<p>Rock fragmentation by blasting is a crucial process in mining operations, particularly when dealing with hard rock masses. The economic value of minerals is intrinsically linked to achieving optimal fragment sizes. However, the heterogeneous nature of geological formations presents significant challenges in controlling the fragmentation outcomes. This study introduces a Vulnerability Index (VI) based Rock Engineering System (RES) framework for both risk assessment and prediction of mean fragment size (X₅₀) in surface iron-ore blasting. The novelty lies in adapting the VI concept to quantify fragmentation risk, incorporation of geological characterization parameters (GSI and BI) alongside conventional blast design parameters, deriving parameter weights from an Expert Semi-Quantitative (ESQ) interaction matrix within a RES framework, and validating the RES-VI predictive model using eight independent blasts and comparing performance to an Adaptive Neuro-Fuzzy Inference System (ANFIS). Data from 26 blasts and 20 influential parameters were used to build the RES model. The analysis revealed that achieving precise fragmentation control remains challenging, with the VI indicating medium to high risk levels (VI = 35.25, Category II), reflecting the inherent complexity of fragmentation behavior under heterogeneous geological conditions. The RES-VI approach not only classifies blasts into risk categories (low-medium, medium-high, high-very high) but also provides a predictive regression for X₅₀ (R² = 0.81; RMSE = 2.24) that outperforms the ANFIS benchmark (R² = 0.44). The integrated RES-VI methodology therefore offers mine engineers a transparent risk-aware tool for optimizing blast designs and mitigating adverse fragmentation outcomes.</p>