With the increasing costs of extracting ores, mines are becoming more mechanized and automated. Mechanization and automation can make considerable contributions to mine productivity, but equipment failures and maintenance have an impact on the profit. Implementing maintenance at suitable time intervals can save money and improve the reliability and maintainability of mining equipment. This thesis focuses on maintainability prediction of mining machinery. For this purpose, a software tool, GenRel, was developed at the Laurentian University Mining Automation Laboratory (LUMAL). GenRel is based on the application of genetic algorithms (GAs) to simulate the failure/repair occurrences during the operational life of equipment. In GenRel it is assumed that failures of mining equipment caused by an array of factors follow the biological evolution theory. GenRel then simulates the failure occurrences during a time period of interest using genetic algorithms (GAs) coupled with a number of statistical techniques. This thesis will show the applicability and limitation of GenRel through case studies, especially in using discrete probability distribution function.

One of the objectives of this thesis is to improve GenRel. A discrete probability distribution function named Poisson is added in the pool of available probabilities functions. After improving and enhancing GenRel, the author carries out two groups of case studies. The objectives of the case studies include an assessment of the applicability of GenRel using real-life data and an investigation of the relationship between data size and prediction results. Discrete and continuous distribution functions will be applied on the same input data. The data used in case studies is compiled from failure records of two hoist systems at different mine sites from the Sudbury area in Ontario, Canada.

The first group of case studies involves maintainability analysis and predictions for a 3-month operating period and a six-month operating period of a hoist system. The second group of case studies investigates the applicability of GenRel as a maintainability analysis tool using historical failure/repair data from another mine hoist system in three different time periods, three months, six months and one year. Both groups apply two different distribution probability functions (discrete and continuous) to investigate the best fit of the applied data sets, and then make a comparative analysis. In each case study, a statistical test is carried out to examine the similarity between the predicted data set with the real-life data set in the same time period. In all case studies, no significant impact of the data size on the applicability of GenRel was observed. In continuous distribution fitting, GenRel demonstrated its capability of predicting future data with data size ranging from 166 to 762. In discrete probability fitting, the case studies indicated to a degree the applicability of GenRel for the hoist systems at Mine A and Mine B.

In the discussion and conclusion sections, the author discloses the findings from the case studies and suggests future research direction.