Predictive Condition Monitoring
Predictive Condition Monitoring
Moving from preventive to predictive maintenance.
Problem
Increasing reliability and lowering maintenance costs are key differentiators in today’s highly competitive mining environment. Understanding mine equipment criticality demonstrates where maintenance effort should be focused to mitigate risk. One of the areas of focus based on criticality is conveying systems. These conveyors are typically single points of failure and, therefore, have high criticalities.
The conveying system’s multiple rotating components, length and remoteness increase the complexity of maintaining these systems. When conveyor operations are closely monitored, it’s possible to predict mechanical failures before they occur. In fact, recent studies indicate up to 90% of machinery malfunctions are predictable. In many cases, it can be performed online. The largest mining company in India seeks online condition monitoring solutions for their critical assets of coal washery.
Solution
As a part of solution implementation, smart vibration sensors were commissioned on conveyor belt systems across the coal washery. By smart sensors we mean the sensors equipped with edge computing features. These sensors understand the vibration patterns, machine kinematics and intelligently push processed data to cloud servers. Next stage of intelligent computing happens on the cloud where the machine learning algorithms work on the data and identify hidden patterns in data.
These intelligent algorithms continuously acquire insights from machine behaviour to predict the next breakdown event. Customers migrate themselves to a platform where they no longer need to worry about next maintenance activity and machine health analytics. It’s all intelligent now!!!
Statistics
Online condition monitoring of 12 critical assets
4% overall savings in maintenance costs
Ability to detect transient and spurious events
Benefits
The solution enabled the stakeholders to diagnose the problem early with the predictive health algorithms doing their job. Now there is no need of unplanned diagnostic plant shutdowns in the middle of the day. With the operational data already been available on Adani cloud data center, it has become much easier to correlate machine health with operational parameters.