Date(s) - Tuesday, 24/08/2021 - Wednesday, 25/08/2021
11:00 am - 2:30 pm AEST
Australia (Perth/AWST): 9.00 am to 12.30 am
Australia (east coast/AEST): 11.00 am to 2.30 pm
Manual interpretation of numerical drill hole data (such as multi-element geochemical analyses, petrophysical data, or wireline logging data) is time-consuming and subjective. Information provided by a geologist’s visual log aids interpretation, although this is also subjective and dependant on a geologist’s experience. In addition, not all geological boundaries are readily detectable visually.
We combine the power of machine learning with multiscale spatial information (using wavelet tessellation method) to provide the geologist with a fast, flexible, and intuitive system for automating drill hole logging. Our research team have recently made significant improvements and added new capabilities to Data Mosaic and a new version will soon be released.
Workflow and solutions are delivered via our interactive web app, Data Mosaic, which allows the geologist to upload their own data, control the workflow, and perform analysis on multiple variables with interactive plots. The process allows the user to input data from a variety of sources allowing the Data Mosaic web app to locate changes in downhole geology, based on those data. The user can then classify the numerical data into specific rock packages, with the aid of scatterplots, and compare the results to the geologist’s visual log, which can also be uploaded to the web app.
This provides insight into the spatial arrangement of the geochemical and mineralogical components of multiple drill holes, which can be incorporated into 3D models, at a scale chosen by the user.
We will provide training in:
For more information on these methods, please consult our research site: