Quantum Optimisation for Mine Planning and Extraction Scheduling

Workshop Description

For mine planners and operations research teams. Covers QUBO formulations for pit shell optimisation, block model extraction sequencing, truck-shovel dispatch, and tailings management. Includes D-Wave quantum annealing and Fujitsu Digital Annealer benchmarks against Whittle Four-X and Gurobi MIP on deposit models with 5,000-50,000 blocks.

Mine planning is one of the largest-scale combinatorial optimisation problems in any industry. A typical open-cut mine has a block model with 50,000 or more blocks, each with grade, tonnage, and geotechnical properties. Scheduling extraction across 20+ periods subject to precedence constraints (you cannot mine a block until overlying blocks are removed), blending targets (mill feed grade requirements), capacity limits (crusher throughput, truck fleet size), and financial objectives (net present value maximisation) generates a problem that Whittle Four-X and Gurobi MIP solve through decomposition and heuristics. Quantum approaches encode these as QUBO problems: binary decision variables for each block-period assignment, quadratic penalty terms for constraint violations, and objective functions representing NPV. Current quantum annealers (D-Wave Advantage, 5,000+ qubits, Pegasus topology) handle sub-problems of 150-200 fully connected variables. Quantum-inspired classical solvers (Fujitsu Digital Annealer) handle 100,000+ variables and produce competitive solutions today. This workshop formulates pit shell, sequencing, and equipment dispatch problems as QUBO, solves them on both quantum and classical hardware, and runs benchmark-specific performance comparisons against industry-standard tools. The honest assessment: quantum-inspired solvers are production-ready now, while native quantum hardware requires the fault-tolerant machines expected in the 2028-2032 window.

What participants cover

• Pit shell optimisation: Lerchs-Grossmann graph closure as QUBO with ore grade, stripping ratio, and slope stability constraints
• Block model sequencing: scheduling 10,000+ blocks across periods with precedence, blending, capacity, and stockpile constraints as quadratic penalty terms
• Equipment dispatch: truck-shovel allocation, haul route assignment, and crusher feed scheduling as combinatorial assignment problems
• Quantum annealing benchmarks: D-Wave Advantage on 150-200 variable sub-problems versus Whittle Four-X on matched deposit models
• Quantum-inspired solvers: Fujitsu Digital Annealer solving 100,000+ variable block scheduling instances on classical hardware at production scale
• Tailings management: waste dump scheduling, dam monitoring sensor placement, and rehabilitation sequencing as secondary optimisation targets