Mining is an asset-intensive industry. Not only is there a lot of different equipment and machinery involved, but these assets are expensive, complex, and often located in remote, relatively inaccessible areas. The mining industry spends billions annually on maintaining this equipment. According to Mining Global, maintenance costs accounted for 30–50% of a mine’s total operating expenses in 2014, and more than 60% of the total mine workforce can be focused almost exclusively on servicing or repairing assets in the field. Increasing asset performance to extend their useful lifetime, decrease downtime, and improve overall equipment efficiency can make a huge difference to the bottom line.
Traditional reactive (“fix it when it breaks”) or pre-scheduled preventive maintenance approaches often result in premature equipment failure, unnecessary expense, and lost production caused by downtime. They also do not account for performance inhibited by unplanned variations (slow-running conveyors, sticky valves, blocked chutes, etc.). Recent ARC Advisory Group research shows that only 18% of assets have a failure pattern that increases with use or age — meaning that a more proactive strategy than preventative maintenance alone is needed to avoid failure in the other 82% of assets.
Fortunately, such a proactive approach is now far more affordable thanks to the phenomenon known as the Internet of Things (IoT). As technology continues to advance, the cost of IP-enabled sensors and other smart devices continues to decrease. Mining equipment and machinery are increasingly equipped with intelligent devices that are able to sense, generate, and transmit data. Compressors, generators, pumps, fans, blowers, heat exchangers, boilers, ovens, kilns, pulverizers, crushers, gearboxes, and condensers are just some of the many assets that can be monitored using sensor data. Predictive maintenance continuously monitors asset performance through sensor data and prediction engines to identify potential issues before they become problems. Schneider Electric has found that an optimized maintenance strategy can reduce the overall maintenance budget by 5% and reduce the number of work stoppages by 50%.
An analytics software program is needed to translate the massive amounts of raw IoT data into actionable intelligence and insight into the health and performance of equipment. In effect, analytics software bridges the gap between data and action by helping operators draw reliable conclusions about current conditions and future events. Predictive analytics software uses historical operational signatures for each asset and compares it with real-time operating data to detect subtle changes in equipment behavior. The software is able to identify changes in system behavior days, weeks, or even months before the deviating variables reach operational alarm levels. This early warning detection and diagnosis creates more time for analysis and corrective action. By closely monitoring the equipment, managers can better understand the real reasons for variations and delays.