Automation of Tribological Analysis of Mining Equipment through Machine Learning

Katia Huaman Lima I.; Gladys Cutipa Arapa M.E.E.; Villafuerte Serna M.E.R. · IEEE Andescon, ANDESCON 2024 - Proceedings (2024) · DOI: 10.1109/andescon61840.2024.10755834

Resumen

In recent years, artificial intelligence has been gaining ground in various fields, and the mining industry is no exception. This article presents a comparative analysis of different machine learning techniques applied to the tribological analysis of mechanical equipment components. Machine learning-based approaches are considered promising for increasing safety and optimizing preventive maintenance schedules, thereby extending the lifespan of mining equipment. For this, oil analysis is crucial. In the project, a methodology based on statistical tests was used to determine anomalies in oil samples. On the other hand, subjective interpretations by those responsible for the analysis lead to significant errors and unscheduled equipment shutdowns. Implementing machine learning promises to automate and refine this analysis, allowing for more accurate and earlier detection of potential failures, thus increasing equipment reliability. Various machine learning algorithms will be evaluated to develop an anomaly detection model for oil samples and a classification model for the component's condition based on its oil sample.

Instituciones: Universidad Nacional de San Antonio Abad del Cusco