Underground mining is a complex operation that requires careful planning. The short-term scheduling, which is the scheduling of the tasks involved in the excavation process, is an important part of the planning process. In this paper, we propose a new method for the short-term scheduling of cut-and-fill mines.

Our problem formulation includes a new aspect of the problem, which is to handle that different excavation locations of the mine can have different priorities. The inclusion of this aspect allows the user to control the output of the scheduling and to direct resources to the locations where they are most needed according to the long-term plans. Our solution method consists of two components: a priority-based heuristic that constructs a complete solution by iteratively solving partial scheduling problems containing subsets of tasks, and a logic-based Benders decomposition scheme for solving these partial problems.

The computational performance of the proposed method is evaluated on industrially relevant large-scale instances generated from data provided by the mining company Boliden. Comparisons are made both with applying a constraint programming solver instead of the logic-based Benders decomposition scheme and with applying a constraint programming solver directly on the complete problem. The results show that our method outperforms the other two methods and yields schedules with a shorter makespan. The used instances are made publicly available to support further research in this area.