An industrial mining complex is an integrated value chain includes the excavation, transportation, processing, and distribution of mineral products as well as the storage of waste products. Critical sources of uncertainty include the characterization of the mineral deposits, commodity prices, equipment performance, and processing facility performance. To maximize long-term value of the mining complex, an optimization model is formulated, accounting for decisions in block extraction, material destination policy, downstream material flow, capital investments, and major operating modes. The size of the model, number of binary and integer variables, and the presence of non-linear transformations in the formulation make optimization with exact methods highly impractical. Metaheuristics have been used to provide good solutions in reasonable execution times. Hyperheuristics have extended these capabilities, using online learning to accelerate optimization. Algorithms amenable for parallel computing present opportunities to scale hardware and further improve optimization results while reducing execution times. The present work reviews these developments and explores how they can be combined for the application of optimizing mining complexes.