First Experimental insights on a Novel Senso Network based Measurement Platform for Avalanche Dynamics

15

Artikel aus einem Buch

200205291

Snow avalanches are an ever-present reality in alpine regions; and they can cause heavy damage to settlements
and infrastructure. However, current modeling and simulation techniques of avalanche dynamics lack a precise description of the innerdynamics, e.g., how single particles, such as debris, or potential avalanche victims are transported. This means that the mechanics of important physical effects, such as particle-size segregation and phase separation inside the flowing avalanche, are currently invisible to us. Thus, understanding the behavior and the dynamics of avalanches is still an important challenge. In this work, we present new insights on AvaRange, a measurement based approach towards a better understanding of these internal particle dynamics. AvaRange adapts ranging techniques that we previously successfully applied to sensor networks to track particles that move with the avalanche. The idea is to place sensor nodes in rugged spheres of varying shape, size, and density and to deploy them in an avalanche release area. After (artificially) triggering the avalanche, the sensor nodes travel with the avalanche and collect ranging data to determine their relative position. The collected data allows to gain insights into particle trajectories and local properties. Here, we present the results of a first measurement campaign. We evaluate the most promising radio based localization techniques, namely Received Signal Strength (RSS) and Time of Flight (ToF) based approaches. The goal is to assess the applicability of these localization techniques in an extreme environment such as avalanches. Our experiments focus mainly on static signal transmission and localization accuracy in snow. The results show that these techniques are suitable for such a challenging environment. The RSS based approaches prove to be simpler to apply but less accurate than ToF based approaches. With ToF based approaches, we finally achieved accuracies of less than a meter. The findings suggest that the accuracy could be improved by: First, further adaption of algorithms to the special requirements of the application scenario in moving avalanches and second, customization of the radio hardware for a bigger transmission range in snow.