Want more Insights from SAS? Subscribe to our Insights newsletter. Or check back often to get more insights on the topics you care about, including analytics, big data, data management, marketing, and risk & fraud.
Why you need asset performance analytics
Better manage your industrial assets to eliminate downtime, shorten repair time and improve the overall health of assets
From manufacturing diapers to extracting oil from the ocean floor, the reliability of the industrial assets and machines in many processes is critical to success. Preventive maintenance is a long-standing tradition in many industries, especially manufacturing and energy production. Today, asset performance analytics is becoming more important as a way to provide a holistic, predictive approach to asset maintenance and performance. It is broadly aimed at optimizing the value of production assets by analyzing asset data to predict future failures and prevent downtime.
“Asset analytics supplies the leading indicators that tell an enterprise when an asset is likely to fail in the future, and on what schedule that asset should be evaluated for maintenance or replacement,” explains Alyssa Farrell, Energy Industry Marketing Manager at SAS. “Comprehensive asset analytics also takes into account the entire performance and lifecycle cost of the assets, how they are networked and their value to the company.”
In both the energy and manufacturing industries, revenues and outcomes rely heavily on the performance of machines, so “if that machine, that piece of equipment or that asset fails, the enterprise is in big trouble,” says Mike Hitmar, Manufacturing Industry Marketing Manager at SAS.
According to Hitmar, asset performance analytics can be used in three major ways to improve preventive maintenance:
- Time-based preventive actions can be taken based on historical data. The ideal frequency for replacing air filters, for example, can be determined using analytics.
- Applying analytics can help reveal how an asset is operating and whether it is degrading in performance. Often, small internal signs that might go unnoticed by workers will be detected by analytics before a failure occurs.
- Asset optimization uses analytics to make sure the asset is operating at peak performance at all times.
Getting started with asset performance analytics
How do you know if asset performance analytics is right for you? And what is a good way to start? Begin by asking these questions:
- Are your overall maintenance costs high compared to your industry peers?
- Are you facing equipment performance challenges that result in poor quality or reliability, missed shipments or factory overtime?
- Are you meeting your SLAs? What about your service contracts?
- Are there opportunities to improve profitability by streamlining operations?
If you answered yes to any of these questions, you should consider asset performance analytics. Farrell suggests starting with an easy win. “I would start with the basic approach of finding a high-profile project that can produce a meaningful, quick, positive outcome using asset analytics. Then rotating that champion who drove the change around to other assets or production centers.”
Asset analytics supplies the leading indicators that tell an enterprise when an asset is likely to fail in the future, and on what schedule that asset should be evaluated for maintenance or replacement.
Energy Industry Marketing Manager
Asset performance analytics and the Internet of Things
Technology is no longer the barrier to gaining insights and intelligence about your assets. High-performing analytics and streaming data from sensors on your assets can help provide answers faster than ever before. The number of vendors and sensor types has exploded recently.
Using sensor data presents many opportunities but also presents some data management challenges. “Inevitably, there will be data quality questions to solve before sensor data can be fully analyzed,” says Farrell. “And I think it’s a big opportunity to improve the management of sensor data so that energy and manufacturing companies can focus on the analytics and reduce the effort that it takes to manage and aggregate and clean up the sensor data.”
Hitmar emphasizes the importance of data preparation for analytics: “We’re really focusing on data quality from an analytics perspective, including the way the data is indexed and organized to allow analytics to be run on it. An example would be how different sensors handle time stamping. Production machines are not necessarily connected to each other, and sensors are not set to record in sync. Sensors might be set up to measure temperature or vibration, but with different time intervals. Analytics wants the data in a nice time series of events, for everything to be lined up in rows and columns, which isn’t always possible with multiple sensor sources.”
Real-world examples of asset performance analytics
More and more often, the uses of asset performance analytics will involve streaming analytics on sensor data. For example, Hitmar describes a diaper manufacturer that was able to reduce rejects, scrap and rework. For a factory that is running all day, every day, a 20-minute improvement in reaction times made a huge difference. These results were made possible by collecting and analyzing more than 5,000 attributes, including streaming data from sensors throughout the factory.
Likewise, Farrell tells us about a field of oil wells that sends streaming data from the ocean floor. “The recent volatility in oil prices has caused an uptick in interest in asset analytics because producers are keenly focused on which assets are most efficient under which conditions.”
For more examples and further discussion about asset performance analytics from Farrell and Hitmar, read the paper Asset Performance Analytics: Optimizing Performance in Asset-Intensive Industries from the International Institute of Analytics.