• Expertise in open-pit and underground mining
    Mine Planning

SHORT-TERM MINE PLANNING

In open-pit mining, short-term planning focuses on creating a dependable production schedule that aligns with the long-term vision. The resulting short-term strategy needs to adhere to mining, processing, and plant limitations while optimizing key objectives such as overall metal yield, net profit, blending, or striping ratio. At APMT, we continually innovate, developing tailored algorithms to offer client-specific solutions on:

Heuristics and mixed-integer programming (MIP) models
Blast-hole data assimilation
Risk-based economic analysis

LONG-TERM PRODUCTION SCHEDULING

In open-pit mining, long-term scheduling determines the extraction timeline for blocks and sets the appropriate cut-off grades, and bench/phase definition. This schedule is optimized to maximize the project's net present value while aligning with corporate strategies, considering mining, processing, and metallurgical constraints, and maintaining balanced stripping ratios. At APMT, we employ advanced algorithms to address and analyze long-term mine planning challenges, including:

Strategic mine planning
Stochastic optimization
Real-options-based strategies under geological uncertainty
Risk-based economic analysis

MINE SCHEDULING WITH REINFORCEMENT LEARNING

At the frontier between AI research and mining application, we reframe the problem of mine scheduling as a reinforcement learning task for short-term and long-term planning. Using a strategic workflow we provide optimal mine plans accounting for:

Uncertainties on geological attributes, metal price forecasting, metallurgical responses, and mining costs
Geometric precedence, open mining faces, and processing capacities

GRADE CONTROL MODEL

Grade Control aims for the integration of data from different sources and supports into the resource models to deliver more accurate local grade and geological attribute estimates. At APMT we focus on delivering:

High-resolution grade control models
Data assimilation with different supports
Robust and reliable data assimilation workflows