Industrial mining and power generation facilities depend on continuous conveyor operation to clear mines and feed furnaces. Almost the entire profit is made or lost with only a few percentage points of uptime. Threats to production availability, such as conveyor motor failures, are nearly impossible to detect ahead of a failure event. Failure to detect impending changes in production processes results in catastrophic damages, unscheduled downtime, unplanned capex, and lost revenue.
Motor Control, Motor Current, and Temperature are available from PLC systems operating the conveyor belt pulleys and are fed into Splunk for long term storage. However, this data is currently not visualized in the HMI.
Customer’s Splunk server is integrated with Falkonry through the Falkonry Splunk App providing a direct data link between Splunk and Falkonry. Falkonry builds an AI model of pulley motor behavior from the data as it starts to flow through Splunk. Falkonry detects the earliest indication of degradation in conveyor motor behavior by examining time-series data. Indication of motor degradation is returned to Splunk from which notifications are generated when an unknown pattern of behavior arises to allow an engineer to investigate the issue. For named behavior patterns, notifications inform operators to idle the errant motor or conduct a controlled conveyor shutdown in order to prevent the pulley wire from breaking.
Instead of running to failure, conveyor operators can detect subtle changes in patterns in the combination of control and monitoring signals in time to avert costly failures. Knowing at the earliest possible time of an impending change in pulley motors can enable controlled or scheduled shutdowns reducing waste, collateral damage, and minimizing production downtime.