At Rio Tinto Bingham Canyon mine, large capacity heavy mining equipment (HME) such as electric and hydraulic shovels, drills, and haul trucks are used for economical and efficient excavation of waste and ore. Unscheduled equipment downtime causes major production losses. It also impacts safety of personnel and adversely affects the environment. Over the years, an extensive amount of data has been generated on equipment reliability and downtime. This data can be mined to provide insights into the failure indicators and areas to focus on in order to improve equipment reliability. Defect Elimination (DE) is currently done by looking at Pareto charts and focusing on those events or breakdowns with the longest duration and the most occurrences. While this approach considers the two most common variables, the computerized maintenance management system (CMMS) and mine monitoring and control center (MM&C) can generate numerous variables. These variables can be simultaneously extracted for knowledge and information to arrive at the optimal decisions for DE. This presentation will highlight the application and benefits of leveraging big data and machine learning on current and historical equipment reliability downtime events. This will allow us to better understand, predict, and control the condition and performance of HMEs. Both supervised and unsupervised machine learning methods will be used to extract valuable knowledge and information from the data for pattern recognition and classification.
“R.A.I.” the Reliability.aiTMChatbot
You can ask "R.A.I." anything about maintenance, reliability, and asset management.