energy efficiency and reliability. Sensor nodes are optimized for low power consumption, extending their operational lifespan and ensuring continuous monitoring. Redundant communication paths and fault-tolerant mechanisms are incorporated to enhance the network's robustness against node failures and external interferences. Field trials were conducted in regions with known landslide activity to validate the system's performance. The tilting of accuracy 0.001 degrees was recorded and is shown by the sensor. The results demonstrate the Wireless Sensor Network’s capability to provide accurate and timely data, enabling early warning and rapid response. The data collected during these trials also contribute to a deeper understanding of landslide dynamics and can inform future improvements to prediction models. To see how the routing algorithm works, consider two Boss Nodes that are not yet communicating. The core of the step is as follows: they both broadcast to the neighbors that their Level is 0. The neighbors process this information and know that both of their Bosses are on Level 1. Already nodes 18, 19, 22 and 23 now must choose the correct Boss node, and the correct node is the one with the lowest Level. This thesis underscores the potential of WSN in natural disaster monitoring and highlights the benefits of using advanced sensor technology and data analytics for early warning systems. The developed system represents a significant step forward in landslide risk management, offering a scalable and efficient solution for communities in vulnerable areas.