Por: Hernando Valdivia Lozada, superintendente de Metalurgia; Héctor Paredes Chirinos, superintendente de Control de Procesos; David Estrella, gerente de Concentradora y Alexander Álvarez, ingeniero de Analítica Machine Learning, Compañía Minera Antamina. 

Abstract

In Compañía Minera Antamina; as part of the constant technological innovation in its polymetallic mineral concentrator, Advanced Data Analytics has been used, through the application of Machine Learning to increase its metallic production of copper equivalent, generating articulated models that maximize the tonnage processed in SAG milling, as well as increase the metallic recovery in copper flotation and zinc flotation, without affecting the capacity of the process or the quality of the concentrates to be produced. 

The initial work of 04 months was focused on the analysis and knowledge of the process, strategy planning, information governance, ideation workshops, application of agile methodology, exhaustive analysis of variables with direct influence on production, determining the real capacity of teams in the operation, establishing bottlenecks, opportunities for improvement and in parallel the data architecture was generated in the cloud.

As of August 2021, we were able to implement Machine Learning models, with the involvement of areas such as Geology, Mine Planning and Operations, Concentrator Maintenance, Concentrator Operations, Process Control, Information Technology, Six Sigma Analysts and the Metallurgy team, with the support of an international consulting firm and the creation of the multidisciplinary work team "MAYTA" (Quechua word meaning "ONE ONLY").

The real benefit obtained was a 3.7% increase in Cu equivalent over the baseline, thanks to the implementation of the recommendations made by models and evaluated by the operational team.