Predictive Maintenance (PdM) monitors the performance and condition of equipment directly during normal operations in order to predict failures ahead of time. The goal of PdM maintenance is to perform maintenance only when necessary in order to prevent unnecessary preventative maintenance tasks from being performed as well as avoid unplanned breakdowns from occurring.
When Predictive Maintenance is employed, organizations are able to consistently test and monitor conditions like corrosion and lubrication. Methods used to accomplish PdM include oil analysis, sound level measurements, vibration analysis and infrared testing.
Preventive vs. Predictive Maintenance
Although the best maintenance programs will include a good balance of both of these, predictive maintenance and preventive maintenance employ different strategies. Preventive maintenance is determined using the expected or average life cycle of equipment or another asset, whereas predictive maintenance is based on what the equipment’s current condition is.
Although it is more complex to get predictive maintenance established, it can save a business more money and time. For example, when vibration measurements are taken on an electric engine based on recommended intervals it will be able to detect wear more accurately and allow the business to replace a bearing or take other action before a total failure happens.
How Predictive Maintenance Works
The current condition of equipment is evaluated by PdM through performing continuous or periodic equipment condition monitoring. A majority of PdM is performed during normal operations of the equipment in order to minimize everyday operations being disrupted. Statistical process control principles are leveraged by this maintenance strategy in order to determine when it is going to be necessary to perform maintenance tasks in the future as well as currently.
For example, instead of changing the oil on a vehicle based on being driven 3,000 mails, with PdM the organization will collect sample data of the oil and the oil will be changed based on asset wear. To be effective, predictive maintenance requires both hardware for monitoring equipment as well as software for generating work orders whenever a potential problem has been detected.
Types of Predictive Maintenance
Vibration analysis: Degradation in the performance of equipment like motors and pumps can be detected through using vibration sensors.
Infrared: Frequently infrared cameras are used for identifying temperature conditions that are unusually high.
Acoustic analysis: Ultrasonic or sonic tests are performed to find liquid or gas leaks.
Oil analysis: Asset wear is determined through measuring the number and size of particles in the oil.