Artificial Intelligence and Digital Twin Technology in Borehole Mining

Borehole Mining (BHM) is a remotely operated, waterjet-based mining method that allows mineral extraction through a borehole without human presence underground. This remoteness is one of BHM’s principal advantages, but it also creates a fundamental operational limitation: the operator has only limited, indirect information about the conditions and geometry of the underground excavation. As a result, important decisions are often reactive and may contribute to ore dilution, under-recovery, premature stope collapse, uncontrolled subsidence, or inefficient use of water and pumping energy. This paper proposes a conceptual approach in which geological characterization, physical-property testing, three-dimensional modeling, and rock-mechanics simulation provide the knowledge base for a BHM Digital Twin (DT). The DT would represent the orebody and hosting rocks together with their relevant physical and geomechanical properties and would provide predicted responses to alternative extraction scenarios. During mining, monitoring data could be compared with these predictions by Artificial Intelligence (AI) and Machine Learning (ML), allowing deviations, patterns, and potential hazards to be recognized and communicated to the human operator. The proposed AI+DT concept would therefore move BHM from predominantly reactive operation toward proactive, human-supervised decision making, with the potential to improve safety, reliability, resource recovery, and environmental performance.