The Future of Asset Performance Management

Combining the power of on-premise software with cloud-based data storage and analytics, AVEVA’s hybrid approach to asset management visualization is poised to change industrial maintenance and connect it more closely to core business functions.

Jim Chappell
Jim Chappell

From a business perspective, asset management has three aspects, said Jim Chappell, vice president of information solutions at AVEVA. Those aspects are: strategize, analyze and maintain. “From the design of an asset to the end of its life, asset information should be constantly analyzed for the operation to be in a continuous optimization mode,” he said.

At the AVEVA World Conference event in Dallas, Chappell explained how AVEVA is advancing its approach to Asset Performance Management (APM) through the combination of its Avantis, PRiSM, Smart Glance and Insight visualization technologies with continuous business optimization in mind. “At the heart of your business, asset management gives you insights into risk. By connecting your asset strategy with corporate objectives, you can better balance financial, operational and safety risks,” he said.

This process of connecting assets to corporate objectives begins with SCADA and sensors and moves into Industrial Big Data to transform the data into information by applying analytics and, ultimately, performing intelligent maintenance to drive business improvement.

Big Data

Chappell explained that AVEVA’s Industrial Big Data is based on the company’s Process Historian and Enterprise Data Management solutions, which provide the required data management capabilities—ranging from time slicing the data to hosting it on the cloud. From this basis, AVEVA’s Insight is used as the industrial platform to leverage these data for OEE, business intelligence, condition-based analytics, predictive analytics, asset management, collaboration, mobile maintenance, and the development of asset models and hierarchies.

Using Insight and Insight Mobile to access all this data in a user-friendly way is key to delivering data in a hybrid cloud and on-premise architecture. With Insight, users can see when alarms occur, how they were addressed, and view and add comments. “Our whole mentality around this is that getting this information should be intuitive and easy,” Chappell said.

AVEVA is also adding mapping to Insight so users can drill down even further to see individual trends, as well as additional details around alarms and events. “It [Insight] doesn’t replace AVEVA’s Alarm Manager,” Chappell said, “but it does give you high-level insights to diagnose and determine the root causes of problems. Also, condition-based alerts can be configured to send notices or trigger work orders when things happen per your established parameters.”

Analytics

Since reducing risk is such a big driver of maintenance strategies, Chappell noted that users need to be alerted of possible issues before they become problems. They also need help in decision-making to better pinpoint root causes. These user requirements are why AVEVA incorporates business intelligence capabilities into its condition-based analytics. The “multi-dimensional analysis [provided by business intelligence] allows users to see specific aspects of energy use, batches, shifts, equipment states and costs,” he said, adding that AVEVA is moving toward bringing even more intelligent analytics to Insight.

Artificial intelligence (AI) is one of these technologies in which AVEVA is investing. It will be rolled out to users via machine learning in Insight’s News Feed function. The News Feed uses machine leaning much the same way Google does with site ranking, Chappell said. “You’ll see what’s relevant to you based on your log in. We figure out what trends will be relevant to you based on what’s on your dashboard at the top of your News Feed,” he added. “With this functionality, we’re headed toward team-based collaboration, in which others on your team will be alerted based on what you learn.”

Whereas machine learning is a dynamic application of automated intelligence, AVEVA’s predictive maintenance functionality is more model based. “It takes historical data and creates a digital signature of the asset to create acceptable bands of operation,” Chappell said. “All new data [that comes into the system] will be matched to these bands. If the data diverges, these differences are identified and displayed to tell you if you’re moving into an anomaly situation—and you’ll see this well before it becomes necessary to take maintenance actions.”

On top of this model-based analysis, AVEVA will be adding prognostics via neural net and deep learning technologies. Chappell said the insights delivered by prognostics will help users determine how fast a situation will worsen, how to do what-if analyses, and tell them if repairs can wait until the next planned maintenance outage.

Maintenance

AVEVA’s technology approach to the maintenance lifecycle includes its Avantis Enterprise Asset Management (EAM) solution, condition management and EAM extensions—which will be rolled into AVEVA EAM Insight incrementally in 2019. “We’re moving toward a lightweight cloud version of EAM that will link to on-premise software,” Chappell said, “which will produce one view for the user. With this one view, we’ll avoid creating an information junkyard that requires users to navigate through a bunch of tabs to see what’s relevant.”

With all the data gathered in this way, AVEVA is developing a common intelligent model focused on assets and/or entities in the plant and all the information around them, such as sensor data, alarms, video, location, OEE, etc. “This is a fundamental shift away from a tag centric system to an asset centric system,” Chappell said. “This enables information to be visible on a mobile device to give you the information you need in the context you need it.”

Providing an example of how this technology is already being used in plants today to transform big data into case management, Chappell described a valving issue that affected an AVEVA customer. Insight collected the bearing sensor data in the cloud, at which point the system determined that the bearing temperature was too high for conditions. This led to an alert being sent to a technician along with a work request. The technician discovered that the oil reservoir contained half oil and half water, which was putting too much pressure on the seal and causing water to leak into the bearings. Insight showed the technician that, under those conditions, the bearing would likely to fail within seven days, so an emergency work order was issued. The system then recommended procedures on how to fix the problem.

Companies in this article
More in Home