3D Stope Sequencing & Grade Blending.
A five-part blueprint showing how an Eve Count engagement documents an industrial problem, a classical baseline, a Hamiltonian mapping, a Nexus trapped-ion compilation and the executive decision gates.
Evidence status
This page is an illustrative method blueprint, not a completed client benchmark. Figures such as the 8.4% MILP optimality gap and 62% circuit-depth reduction are candidate comparison inputs to be reproduced and verified during a paid engagement; no live hardware advantage or physical-qubit crossover threshold is claimed here.
Inside the feasibility blueprint.
Part 01
The operational problem
A multi-pit copper/gold operation with non-local geometric extraction cones and dynamic smelter silica tolerances below 3.2%.
Report evidence: extraction precedence graph, block model assumptions, blending constraints, time horizon and agreed objective.
Part 02
The classical baseline
Blueprint baseline input: Gurobi 11 MILP terminates at an 8.4% optimality gap after six hours on a 64-core HPC node, prompting block aggregation.
Verification required: model and instance size, solver logs, configuration, hardware specification, random seeds, incumbent and bound. These supplied figures are not a published measured result.
Part 03
The Eve Count formulation
Map the extraction precedence matrix and blending penalties to a high-density Hamiltonian, investigating native trapped-ion ZZ(θ) interactions.
Report evidence: encoding, penalty calibration, constraint satisfaction, qubit requirements and a comparison with the original objective. A Hamiltonian mapping alone is not an advantage result.
Part 04
The Nexus benchmark
Compile via pytket to a trapped-ion emulator on Nexus. The supplied blueprint sets a 62% circuit-depth reduction against a planar swap-mapping baseline as a comparison figure to validate.
Verification required: matched circuits, compiler versions, optimisation settings, gate counts and noise assumptions. Emulator compilation cannot substantiate the supplied claim of zero gate-fidelity degradation on live hardware.
Part 05
The executive outcome
A validation report identifies measured resource costs and a conditional hardware-scale horizon for testing trapped-ion QAOA against classical heuristic aggregation.
Decision gates: reproduce the baseline; validate the encoding; assess noise sensitivity; estimate scaling; decide whether to commission a pilot. No algorithmic crossover or exact physical-qubit threshold is established by this blueprint.