This article describes a new model aiming at optimizing short- and medium-term underground mine scheduling. The complexity of the problem to solve and the frequency at which planners have to revise these schedules are among the main motivation for developing such a model. In order to address this problem, a Mixed Integer Programming model was developed with a variable time discretization to accurately represent both short- and medium-term operational constraints in a single model. Results of a preliminary model are presented with explanations an in-depth analysis. An improved formulation is also described with it's associated results and benefits. Further testing with scenarios similar to long-term planning show very promising results for the possible application of our modified formulation to existing long-term model.