Bid for transformational AI centre to support resource sustainability
A proposed Australian Research Council (ARC) Centre for AI in Resources will explore how Artificial Intelligence (AI) technologies can improve the profitability and sustainability of mining and mineral processing operations.
The University of Melbourne is developing the ARC Centre bid in conjunction with partners in the mining and mining equipment, technology and services (METS) sector to support the adoption of AI.
If successful, the centre will run for five years and will deliver outcomes that directly benefit its industry partners. Commercial outcomes are an essential focus of the centre, which would be part of the ARC’s Industrial Transformation Research Program, with contributions from the Australian Government, the University of Melbourne and industry partners.
Enterprise Professor for the University’s Sustainable Resources research program, Dr Chris Goodes, says the centre will build on world-leading research in this field already underway at the University of Melbourne.
It will allow us to tailor AI technologies to address issues specific to the mining industry,
he says.
This includes addressing industry challenges such as the increasing scarcity and complexity of mineral deposits.
Community expectations underpinning the industry’s social license to operate are also increasing,
Dr Goodes says. While the resources sector supplies many crucial minerals and metals, which provide multiple benefits for society, it is also expected to reduce its energy consumption, reduce greenhouse gas emissions and minimise waste in its operations.
Dr Goodes says AI has the potential to help do this. It can consider the combined performance of many individual operating units and optimise outcomes for multiple objectives such as safety, output or quality. It can do this across multiple time frames, from a single shift to a decade.
However, to develop and implement the technologies needed to do this, we need strong partnerships that bring together miners, researchers and the METS sector. New technologies traditionally come from the METS sectors, while the miners themselves are the users of these technologies, rather than the developers.
The centre’s proposed research programs will be finalised in conjunction with the bid participants, with the three broad categories including Humans and AI, AI assurance and AI engineering lifecycle.
Professor Adrian Pearce, who leads the Sustainable Resources research program, says the ARC centre will provide practical solutions to the challenges of implementing an AI system across a mining operation.
What are the questions a miner needs to consider in the design and development of such a system? And how do you go about integrating AI with existing networks of autonomous equipment, sensors, data and other communication systems?
To address these issues, the centre must consider accountability and ethics in the deployment of AI systems, as well as the training and skills staff require to optimise how they use and interact with an AI system.
Other issues include technical testing protocols and AI assurance and validation procedures, standardised data protocols across diverse sensor and communication networks, and AI system security.
Potential partners in the proposed ARC Centre for AI in Resources are urged to contact Professor Adrian Pearce, adrianrp@unimelb.edu.au, for more information, with participation expected to be finalised by October 2020. The ARC submission closes in November. If the bid is successful, the centre will begin operating in late 2021.