Oral Presentation Sub22 Conference

Including a robot at the table in identifying exploration targets (17283)

Suzanne Hunt 1
  1. OZ Minerals, Adelaide, SA, Australia

There is much still to learn about our earth and geoscience, it is quite breath-taking how much we still don’t know and are learning. An example will be provided of recent scientific news on the origin of heavy elements eg copper and its interstellar creation. Proving even the basics on the origin of metals still require some unravelling…

With computing power continuing to increase at a rate exceeding most conservative estimates. The future of geoscience is integrating rapidly into the world of datascience. Key components of this change are sensors and advanced instrumentation, big data and artificial intelligence, spatial data characterisation, decision making algorithms and computational imaging with visualisation systems.  These systems and process flows will themselves communicate, learn, and transfer information to management.

Four broad categories can be defined in terms of tools being used for exploration with some examples provided from recent work at OZ Minerals which demonstrate these capabilities and examples across many technology partnerships will be provided. To apply the knowledge we gain to execute our programs of work, we need to consider the best approach to do this.

At OZ Minerals our framework of systems and behaviours we call The OZWay, guides us while giving us the freedom and pathways to achieve our aspirations and purpose. The OZWay shows our ecosystem; the context, choices, enablers, work, and performance that live within the world of OZ Minerals. It is our way of working and how we work together.

OZ Minerals looks at the world of value creation through the lens of our stakeholders – our workforce, communities, shareholders, governments, customers, and suppliers. Every group is just as important as the other. When we make decisions, it is based on what value we can create for all six groups. We also assess our risks (which for us means threats and opportunities) from the perspective of value creation for these six stakeholder groups. This is an approach we take to all our work, including mineral exploration. We seek to explore ethically and responsibly, mine and sell modern minerals. In doing so, we are contributing to a low carbon future and economic wellbeing which, in turn, helps us achieve our purpose of making lives better.  

Our industry has a poor record of discovery globally. We are at a unique moment in the energy transition. We are differentiating ourselves by thinking about things differently.

In terms of machine learning/ artificial intelligence (ML/AI) work, we consider this conceptually by thinking about our Exploration Product and our Pipeline. In advancing and selecting new targets we use AI/ML to add another layer of information into the ranking mix by adding another facet to the discussion. You can think of this as having our traditional geological team members together with a robot around the table. The robot can challenge traditional thinking and the geologists can challenge the robot, asking for Explainability and why weighting of certain layers is occurring. Full collaboration with the Machine derived information. 

The four tools we use at OZ Minerals include:

  • AI discovery tools, potential maps and big data targets at the country and commodity level
  • Unconventional geological characterisation to support rapid greenfield target testing
  • ML trained near miss fingerprinting of world class deposits to fully evaluate each target we drill, plus scanning global drilling results
  • Groundwater research studying chemical isotopes as vectors for hidden mineralisation – underway in some of our key terranes with encouraging results.

To have a robot at the table, we need the robot to provide explainability and be transparent about why they have selected a particular location, clearly expressing why datasets have been given certain weightings, so the overlap between traditional and data driven targets is clear. This creates buy-in with the traditional geological methods.

  • Caption::
  • Acknowledgements: