The project aims at developing a generalized model or unified model applicable for ore dilution minimization and stope stability, regardless of the stope size, mining methods and site conditions. It is motivated by the need for further studies to continuously improve understanding towards the complex rock mass behavior surrounding stopes as unplanned stope instabilities and dilution pose a serious threat to production and safety. This project has two major components (work packages): stope performance prediction using the stability graph methods, and stope stability analysis based on numerical modelling. The tasks to be delivered include: compilation of a database of stope performance and characterization of stope performance; improving the graph methods: probabilistic, machine learning and rock engineering system approaches; investigating the effect of stress relaxation; approach toward a generalization of the method; determining how the influencing factors (orebody geometry, stress relaxation, faults, etc...) affect the stope performance via numerical modelling using FLAC3D, Map 3D and/or Rocscience.