Cobble motion characterisation with smart sensors through laboratory experiments for ground-based landslide monitoring

Landslides often contain boulders on their surface or within the landslide body. Embedding sensors inside boulders within a landslide may help monitor its movement and dynamics. In this study, smart sensors were tested for tracking movements of a cobble, estimating its magnitude and mode of movement in dedicated laboratory experiments. The cobble was embedded with a sensor equipped with accelerometers, gyroscopes, and magnetometers. The experiments consisted of letting the cobble travel down an inclined plane. By changing the angle of the inclined plane, the cobble showed different modes of movement such as rolling and, when embedded in a thin sand layer, sliding. While travelling down the slope, the cobble was tracked to infer its position from camera videos. Raw sensor data were used for motion detection and discerning the mode of movement. Sensor-based acceleration and camera-based position were fed to a Kalman filter to derive the cobble velocity and compute the total kinetic energy to characterise the motion magnitude. Furthermore, LoRaWAN wireless transmission was tested by burying the cobble in sand layers of different thickness. The experiments contributed to understanding how the sensor functions and may be applied in the field for landslide monitoring, modelling and early warning systems.