Our Introduction to Predictive Maintenance article explained predictive maintenance and discussed the difference between reactive and proactive maintenance. In this article, we will break down popular maintenance strategies in the mining industry.

Preventing equipment failures and reducing unplanned maintenance is critical to productivity and the bottom line. But which maintenance strategy is best when it comes to maintaining equipment and preventing costly downtime? What is the difference between preventative maintenance, condition-based maintenance and predictive maintenance?

Preventive maintenance (PMs)

Preventative maintenance (PM) depends on fixed factors such as service recommendations to maintain mining equipment and assets. Preventive maintenance is completed at predetermined intervals and does not consider the equipment condition.

While preventative maintenance aims to prevent and reduce equipment downtime, machines don’t wait for routine maintenance intervals to fail.

Condition-based maintenance (CBM)

Unlike preventative maintenance that relies mostly on routine maintenance intervals, Condition-based maintenance continuously monitors equipment performance with a combination of sensors, scheduled tasks and visual inspections.

Condition-based maintenance can significantly reduce maintenance costs compared to preventative maintenance or a run-to-failure strategy as maintenance and repairs are based on equipment’s current condition.

Condition-based maintenance shares common traits with predictive maintenance using sensor data to measure and predict equipment failure, but Predictive maintenance takes the usage of the data a step further.

Predictive maintenance (PdM)

Predictive maintenance is a maintenance strategy that forecasts machine failure and helps identify underlying problems that need to be fixed.

Predictive maintenance takes the sensor data and uses algorithmic formulas to predict machine failure. Our team of data scientists developed DINGO’s predictive analytics models using real failure data and then validated by maintenance experts.

By continuously improving the health and performance of your equipment, DINGO’s solutions increase availability, extended component life and reduce operating costs.

Case Study: Predictive Maintenance in the Real World

A Dingo customer was experiencing unexpected failures in their Ball Mill drive gearboxes and pinion bearings. There was a large amount of condition monitoring data available (vibration, pressure & temperature sensors) but only a small number of failure records.

The challenge was to isolate the failure mechanisms (no failure classifications provided) and develop a predictive model that would allow component repair/replacement to be scheduled during quarterly plant shutdowns.

Dingo’s data science team developed an anomaly detection algorithm that determined the indicators whose degradation factors correlated best with gearbox failures, separately to pinion bearings.

From the outputs of the anomaly detection algorithm, Dingo created a predictive model that could predict the probability of failure over time for each gearbox and pinion bearing. 

The model was back tested against previous failure data to determine the appropriate confidence level of the model.

Back testing results show 85% accuracy in predictions, correctly identifying past failure events separately to repair events.

Implement Predictive Maintenance at Your Operation

Unlike traditional maintenance methods, DINGO’s Trakka provides early warning indicators and insight that increase availability, fewer unplanned failures, and a reduction in overall maintenance time and cost.

For more information, contact a DINGO expert.