A team from the University of Adelaide’s Australian Institute for Machine Learning (AIML) has come second in the OZ Minerals Explorer Challenge. The DeepSightX team’s analysis of mineral data from the Mount Woods project area predicts where new deposits of elements and minerals could lie.
“The DeepSightX team exploited multi-disciplinary skills at the intersection of artificial intelligence and geoscience to analyse public data sets from the Mount Woods project area near Prominent Hill,” says the University of Adelaide’s Associate Professor Javen Qinfeng Shi, Director, Probabilistic Graphical Model Group, Director in Advanced Reasoning and Learning, AIML and Associate Professor/Reader, School of Computer Science.
“The team won 2nd prize in the challenge and with it $200,000.
“Researchers from AIML and the Institute of Minerals and Energy Resources (IMER) - both hosted by the University of Adelaide – collaborated with industry experts in minerals exploration (Austrike Resources) and geoscientific modelling (Gondwana Geoscience).
“The team developed a drilling exploration plan that took advantage of the overwhelming data available, while being justifiable from a geoscientific perspective.
“We achieved this by integrating the latest concepts from mineral systems modelling, with recent breakthroughs in deep learning - artificial neural networks and algorithms inspired by the human brain that learn from large amounts of data - and computer vision.
“The team began with a deep dive into the available data. Our engineers developed a variety of interactive visualisation tools, in order to facilitate our domain experts’ ability to explore the data. In parallel, our geoscientific modelling experts produced state-of-the-art models of TMI, Gravity, and Magneto Telluric (MT) data for the Mount Woods region.”
The DeepSightX team includes: Javen Qinfeng Shi, Hao Zhang, John Anderson, Karl Hornlund, Dong Gong, Ehsan Abbasnejad, Zifeng Wu, Lingqiao Liu, Matthew Zengerer, Christopher Matthews and Anton van den Hengel. Graham Heinson provided valuable advice.
Within three months the DeepSightX team developed a world-class predictive modelling capability, which illustrates the disruptive potential of machine learning when paired with expert domain knowledge. DeepSightX’s model will be tested in real life, with the top targets scheduled to be drilled by the end of 2019 with the prospect of unearthing the next big Australian mineral deposit.
One thousand global participants from sixty-two countries dug through more than six terabytes of public and private exploration data from OZ Minerals’ Mount Woods tenement in northern South Australia in the data-driven Explorer Challenge.
“The prize recognises DeepSightX as a new and competitive minerals exploration AI team that applies a multi-disciplinary approach with the requisite domain expertise to tailor target modelling to each new geological and commodity setting as these arise. It also demonstrates that South Australia has exceptional local talent,” says Associate Professor Shi.
Associate Professor Javen Qinfeng Shi, Director, Probabilistic Graphical Model Group, Director in Advanced Reasoning and Learning, Australian Institute for Machine Learning, Associate Professor/Reader, School of Computer Science, The University of Adelaide.