# Geotechnical Sensor Analysis – Kriging or Gaussian Process Regression for Fracture Risk in Mining.

The ability to make informed and quick decisions is invaluable in any operation. Mining operations are particularly safety aware. For geotechnical experts to assess risks and decide on the appropriate procedures quickly, can save lives and save money.

In mining operations there are sensors that measure cave depth, ground failure and fracture risk just to mention a few. Analysing all these measurements independently complicates the interpretation and can delay reaching a conclusion and a resulting decision.

We would like to demonstrate how data from various sources can be combined into a single visualisation that conveys all the information that a technical expert needs to decide on an appropriate course of action. The images below illustrate how sensor data can be used to generate images that assist geotechnical experts visualise contour plots of the rock body to highlight fracture zones and closure areas.

## Visualise 3D Plot of Open Cast Mine

The 3D plot illustrates the geometry of an open cast mine. This data would typically come from sensors in a real operation. The circles represent fracture risk in the rock and their diameters are proportional to the magnitude of risk. The fracture risk data would typically be obtained from drill samples on a real mine. This a very useful visualisation of the raw data.

## Interpolate Fracture Risk

Only being able to visualise the fracture risk at the sample points is not very valuable. Typically, Kriging or Gaussian Process Regression is used to interpolate the fracture risk between the sample points and allows us to visualise the possible fracture risks at points other than at the sample points.

Here is a 3D volume plot of the open cast mine. A colour scale is included that represents the interpolated fracture risk at each point. Contour lines are added to aid visual inspection.

Here a single image is presented that captures the geometry of the mine, the interpolated fracture risk and contour lines. In the code used to generate the image (download here), you have the option to inspect the fracture risk at various slice thicknesses. This level of interactivity allows for geotechnical experts to inspect the dissection of the mine at multiple points.

This single image is an example of how visualisations can provide geotechnical experts with the information they need to make safety critical decisions in seconds. The automation of data processing and visualisation further contributes by reducing the time spent crunching data and enables technical experts to spend time applying their expertise and making key decisions instead.