Image default

O3 Mining Identifies Over 25 Targets for Exploration with AI Modeling

O3 Mining Inc. (TSX.V:OIII) (“O3 Mining” or the “Corporation”) is pleased to announce it has engaged Mira Geoscience Ltd. (“Mira”) to conduct modelling and exploration targeting over its Alpha property in Val D’Or, Québec, using artificial intelligence (“AI”). O3 Mining seeks to actively minimize exploration risks and mitigate costs by using proven cutting-edge technology.

Mira used drilling and mapping databases, geochemical samples, Induced Polarization (IP), Electromagnetic (EM), magnetic and gravity datasets and other public data, to provide a regional scale targeting model of the Alpha Property to assist with resource expansion and regional exploration. A total of 661,760 metres in 2,400 drill holes, over 2,000 surface assays, 160,000 assays for gold and base metals plus 25,000 multi-element assays, 400 line kilometres of IP, 5,500 line kilometres of Mag and over 700 line kilometres of EM were used.

“The Alpha property is an 80 square kilometre property located in the heart of the Val D’Or district hosting approximately 40 historical gold and copper-gold zones in numerous geological environments. Therefore, prioritization of drilling targets is key to maximize probabilities to rapidly discover significant mineral deposits on the property” says Jose Vizquerra, President and CEO of O3 Mining. “The application of the Mira innovative AI methodology is part of an integrated strategy to accomplish this task and come up with the best targets, which in turn will produce the best results and bring the most value to our shareholders.”

The deliverables produced for O3 Mining include the following:

  1. Regional-scale interpretation including surfaces from structural and stratigraphic modelling from the SKUA platform in the Gocad Mining Suite© software (“GMS”);
  2. Intrusion surfaces from unconstrained and geometry inversion of magnetic data;
  3. Leapfrog dynamic model; and
  4. Targeting voxet with constrained inversion results, geochemistry interpolation, feature engineered variables and prospectivity score results

Using the data populated into a 3D voxet comprising cells of 25m x 25m x 25m, Mira conducted different targeting exercises combining knowledge-driven and data-driven (supervised machine learning) methods. The knowledge-driven exercise combined the experience of O3 Mining’s geologists using an index overlay approach to generate mineralization probabilities throughout the model. The drilling information was used to construct predictive models using a random forest classification approach. Once the predictive model was deemed strong enough in making an accurate prediction on known mineralization, it was applied to the voxet to estimate the likelihood of each cells to be mineralized. By combining the different prospectivity scores produced by the targeting exercises, Mira produced a set of mineral prospectivity indices (MPI) for the different types of mineralization observed on the Alpha Property (Figure 1).

In parallel with the collaborative work with Mira, O3 Mining is currently executing a stripping campaign on selected mineralized areas to improve understanding of the geology and mineralization controls within the four sectors on the property.  The results are fed back into the Mira modelling to ensure that the AI process is supported with verified field observations.  This work started in June 2020 and will extend into September 2020. The results will be made public once the program is completed and full assay results are processed. O3 Mining believes that the combined Mira modelling, the in-depth knowledge of the property geology and the results of the stripping campaign will provide a strong base to the upcoming aggressive 150,000 metre drilling campaign slated to be executed between September 2020 and April 2021. 

Related posts

Anonos recognized in 2023 Gartner Emerging Tech Impact Radar: AI in Insurance

Business Wire

FaceMe® ranks First in NIST FATE Test for PAD

Business Wire

CodiumAI released an open-source pull-request AI agent

PR Newswire