Open AccessUnderground Mine Drilling Predictions using Artificial Neural Networks for Short Term Mine Planning
Author(s) -
J. Del Rio,
Sebastián Parra Arenas,
Giovanni Franco Sepúlveda
Publication year - 2020
Publication title -
journal of mining engineering and research
Language(s) - English
Resource type - Journals
ISSN - 0719-9961
DOI - 10.35624/jminer2020.01.11
Subject(s) - artificial neural network , productivity , term (time) , adaptation (eye) , mining industry , engineering , data mining , drilling , plan (archaeology) , computer science , artificial intelligence , mining engineering , geology , mechanical engineering , paleontology , physics , quantum mechanics , optics , economics , macroeconomics
Performance of jumbo operators directly affects productivity in current mining opera-tions. Implementing tools to attack low productivity sources is the main focus of the mining industry due to the transition from “easy” deposits to remote, low grade and deeper deposits. Moreover, im-proving equipment productivities strongly impact by producing higher values of outputs with the same values of inputs. This paper presents a method to predict the performance of jumbo operators, seeking to link this predicted data into an efficient short mine planning. During 2018 several data related to jumbo operations was collected in order to create a manual control room to analyze real downtimes and their main sources. After data collection, typing, filtering, and adaptation were performed in order to prepare the database to train the Artificial Neural Network (ANN). Once the ANN was trained, an R2 of 0.9998 was obtained between the real value and the predicted value, finding that the proposed methodology is of great help when carrying out the mining planning in the short term within the mining operation.