Machine Learning in Geometallurgy: A Review of Advances and Case Studies from Peru's Mining Sector

Aguilar M.Y.B.; Torres-Guerra J.A.; Soto-Juscamayta L.M. · Ingenierie des Systemes d'Information (2025) · DOI: 10.18280/isi.301006

Resumen

The integration of artificial intelligence (AI), particularly machine learning (ML), into geometallurgy provides an important opportunity to optimize mineral processing and mine planning. This study synthesizes recent research on ML-based geometallurgical applications and examines advances and challenges within the Peruvian mining sector. A non-experimental, descriptive methodology was employed through a systematic literature review in Scopus, ScienceDirect, and Web of Science (2013–2023). The search identified 312 records, reduced to 238 after removing duplicates. Following title and abstract screening, 170 studies were excluded, and 33 publications met the inclusion criteria, all reporting ML models incorporated into geometallurgical workflows. The selected studies were classified into six application categories, and two Peruvian case studies were examined: Sociedad Minera Cerro Verde, focused on copper concentration improvement, and the Minsur–Pucamarca Unit, centered on gold leaching optimization. Internationally, research is dominated by supervised classification algorithms for mineralogical prediction, while in Peru successful implementations are mainly associated with computer-assisted decision-making in operational contexts. At Cerro Verde, the use of Random Forest and Gradient Boosting models led to a 6.5% increase in copper production and a 0.8% rise in recovery. At Minsur, the Optimus Leach system improved gold recovery prediction accuracy (R² = 0.81) and generated USD 1.4 million in economic benefits during its first year. Overall, the findings indicate that ML-enabled geometallurgy can enhance efficiency, profitability, and sustainability when supported by high-quality data, adequate instrumentation, and multidisciplinary teams, contributing to the digital transformation of Peruvian mining. Copyright:

Instituciones: Universidad Nacional Mayor de San Marcos