Scientists at the Indian Institute of Technology (IIT) Mandi have developed a new landslide early warning system. The system uses machine learning and advanced analytics. It can detect ground movements as small as less than one millimeter. The technology was tested in three landslide-prone areas in Mandi district. Researchers report over 90 percent accuracy in landslide prediction during trials.
Real-Time Monitoring System
The system collects real-time data using sensors. This data is transmitted to alert systems. It was developed with the National Mission on Himalayan Studies. Himachal Pradesh experienced about 140 major landslides during last year’s monsoon season. This new technology aims to reduce damage and casualties.
How the System Works
Installed on hills, the system triggers alerts for minimal ground shifts. These shifts are less than one millimeter. Alerts go to “warning poles” in vulnerable areas. This includes valleys and roads near hills. The poles activate hooters and blinkers. Alerts are also sent via SMS. A dedicated Android-based web application notifies the District Disaster Management Authority (DDMA) control room.
Professor Varun Dutt from IIT Mandi stated the system offers millimeter-level resolution. This is much more precise than satellite data, which typically provides information at a 20-meter resolution. Satellite data tracks ground movements over weeks or months. The new system detects finer, faster changes.
Integration with Satellite Data
Future enhancements will integrate data from the NASA-ISRO Synthetic Aperture Radar (NISAR) mission. This mission, launched by the Indian government, will improve warning accuracy and timeliness. Currently, satellite data is received every two weeks. Combining this with the new system’s real-time data provides more robust warnings.
The system’s deployment in Mandi district aims to safeguard communities. It provides critical lead time for evacuations and mitigation efforts. This is particularly important in the Himalayan region, which is prone to frequent landslides.