Landslide Risk Management Using Sensors,
Networks, and Computing
Approximately one third of the land area of the Philippines consists of steep mountains and slopes making it prone to landslides that result in significant loss of life and property. The practice of monitoring landslides in real time using field instrumentation is well established in developed countries. However, cost and logistical constraints make it impractical to use commercially available field instrumentation for monitoring landslide prone areas in the Philippines. The cost of importing and installing such devices is generally far beyond what most communities can afford.
This on-going research aims to develop a cost-effective and accurate method for forecasting landslides in soil slopes by extending the observational method widely used in geotechnical engineering. Two key components of this approach are a predictive model for calculating observational parameters in the field, and a monitoring system for measuring observational parameters, and iteratively refining predictions made by the model. Wireless sensor networks (WSN), deployed at strategic points on a slope, periodically measure parameters such as ground movement, rainfall intensity, moisture content, and pore water pressures, and transmit collected data to a central repository for processing and analysis from which landslide forecasts and/or predictions can be made.
At this stage, initial designs and prototypes of instrumentation have been made. A system consisting of alternative instrumentation for slope movement and soil-water status is currently under development and shall be tested on a miniature slope model (Fig. 1). The landslide monitoring system will consist of three parts: an array of sensor columns, motes (integrated microcontrollers and radio platforms) and a Graphical User Interface (GUI). The sensor column arrays, buried vertically underground, are composed of pipe segments each containing a tri-axial accelerometer for tilt measurements and capacitive type sensors for soil moisture measurement. A separate column devoted to pore water pressure (PWP) measurements will be developed. Measurements taken in each segment can be accessed via Controller Area Network (CAN) communications protocol. The sensor columns can communicate wirelessly with a central base station through the radio motes placed above ground. The motes form a wireless sensor network that will collect, store and forward data for post-processing and display. A Python-based GUI will be developed for data interpretation. Other parallel efforts include development of cost-effective time domain reflectometry instruments for detecting underground slip surfaces and algorithms for relative position measurements using wireless motes. This work will give an update on the status of these efforts, including prototypes that have been built, and the experimental results arising from initial laboratory testing.