Leverage Predictive Analytics For Motors
EP Editorial Staff | April 25, 2019
The Internet of Things is improving countless operations in plants, including the way motor health is managed.
By Jim Sanderson, Motion Industries and Thomas Schardt, Nidec Motor Corp.
Predictive analytics, which uses existing data to identify patterns and predict future outcomes, offers substantial value in industrial settings. Manufacturers can use it, for example, to know in advance if a mechanical issue could result in failure of a motor or driven equipment. Likewise, condition monitoring—a major component of a predictive-maintenance (PdM) strategy that monitors an asset’s actual condition to guide maintenance needs—provides valuable insight into the health of critical equipment that allows corrective action(s) before a catastrophic breakdown occurs.
Both of these approaches can benefit the millions of installed industrial motors that are not currently being monitored. That’s because every new motor begins to deteriorate the minute it is put into service.
It helps to think of a new motor the same way we think about a new car. When you drive a new car off the lot, deterioration begins—even though it is almost impossible to detect. This is why automobile manufacturers recommend scheduled maintenance. The desired outcome is to keep the vehicle running in top condition over time.
In industrial environments, scheduled maintenance—sometimes called preventive maintenance (PM)—serves a similar purpose. This strategy has grown more common as sites have shifted away from reactive mindsets.
THE GOOD NEWS
Reactive maintenance is not likely to disappear altogether. In applications where downtime costs are less critical, companies will continue to run certain motors to failure, replacing them as necessary.
For critical equipment, downtime can be costly. Owner/operators of such equipment face financial consequences if these assets aren’t properly maintained. For this reason, many have implemented PM programs to prevent motor failure and help assure maximum uptime.
In many cases, preventive maintenance is performed using route-based, periodic monitoring. This approach requires either an in-house or third-party expert to periodically travel to an industrial site to collect motor-performance data, and similar information from other assets. The data can include vibration levels, temperature, speed, and other applicable parameters.
This information is taken back to an office, where it’s uploaded into a database and compared to prior sets of asset operating data. The process creates operating trends that can be reviewed to identify potential problems. Trend information can also support decision-making as you consider whether to schedule maintenance or perform complete replacement of failing motors.
The above method does have its drawbacks, though. One is lag in response. Production schedules and the speed of operation will change. Temperature and vibration may reach critical limits under different speed or load conditions. From the time of data collection and evaluation until implementation, a failure could occur. The good news is, approaches to monitoring and, in turn, analytics have improved.
THE BETTER NEWS
As helpful as route-based monitoring can be, it offers just a fraction of the value to be gained when the IoT (Internet of Things) is added to the mix. When motors are equipped with IoT technology, route-based monitoring can be transformed to predictive maintenance (PdM). The same well-trained experts can be used to provide remote continuous condition monitoring, which is the foundation for improved machine reliability.
Continuous condition monitoring leading to PdM is scalable and can be expanded to hundreds of IoT-enabled motors, depending on the service provider’s capacity and owner/operator’s preferences. At whatever scale is chosen, the goal is the same: to predict equipment failure before it happens and so avoid consequential damages.
Unexpected breakdowns not only lead to costly downtime, but also further damage to equipment. IoT-enabled technologies help maximize equipment life without increasing failure risks. It not only identifies the best time for maintenance, but also the scope of work that should be performed.
Demand for continuous condition monitoring is increasing among small- and medium-sized enterprises. Sensors, processing power, communication capabilities, and data-storage costs have dropped dramatically in the past few years. This has put IoT technologies within reach for many more industrial monitoring and control-system applications than thought possible two or three years ago. The situation has great significance for a site’s electric-motor fleet.
Advances in reliable, automated wireless sensors have led to affordable multifunctional, wireless monitoring. Once collected, data is processed and put in some form of graph or chart, then personnel can begin to determine the equipment’s current health. The plant operator can trend these critical parameters and set thresholds for earlier warning, in real time.
Although this is a good format to monitor critical parameters and react as they move out of acceptable ranges, the ultimate process monitoring and evaluation uses predictive analytics. Predictive analytics adds a more sophisticated approach to this monitoring, to gain even more value and address issues that may arise. In this approach, algorithms learn, model, and identify patterns of the rotating equipment, not just the process limits. This provides the ability to trend individual processes over time and determine a course of action.
With predictive analytics and the use of wireless sensors, the trending process can be adjusted for the machine operation and determine patterns for early detection and protection. By adding such analytics, a site can create a true PdM system. As operations transition toward continuous real-time monitoring of motor health, the number of IoT-equipped motors will continue to grow. So will the operating efficiency and productivity of the plants these fleets serve. EP
Jim Sanderson is a division electrical engineer at Motion Industries, Birmingham, AL. Thomas Schardt is senior director of IoT for Nidec Motor Corp., St. Louis. Visit MotionIndustries.com/efficientplant and TheMotorSpecialist.com.