2020 Recipients

Mining operations are motivated to identify underground rock mass instabilities. However, geotechnical professionals often have insufficient time to investigate complex rock mass phenomena in detail, meanwhile more geomechanical instrumentation data is collected than ever before. The developing field of machine learning has been shown to ease the burden of data manipulation and reduce the bias introduced in a manual process. Josephine’s research lies at the intersection of rock mechanics and emerging machine learning algorithms. She is developing data-driven methods to predict rock mass behaviour in underground mining excavations. Machine learning algorithms have been found to be highly useful in ore prospecting, and for autonomous mine vehicles. The challenge in applying these algorithms to rock engineering is combining the various input data formats, developing algorithms to be interpretable, and validating the predictions using conventional methods. Josephine’s research objective is to produce a framework for applying machine learning to geomechanical datasets so that there is a blueprint for how these may be applied in mining practice. She is working with a collection of operating mines, instrumentation manufacturers, and mining consultants to improve the prediction of rock mass stability hazards, and to improve the efficiency of the rock support rehabilitation scheduling.

My M.Sc. research project focuses on gaining insights into ore-forming processes of seafloor massive sulphide (SMS) deposits on the modern seafloor, specifically at an actively-forming deposit (VOLPA) in the most tectonically- and volcanically-active place on Earth—the northern Tonga subduction zone with an industry partner SM2. This study involves an integrated program of geological and structural mapping at regional to local scales to determine the tectonic and volcanic controls on ore formation. This is integrated with petrological and geochemical studies of ore assemblages and fresh volcanic rocks, including major- and trace-element geochemistry, S-isotope systematics, and 226Ra/Ba dating of hydrothermal barite, in order to understand the geodynamic, volcanic and hydrothermal history of the site. The results of this study will add to a growing work on the SMS deposits on the seafloor and specifically along the Tonga-Kermadec arc, providing new insights into the relationship between geodynamics, volcanism, and hydrothermal venting. This work has direct implications for the emerging seafloor mining industry and may advance our understanding of VMS formation in ancient terranes where the geodynamic setting is difficult to constrain.