Updating Case Studies and Practical Insights

This chapter presents three industrial-scale applications of the updating framework introduced in Chap. 3. The first application in a gold mine applies ball mill monitoring data to continuously update the grade control model in terms of the Bond Work Index prediction. A more precise characterization of the input material in the comminution process allows for a more precise adjustment of control parameters. The second example demonstrates how coal quality data from a cross-belt online sensor are used in an open-pit mine to locally improve the coal quality model in real time. The third application focuses on the mineralogical characterization of a polymetallic ore vein in an underground mining operation. Data from hyperspectral imaging (Donner et al. 2019) and point sensor data using Laser-Induced Breakdown Spectroscopy (LIBS) or X-Ray Fluorescence (XRF) technologies (Desta and Buxton 2019) are used to continuously improve the prediction of vein geometry and the local mineral composition. All three case studies present industrial-scale results with a technology readiness level TRL 6 defined according to NASA’s definition as the “System/sub-system model demonstration in an operational environment.” This Chapter concludes with a summary of lessons learned and practical aspects when implementing a continuous feedback loop.