Abstract
Landslides threaten the infrastructure and safety of communities. Soil conditions can predict landslide threat, but the cost and complexity of sensing systems for documenting hazardous conditions across a heterogeneous spatial area prevent widespread utilization. The SitkaNet system is a low-cost, easier to install alternative that allows for numerous sites to be monitored with real-time reporting and expands the accessibility of data-driven landslide forecasting. Using a combination of industry-proven sensors and cheaper alternatives, each SitkaNet node can measure the rainfall, six soil moisture sensors at varying depths, water table, atmospheric pressure, humidity, and temperature at each site for less than one-fifth the cost of existing solutions (<$1000/node). The SitkaNet nodes transmit data wirelessly at five-minute intervals over LoRa network to an Ethernet connected hub instead of more traditional on-site cellular or satellite methods. The node electronics are packaged with 3D printed components in a small waterproof case mounted on a hand-driven well-point utilized for the water level measurement. Each node is intended for operation for more than six months on a lithium-ion battery pack: no solar panel is needed, so amenable to low-light sites. The installation process is streamlined which allows for a node to be installed in less than a day compared to multi-day procedures required by other systems.