Roussos Dimitrakopoulos



Sustainable Mineral Resource Development and Optimization Under Uncertainty

Tier 1 - 2004-11-01
Renewed: 2012-04-01, 2019-12-01
McGill University
Natural Sciences and Engineering Research Council

514-398-4986
roussos.dimitrakopoulos@mcgill.ca

Coming to Canada From


The University of Queensland, Brisbane, Australia

Research involves


Developing new, risk-based modelling technologies for holistic mine planning, design, and production scheduling, founded on stochastic modelling, optimization and artificial intelligence.

Research relevance


The research is contributing to sustainable development of mineral resources, by developing, testing and practicing advanced and improved decision-making tools for planning industrial mining complexes and extracting resources.

Sustainable Mineral Resource Development Under Uncertain Conditions


Sustainable development of mineral resources is critical, as it ensures the supply of raw materials and metals. Society needs to meet its present needs for metal and minerals, while preserving future generations’ ability to meet theirs.

To that end, Dr. Roussos Dimitrakopoulos, Canada Research Chair in Sustainable Mineral Resource Development and Optimization Under Uncertainty, considers risk quantification and management in mine planning, design, and production scheduling, based on his framework of stochastic (uncertainty-based) mine planning optimization, from the strategic level to the operational. His new framework considers the simultaneous stochastic optimization of mining complexes or “mineral value chains.” The framework looks at these complexes, or chains, as an integrated business, from extracting materials to delivering sellable products to customers / the “spot” market.

New, associated models and methods further build on the approach, using artificial intelligence and stressing self-learning. The resulting new techniques let engineering production systems learn and respond to incoming production information collected by a wide range of online sensors already available in industrial mining complexes.

Dimitrakopoulos’ research overcomes the limits of conventional optimization techniques in the strategic planning of metal mines and related mining complexes, given supply (metal and mineral deposit; or mining, processing or environmental) uncertainty, and demand (market) uncertainty. His stochastic framework addresses uncertainty stemming from mineral deposits’ supply of raw materials, mining operations, and fluctuating market demands.

Dimitrakopoulos’ new technologies have been tested in several real-world case studies at different mining complexes, and on various commodities. Results show mining production forecasts can be improved from five to 25 per cent; reserves can be increased by five to 15 per cent, and net present value can grow from five to 30 per cent as a result of his methods. His work takes advantage of big data, machine learning, hyperheuristics, and integrated stochastic optimization of very large industrial systems.

Dimitrakopoulos’ chair was originally awarded in 2005. Since then, his research has led to establishing the COSMO-Stochastic Mine Planning Laboratory and COSMO Mining Industry Consortium, which brings together AngloGold Ashanti, Barrick Gold, BHP, De Beers, IAMGOLD, Kinross Gold, Newmont Mining, and Vale—companies representing about 75 per cent of mining activities on the globe. The companies have supported Dimitrakopoulos’ research and testing of methods at mines around the world.