Data mining in mining engineering: results of classification and clustering of shovels failures data. Issue 2 (17th February 2017)

• Record Type: Journal Article
• Authors: Dindarloo, Saeid R.; Siami-Irdemoosa, Elnaz
• Is Part Of: International journal of mining, reclamation and environment. Volume 31:Issue 2(2017)
• Page Start: 105
• Page End: 118
• Publication Date: 2017-02-17
• Subjects: Shovels failures -- classification -- clustering -- data mining -- reliability -- support vector machine; Mining engineering -- Periodicals; Mineral industries -- Environmental aspects -- Periodicals; Abandoned mined lands reclamation -- Periodicals; 622.292
• DOI: 10.1080/17480930.2015.1123599
• Languages: English
• ISSNs: 1748-0930

Abstract

Abstract: The high ownership cost of mining equipment mean that downtimes are expensive and should be avoided with smart and efficient maintenance planning. Modern mines have large data sets on equipment performance and reliability, from dispatch and manufacturer health monitoring systems, that can be mined for more efficient maintenance planning. This study explores the application of classification and clustering approaches for pattern recognition and failure forecasting on mining shovels. The failure behaviour of a fleet of ten mining shovels during 1 year of operation was investigated using these techniques. The shovels were classified into four clusters using k -means clustering algorithms. Future failures were predicted using the support vector machine (SVM) classification technique. Historical failure and time to repair data were used to predict the next failure type for all shovels. The SVM technique was shown to be successful with prediction accuracy of over 75%. This is the first attempt (to the best of our knowledge) that the failure type is predicted based on historical failure/repair data for mining equipment. Clustering shovels based on their reliability can be used for equipment allocation and maintenance planning. These objectives cannot be achieved with traditional reliability modelling. Successful application of these techniques will be valuable input for decision-making during preventive maintenance scheduling.