A wide variety of data is available thanks to digital drive technology, which allows operators to keep the status of a drive or machine under constant observation. Combined with suitable analytical tools, this could be even used for predictive maintenance.
Traditional maintenance methods, in which systems are inspected at regular, set intervals, are not based on the actual condition of the systems in question. This means that systems may be checked either too early or too late.
While premature maintenance is acceptable and has little effect on cost, belated maintenance could mean machine damage or failure. In other words, with consequences that are just as serious and costly. Condition-based maintenance of drives is necessary to avoid this outcome, for instance.
Digital drive technology provides a key foundation for this process, so permanently recording machine and drive data. As a result, the actual condition of a component or system to be continuously monitored and conclusions to be drawn about their functionality.
Condition monitoring makes it possible to intervene in good time as soon as an unusual change in the machine’s condition is detected. Indeed, before the drive has the chance to fail.
When Drives Become Current Sensors
A vast array of different data sources are possible in this area. Electric motors and their actuators provide valuable information. Frequency converters, for example, collect a large amount of data that enables operators to reach conclusions about motors’ working condition.
Also a direct indication can be given, if it is necessary to check the bearings of motor and driven components. This is when a little more energy is needed to reach the same speed even though the process has not changed.
Further data can be recorded via additional sensors to gain a deeper understanding of the drive’s condition, as well as to predict possible wear, tear and failures.
Temperature sensors act as important indicators with electric motors, for instance. Because overheating is a sign of a defect that may result in future system failure. In addition, acoustic and vibration sensor data can be a useful source of information about the status of roller bearings. Or they drives as a whole, with vibration sensor technology offering particularly beneficial results in this regard.
Detailed manufacturer databases are available especially for the bearings used in industrial drives. These databases contain all components of each bearing type with their characteristic vibration frequencies. Therefore, with such manufacturer databases, the individual frequencies can be clearly identified and allocated.
Outliers from the typical frequency spectrum indicate early-stage damage.
Condition Monitoring with MEMS
Market research by Technavio has found that the use of vibration sensors in predictive maintenance processes is one of the factors driving solid growth in this segment. The global vibration-meter market is predicted to grow by 5 % a year on average between 2019 and 2023.
According to Technavio, micro-electromechanical sensors (MEMS) are seeing increased use in this area. Because they offering a cheaper, smaller and more energy-efficient alternative to conventional sensors.
MEMS is also part of a new system that uses the ultrasonic output from a machine or drive. The small MEMS microphones measure noise from 10 to 50-plus kilohertz. Changes within this range can point users to damage in fans’ bearings or dirty air inlets.
These systems’ sensitivity needs to be higher than that of conventional vibration or acoustic sensors, allowing damage to be detected at an even earlier stage.
Drives can also be retrofitted with modern condition-monitoring systems. Various drive manufacturers have developed compact sensor tags that are simple to attach to the drive’s housing. They also provide information about the operating parameters, such as vibrations, temperature and overloading. Additionally, they can then send the data to a smartphone or to the cloud via wireless communications interfaces for further analysis.
While condition monitoring only supports recognition of the current situation, predictive maintenance can be used for long-term planning of when maintenance will be necessary or how long a component will continue to work reliably.
This means higher system availability, reduced costs, a longer drive service life and – most importantly – no unplanned downtime. To this end, patterns are detected from the condition-monitoring data to create insightful predictions.
Artificial intelligence (AI) is playing an increasingly important supporting role in this process, as the amount of data generated by the drives and sensors is too large to be quickly evaluated and interpreted by human talent alone.
The plant operator can only benefit from the data that has been collected once it is processed by suitable methods of mathematical AI analysis.
Today, these methods take the form of analytical tools requiring high computing capacity, mostly stored in the cloud. As a result, the communication capability of digital drives is essential. The data can only be transferred from the motor to the plant controller to the cloud through the appropriate interfaces.
With the growing intelligence of the drives themselves and the effectiveness of their controllers, however, these analyses can increasingly be shifted to the drives.
This way, at the very least, simple factors can be evaluated locally in a direct process and displayed on a device such as the plant operator’s smartphone.