The iron ore industry in Australia has, in recent years, experienced an unprecedented growth phase. But now that major expansions are moving into the operation phase, new challenges have emerged. Whereas initial planning may involve a few modelling studies, ongoing performance monitoring and improvement requires a different, more intense approach. This is especially important in the face of changing factors such as commodity prices, exchange rates and labour markets. Response time is key. As a mining operation proceeds from initial planning to daily operations, the time window for analysis shortens, going from months to weeks to days as the business matures.Simulation models and analysis tools are typically built and run by specialists in the planning phase, but then what? The investment has been made, but realised only once. What if further analysis could be carried out in-house, re-using existing models using up-to-the-minute data? Can we create virtual mines that run in parallel to live operations that allow computers to identify areas of a value creation or operational improvement? These questions can be answered by the innovative application of big data, simulation and predictive analytics. The future will be having this all integrated into one platform that offers, for the first time, an opportunity for non-experts to apply existing modelling and analysis assets - combined with live data - to deliver support for continuous improvement, and the ability to respond quickly to external changes in the best possible way, before they become big problems. These techniques will represent a step change in the ability for operations to maximise efficiency and will underpin the intelligent operations of the future. These issues are addressed in this paper.CITATION:Schneider, M S and Grigoleit, M T, 2015. Maximising productivity through big data, simulation and predictive analytics, in Proceedings Iron Ore 2015, pp 475-480 (The Australasian Institute of Mining and Metallurgy: Melbourne).