Unlike conventional sampling, which is based on discrete campaigns with frequencies on the order of days, LIBS technology enables real-time data acquisition with measurement frequencies of approximately 1000 Hz (1000 pulses per second). This high temporal resolution allows capturing short-term variability of the transported material, providing a more representative characterization of the grades delivered to the process and a better understanding of ore dynamics within the operation. 3.2 Calibration Strategy The initial stage includes system calibration under controlled conditions, using representative ore samples previously analyzed in the laboratory. For this purpose, approximately one tonne of representative material from Chuquicamata Underground was collected, covering the full range of expected concentrations of Cu, Mo, Fe, and As. The sampling campaign included the selection of 50 extraction points, from which approximately 20 kg of material was collected per point, following a prior assessment of grade representativeness. These samples cover the expected range of elemental concentrations, enabling the development of calibration curves, the establishment of relationships between spectral intensity and concentration, and the evaluation of repeatability, accuracy, and detection limits. This phase corresponds to factory validation, where the analytical performance of the system is verified prior to field installation. 3.3 Field Integration and Operational Implementation During the commissioning stage, the LIBS system is integrated into the plant operational control platforms (PI System), enabling real-time visualization and use of grade information under operating conditions. In this phase, tests are conducted with operating material, together with validations in controlled environments such as pilot plant conditions, to ensure that detection limits, accuracy, and sensitivity are representative and suitable for the characteristics of Chuquicamata Underground ore. System integration allows continuous grade measurement to be incorporated as a key operational variable, enabling its direct use in decision-making. In particular, real-time data supports the optimization of reagent dosing, improved control of feed quality, and increased metallurgical stability of the process. This approach supports the transition toward a data-driven operational model, reducing uncertainty in grade estimation and strengthening mine–plant reconciliation through improved traceability and representativeness of information. 3.4 Operational Validation, Reconciliation, and Monitoring Once in operation, a continuous validation process is carried out, based on systematic comparison between LIBS analyzer measurements, laboratory assay results, and metallurgical balance data. This analysis allows evaluation of system accuracy and reliability, the establishment of correlation factors between different data sources, and the quantification of the reduction in the mine–plant reconciliation gap. In parallel, an assisted operation and monitoring phase is implemented, aimed at stabilizing
RkJQdWJsaXNoZXIy MTM0Mzk2