Research Projects

(Rationale) Land subsidence, a gradual sinking of the Earth's surface caused by factors such as excessive groundwater extraction, mining activities, and natural soil compaction, poses significant threats to infrastructure, ecosystems, and human life. Conventional land subsidence monitoring techniques frequently have gaps in their temporal and spatial coverage, which renders them inadequate for real-time or large-scale applications. With the growing availability of satellite imagery and advancements in data analytics, an opportunity exists to develop a more effective solution for subsidence monitoring. This project aims to develop an advanced framework for Interferometric Synthetic Aperture Radar (InSAR) based deformation detection by utilizing state-of-the-art machine learning, deep learning techniques, and Sentinel-1 and Sentinel-2 satellite imagery to monitor, analyze, and predict land subsidence in near real-time.This system will overcome the shortcomings of existing techniques by utilizing AI-based predictive modeling, effective preprocessing, and automated data collecting. The framework will also feature an alert mechanism to educate stakeholders about major subsidence events, enabling prompt response to limit hazards. (Scientific Objectives) 1. Design an automated system for retrieving Sentinel satellite imagery and associated geographic data from internet archives and near real time data, ensuring continuous data availability. 3. Develop and implement machine learning and deep learning algorithms to predict land subsidence trends using historical and near real-time satellite imagery. 4. Create an automated alert mechanism to notify stakeholders via SMS and email when significant subsidence patterns are detected, with varying levels of urgency (e.g., Alarm, Alert, and Evacuate). 5. Validate the framework on subsidence-prone regions like the NF Railway Construction Project (Bairabi to Sairang), which collapsed in June 2023, using historical data to assess its accuracy and applicability. (Hypothesis) 1. Land subsidence trends can be precisely identified and quantified by using high-resolution satellite data. 2. Future subsidence trends can be accurately predicted by using both recent and previous satellite data. 3. A streamlined workflow integrating automated data acquisition, preprocessing, and analysis ensures feasibility for real-time applications. 4. Early alerts through an automated system can help stakeholders take proactive measures to prevent infrastructure damage and ensure public safety. 5. Test the framework on real-world subsidence-prone regions will validate its reliability and scalability across different contexts. (Field Research) The NF Railway Construction of B.G. Rail Line Project (Bairabi to Sairang), which collapsed in June 2023, serves as a critical test case for validating the proposed framework.