The demand for mineral resources is increasing, and companies are forced to mining lower grade and more heterogeneous orebodies to maintain supply. Such high variability leads to higher costs, more complex processing plants, and increased environmental impacts. Mining companies therefore need models that can accurately predict processing costs from exploratory drill core data. As ore grade decreases in the mine, the total energy consumption per tonne of ore will increase, as well as the volume of water and diesel/electricity. Enhanced knowledge of orebody characteristics is thus vital information to optimize profitability, ideally involving a geometallurgical program at the early stage of mine development that can help identify potential processing issues prior to major capital investment. Due to steady advances in technology, machine learning and the use of advanced statistics, much data is collected during drilling yet is not used to its maximum to build reliable geometallurgical models. Machine learning and deep learning are becoming an integral part of industry's drive to greater optimization and efficiency and have the capacity to deal with large, multi-dimensional datasets. They are attracting research attention and implementation across the mining sector. This session invites research contributions that seek to apply machine learning, image analysis and advanced modelling to solving problems in mining and related industries.
Key topics: Machine learning, Image analysis, Geometallurgy/mineral processing, Statistical process control