On the Use of Evolutionary Optimization for the Dynamic Chance Constrained Open-Pit Mine Scheduling Problem

Researchers Ishara Hewa Pathiranage and Aneta Neumann have published a paper detailing a breakthrough in applying evolutionary algorithms to one of mining's most complex optimization problems. The paper, "On the Use of Evolutionary Optimization for the Dynamic Chance Constrained Open-Pit Mine Scheduling Problem," addresses the critical challenge of scheduling extraction when both economic values of ore blocks are uncertain and mining/processing capacities change dynamically over time. Traditional approaches often treat these uncertainties in isolation, but the real world presents them simultaneously—equipment failures reduce capacity while commodity prices fluctuate.

The team's solution employs a bi-objective evolutionary formulation that doesn't just seek maximum profit but also minimizes risk by reducing profit standard deviation. Their key innovation is a diversity-based change response mechanism that repairs infeasible solutions and introduces new feasible ones when changes are detected. Tested across four multi-objective evolutionary algorithms on six mining instances, this approach consistently outperformed baseline re-evaluation methods at various uncertainty levels and change frequencies. The research will be presented at the 2026 IEEE World Congress on Computational Intelligence (WCCI), marking a significant step toward more adaptive industrial optimization.

This work bridges a gap between theoretical evolutionary computation and practical industrial constraints. By designing algorithms that respond to both value uncertainty and capacity changes in tandem, the researchers provide mining operations with tools for more resilient planning. The methodology could extend beyond mining to other industries facing similar dual uncertainties in scheduling and resource allocation, from logistics to energy production.