Laboratory and Field Performance Assessment of LPWA Tilt Sensors for Slope Monitoring and Disaster Early Warning
Discuss this preprint
Start a discussion What are Sciety discussions?Listed in
This article is not in any list yet, why not save it to one of your lists.Abstract
This study evaluated the performance of tilt sensors equipped with Low-Power Wide-Area (LPWA) communication through controlled laboratory experiments and extensive field observations on natural slopes. Eight tilt sensors (A–H), varying in resolution, accuracy, installation type, and communication protocol, were examined. In the laboratory, a high-precision testing machine was used to apply precise tilt angles (0.0001º resolution, 0.0002º accuracy), enabling the quantitative assessment of each axis. Sensor accuracy was specifically evaluated using the 3σ (three-sigma) criterion as an index of noise, and Tilt Sensor A was found to exhibit the highest precision. Field observations at three sites in Japan assessed the influence of environmental factors, such as rainfall and temperature, on the tilt sensors' long-term stability. Composite tilt angles and inclination rates were derived from the data to quantify slope deformation. Buried Tilt sensors demonstrated high stability with minimal susceptibility to temperature effects, while above-ground tilt sensors with coarser resolution were more significantly affected by external conditions. At Site A, the tilt sensors successfully captured slope failure—from initial creep to accelerated deformation—demonstrating their effectiveness as early warning indicators. The observed relationship between the inclination rate and the remaining time to failure was consistent with previous findings. Crucially, a strong correlation was found between the catalog-specified accuracy and the field-observed noise level, indicating that tilt sensor accuracy reliably predicts field stability. These results provide essential, practical insights for the selection and operation of tilt sensors in slope monitoring and early warning systems.