Geologist in the Loop: A Hybrid Intelligence Model for Identifying Geological Boundaries from Augmented Ground Penetrating Radar

Common industry practice means that geological or stratigraphic boundaries are estimated from exploration drill holes. While holes provide opportunities for accurate data at a high resolution down the hole, their acquisition is cost-intensive, which can result in number of drilled being reduced. In contrast, sampling with ground-penetrating radar (GPR) cost-effective, non-destructive, and compact, allowing denser, continuous acquisition. One challenge GPR subjectivity challenges associated interpretation. This research presents hybrid model geologist machine learning identification lateritic deposit. allows an auditable, probabilistic representation geologists’ interpretations feed into planning optimising campaigns terms density location