Senslide

Abstract

We describe the design, implementation, and current status of Senslide, a distributed sensor system aimed at predicting landslides in the hilly regions of western India. Landslides in this region occur during the monsoon rains and cause significant damage to property and lives. Unlike existing solutions that detect landslides in this region, our goal is to predict them before they occur. Also, unlike previous efforts that use a few but expensive sensors to measure slope stability, our solution uses a large number of inexpensive sensor nodes inter-connected by a wireless network. Our system software is designed to tolerate the increased failures such inexpensive components may entail. We have implemented our design in the small on a laboratory testbed of 65 sensor nodes, and present results from that testbed as well as simulation results for larger systems up to 400 sensor nodes. Our results are sufficiently encouraging that we intend to do a field test of the system during the monsoon season in India.