1. INTRODUCTION In open-pit mining, effective grade control at the mine face remains a persistent operational challenge due to the disconnect between digital grade models and real-time field decision-making. Conventional grade control practices rely heavily on printed grade maps, handheld devices, and screen-based visualization tools that provide static representations of ore distribution derived from historical data [1]–[3]. These representations are often outdated by the time they are used at the bench and fail to reflect rapidly changing geological and operational conditions during excavation. As a result, operators must make critical decisions without timely, spatially accurate information, increasing the risk of ore–waste misclassification, dilution, and resource loss, particularly in deposits characterized by high geological variability[2], [4]. Although significant progress has been made in digital mining technologies, including realtime mining (RTM), GIS-based control systems, and online grade monitoring frameworks, their practical impact at the mine face remains limited [4], [5] . Many RTM implementations focus on backend data assimilation, model reconciliation, and post-processing workflows, with updated grade information often becoming available only after production activities have already taken place [6]. This delayed feedback loop reduces the ability of mine personnel to respond dynamically to grade variability during drilling, blasting, and excavation, thereby limiting the operational value of otherwise advanced digital systems. Consequently, grade control remains largely a retrospective analytical task rather than an active, real-time operational process. A further limitation arises from the cognitive burden placed on operators who must translate two-dimensional plans or screen-based three-dimensional models into the physical geometry of the mine bench, which is particularly challenging in complex geological settings with irregular ore-waste boundaries[7], [8]. This mental reconstruction of spatial information is errorprone and becomes increasingly difficult in complex geological settings, where ore-waste boundaries are irregular and grade transitions occur over short distances[7], [9] . Although advanced technologies such as 3D visualization, AR, VR, and geostatistical modeling techniques have been proposed to improve spatial understanding, these tools are typically confined to office environments and do not provide real-time, context-aware visualization at the point of extraction [10], [11]. As a result, the gap between digital models and physical operations persists at the bench level. High-precision positioning technologies, such as GNSS and machine guidance systems, have significantly enhanced the spatial accuracy of mining equipment; yet their integration into operators remains suboptimal. Operators often interact with these systems through vehicle control interfaces or dashboards, limiting their usefulness for human-centered decision-making[12]–[14]. Although operators work in close proximity to highly accurate spatial data, they often lack intuitive access to this information in a form that aligns with their physical view of the mine face. This disconnect constrains effective human–machine collaboration and undermines trust in digital systems, as operators cannot easily verify or interpret model-derived information in real time [15]. Bridging this interface gap requires visualization approaches that deliver accurate, reliable, and immediately interpretable spatial information directly to the operator in the field. As mining enterprises began to adopt digital twin methodologies and data-driven optimization techniques, the lack of real-time, spatially precise visualization at the mine face constitutes a significant obstacle to effective grade control. This paper seeks to address this
RkJQdWJsaXNoZXIy MTM0Mzk2