Maintenance has been evolving with new technologies and strategies since the days of CH Waddington during World War II who questioned why the Royal Air Force (RAF) was performing maintenance the way it was – grounding about half the planes at a time for maintenance following a mission. His theory was that the regular maintenance (preventive or planned maintenance) was increasing breakdowns. He and a handful of other scientists recommended performing maintenance based on the condition of the equipment. And after five months of trying the new procedure, the number of available planes at any given time increased by 61 percent.
Since then, manufacturers used preventive maintenance strategies including sensors placed in devices to determine when equipment might fail. But the results weren’t consistent because the data was difficult to access. Now, with today’s IIoT, machine learning and artificial intelligence, predictive maintenance is a reality.
What is Predictive Maintenance and What are the Benefits?
Predictive maintenance is based on detecting small changes and aberrations in normal operations which usually indicates a larger problem. From digital preventive maintenance came predictive maintenance (PdM) that uses data-driven maintenance strategies to analyze operation and predict and prepare for potential failures. With 24/7 remote monitoring, data-driven insights from machine learning and innovative predictive analytics technology to alert about potential equipment failures, manufacturers can benefit in many ways. Cost savings and ROI of predictive maintenance include:
- Reduced downtime
- More targeted maintenance
- Higher productivity
- Efficient inventory management
- Enhanced data analysis
- Reduced labor and material costs
- Increased plant safety
- Optimized maintenance activities
- Increased overall equipment effectiveness (OEE)
Predictive Maintenance via Condition-Based Monitoring
Another transformative step in evolving maintenance strategies and capabilities came the advent of condition-based monitoring (CBM) which monitors key performance indicators (KPIs) to identify anomalies. Companies can check through measurements, visual equipment inspections, reviews of performance data or scheduled tests, as well as through IoT and historical data. The KPIs are gathered at certain intervals, or continuously—as is done when a machine has internal sensors. CBM can be applied to all assets.
CBM, like all predictive maintenance, also operates on the principle that maintenance should only be performed when there are signs of decreasing equipment performance or an upcoming critical failure. Compared to traditional preventive maintenance, CBM only requires equipment to be shut down for maintenance on an as-needed basis, increasing the time between maintenance repairs.
CBM can reduce machine downtime by 30 to 60 percent and increase machine life by an average of 30 percent. Predictive maintenance plays a key role in detecting and addressing machine issues before it goes into complete failure mode. According to a PWC study, predictive maintenance improves uptime by 51%. Using predictive maintenance, companies can avoid accidents and can achieve increased safety for their employees and customers.
Implementing a Successful Condition-Based Maintenance Program
FactoryTalk® Analytics™ GuardianAI™ is a new software by Rockwell Automation that provides predictive maintenance insights via continuous condition-based monitoring. The software helps maintenance engineers get the right information at the right time to optimize maintenance activities and reduce unplanned downtime.
Armed with this information, maintenance engineers have the insight to understand the current condition of the assets on the plant floor. They receive early notice as soon as an asset begins deviating from normal.
Use Your Existing Variable Frequency Drives as Sensors
When using FactoryTalk Analytics GuardianAI, there’s no need to purchase additional sensors or monitoring equipment. The software provides early warning of potential asset failures based on data that’s already available from variable frequency drives (VFDs). FactoryTalk Analytics GuardianAI software uses the VFD’s electrical signal to monitor the condition of a plant asset. When it detects a deviation in the electrical signal, it alerts the user to the anomaly so that manufacturers can investigate and plan the correct response. FactoryTalk Analytics GuardianAI provides premier integration with PowerFlex® 755, 755T and 6000T drives for key process applications like pumps, fans, and blowers.
No Data Science Required
When deploying innovative solutions in an operations environment, time to value is key. FactoryTalk Analytics GuardianAI software saves time with intuitive and streamlined workflows via a self-service, browser-based experience. Just deploy the application on an edge PC, specify your drive and asset information, and train the predictive maintenance model on live plant data with no impact to operations. When the training is complete, the software will automatically switch to monitoring mode and you can oversee the condition of your plant assets.
Starting from an overview of all assets, you can select any at risk asset to learn more about its condition. You’ll discover key information like the root cause of the deviation, how far it peaked above baseline and deviation duration. You can also include context about the severity of the failure risk and the estimated time to resolve the issue. These details support your maintenance team with the prioritization and planning required for repair.
Advance from anomaly detection to anomaly identification
FactoryTalk Analytics GuardianAI software comes out-of-the-box with embedded expertise about the most probable cause of failure for common plant asset types. If you’re monitoring a pump, fan or blower application, FactoryTalk Analytics GuardianAI understands and recognizes the electrical signature of the associated first principle faults and will provide this context when it alerts you of a deviation. By providing maintenance engineers with information about what type of failure is about to occur, you can reduce investigation time and minimize any downtime required.
The embedded expertise provides a great start for anomaly identification. But you’re not limited to the out-of-the-box functionality. You also have the flexibility to train FactoryTalk Analytics GuardianAI software on process specific faults. After you investigate and identify the source of the issue, you can label the anomaly. When the same issue occurs again, the software will recognize it and notify you.
Analyze at the edge
FactoryTalk Analytics GuardianAI software is deployed, learns and runs right at the edge for near real-time predictions.
Conclusion
Since CH Waddinton and his mission to keep RAF planes in the sky, manufacturers have been seeking to drive more efficient maintenance decision-making and derive more value from equipment. Evolving from reactive and proactive to preventive and predictive, maintenance engineers are now empowered with easy-to-use machine learning through an intuitive user experience that doesn’t require data science knowledge. Find out more at FactoryTalk Analytics GuardianAI.
Source: www.rockwellautomation.com