An Australian university wants to fix mining's biggest data waste: the data nobody uses

The University of Queensland has installed a state-of-the-art mineral analyzer to research how to turn geological data generated across a mine's life into better decisions — not just archived reports.

The University of Queensland (UQ), in Australia, has installed a TESCAN Integrated Mineral Analyser (TIMA) at its Natural Resources Innovation and Characterisation Hub (NRICH). The instrument, which measures mineral composition, grain size and mineral liberation with high precision, will support the research of Pia Lois-Morales, a geologist specializing in geometallurgy and the inaugural recipient of the Jim Askew Evolution Mining Fellowship at UQ's Sustainable Minerals Institute.

The problem this research tackles isn't a lack of data — it's the opposite. According to Lois-Morales, a deposit's geological characteristics are collected and generated at practically every stage of the mining value chain, from exploration to processing, which makes geology one of the industry's richest sources of information. The bottleneck is structural: equipment like TIMA already exists in industry, but there it operates under pressure to deliver immediate operational decisions, with no time to explore the full potential of the data it generates. Having the same equipment inside an academic setting, without that shift-by-shift pressure, opens room to investigate how to collect better data, refine data-treatment workflows and, above all, test how to integrate datasets that today stay siloed between geology, mining and processing — work that day-to-day operations simply don't have time to do.

Every mine accumulates, over decades, an enormous volume of geological data that rarely crosses the boundary between the team that collected it and the team that decides, in processing or mine planning, what to do with that ore. Closing that gap without relying on more sensors or more drill holes — just using what already exists better — is the kind of gain any mature operation, with decades of data sitting in the archive, can replicate once the method is validated.