Laboratory simulation to understand translational soil slides and establish movement criteria using wireless IMU sensors

Abstract

Landslide monitoring and warning using inertial measurement units (IMUs) has shown the potential for remote and real-time applications. However, the studies conducted using the IMU sensors are limited to rainfall-induced landslide detection using soil moisture sensors and accelerometers for predicting slide and measuring tilt, respectively. The tilting of the slope might not occur during a slow-moving translational slide, and it may not always be possible to accurately record the soil moisture condition. The use of raw acceleration data, which is the combination of linear and gravitational accelerations, for calculating tilt or motion is another drawback of the existing studies. Hence, there is a need for a better approach to monitor slides. This paper presents two methods to define movement thresholds and criteria to identify the translational soil slides based on our understanding of the sensor data recorded during the two laboratory experiments. BNO055 sensor devices (IMU sensors) with 3-axis accelerometers and 3-axis gyroscopes were selected for this study. The linear accelerations, gravitational accelerations, and angular velocities were utilized to understand the translational soil slides by correlating the sensor behavior to that of the slope. The interpretation of the movements during the failure at each sensor location was further verified by referring to the videos recorded by two pi-cameras. The outcomes of this study confirm the applicability of the proposed IMU sensor system and the movement thresholds for effective and reliable monitoring and warning of translational soil slides.