347 resulting framework provides a coherent analytical methodology for the interpretation of CAS interaction alerts in complex mining environments, aligned with site-specific operational interpretations. As illustrated in Figure 1, the framework is structured into three sequential and interconnected stages. Stage 1 constructs the spatial context of the mine through intelligent segmentation. Stage 2 performs contextual interpretation and prioritization of CAS interaction alerts using a fuzzy multi-expert system. Stage 3 consolidates historical results to analyze spatial and temporal risk patterns, enabling the identification of operational hotspots from a preventive safety perspective. Figure 1 – Proposed framework for CAS alert prioritization and risk analysis. Although this work focuses on Stages 2 and 3, Stage 1 is included to preserve the end-to-end methodological flow. The framework was implemented within the Hexagon Mining CAS environment and evaluated using real operational data from large-scale open-pit iron and copper mining operations in Brazil and Peru. 3.1. Intelligent Mine Segmentation (Stage 1) Stage 1 introduces spatial context into CAS alerts by automatically segmenting the mine into functional zones, enabling location-consistent risk evaluation. The mine is modeled as a dynamic set of zones, each defined by distinct operational conditions, interaction patterns, and safety constraints. This differentiation is essential, as parameters such as allowable speeds, minimum safety distances, and reaction times vary significantly between haulage routes and operational areas (Ruff, 2007). The segmentation relies exclusively on standard CAS telemetry from mobile assets. Input variables are organized into complementary operational dimensions that capture equipment mobility behavior, spatial occupancy characteristics, and contextual interaction patterns. A hybrid data-driven and expert-driven approach is employed to derive functional zones from these telemetry patterns. The method combines operational criteria defined by mine specialists with unsupervised spatial clustering techniques capable of handling noise, irregular sampling, and non-uniform spatial densities, which are characteristic of large-scale mining operations. This type of clustering approach enables the identification of coherent operational regions without requiring predefined zone boundaries. The resulting segmentation produces a dynamic set of functional zones that is periodically updated using incoming telemetry, allowing the spatial model to adapt to operational variability while providing a consistent contextual structure for subsequent stages. 3.2. Intelligent CAS Alert Prioritization Using a Fuzzy Expert System (Stage 2)
RkJQdWJsaXNoZXIy MTM0Mzk2