Application and extension of machine learning, statistics, process analytics, and process automation techniques to improve mining operations

DRAFT CONTENT taken from press release

The cost of mining for minerals could be about to plummet thanks to the Cooperative Research Centre known as CRC ORE and data science researchers from Queensland University of Technology (QUT). Researchers working with CRC ORE have been applying their collective expertise to develop a roadmap for the application of advanced data analytics to streamline operational processes in hard rock metal mining used to source minerals such as gold, silver, iron or zinc. As a result, the CRC ORE is investing $750,000 in QUT to continue working with them for another two years.

QUT academic co-leaders, Professor Kevin Burrage and Professor Arthur ter Hofstede, and chief investigator Dr Moe Wynn are supervising researchers from the Schools of Information Systems and Mathematical Sciences, Dr Suriadi Suriadi and Sander Leemans as they work alongside domain experts at CRC ORE. The team is working on the application and extension of machine learning, statistics, process analytics, and process automation techniques to improve mining operations. During meetings with CRC ORE, it was clear that data analytics expertise of the QUT team could deliver additional functionality to CRC ORE’s specialised software.

The QUT team is analysing data collected from ore processing plants, extending their research into this new sector.

“By working closely with domain experts from CRC ORE, we have been able to quickly make sense of the data and identify opportunities where our experience in the application of data analytics in other industries can be used to improve mining operations,” Dr Suriadi said.

The QUT researchers have conducted similar projects within industries such as healthcare, insurance, and climate modelling, and their experience is highly relevant to the project at CRC ORE. CRC ORE is now funding QUT to see the continued delivery of the plans that Dr Suriadi and Mr Leemans have put in place.

By working closely with Nick Beaton, CRC ORE’s General Manager of the Integrated Extraction Simulator (IES), researchers developed a proposal for the development of high fidelity simulation process models, a validated Integrated Analysis Method (IAM), and an initial assessment of a novel dynamic forecasting model to allow better process control strategies in the mining process.

Similar to QUT, CRC ORE fosters collaboration between industry and researchers working on real life issues for the mining industry. The QUT researchers needed a good understanding of the mining industry in order to identify opportunities for genuine, effective applications for their work in the context of a mining operation. Throughout 2016, Dr Suriadi and Mr Leemans based themselves at CRC ORE for two days a week. At the same time, ORE staff were learning about data science methodology from and interacting with Suriadi and Sander on a daily basis.