Applications

Slope stability monitoring using networks – Currently mining companies use radar technologies to produce data, every few minutes, on the surface movement of slopes with sub-millimetre precision, resulting in very large data sets.

One approach was to take this information and convert the numbers into a network that allowed the research team to extract the concealed patterns on motion and how they were changing in space and time, despite the high-density and thus challenging nature of the information. The monitored locations on the slope represent the network nodes. Those locations with very similar surface movements were linked together. This process was then repeated until a network was established. The time evolution of the properties of this network were then analysed.

Stable locations that barely moved were identified, as too were unstable locations that had a lot of movement. But it isnt just about movement. Many parts of the slope can move without becoming detached from the rest of the slope. Slope failure – and more broadly failure of granular bodies – has a distinct signature where different pieces of the body begin to move coherently. This means member locations of each piece move “as one”. As failure draws near, this pattern of division in surface movement can be detected in the time evolution of certain properties of the network. When the body reaches this state of divided motion, the likely location and timing of failure may be predicted with reasonable accuracy.