Machines Have Their Say

Nov. 3, 2009
Machines Are Getting Better at Telling Us What's Wrong with Them. More and Better Sensors and Sophisticated Algorithms Make Predictive Maintenance Practical—and Help Close the Loop Between the Plant Floor, EAM Systems and ERP

This article was printed in CONTROL's November 2009 edition.

By Nancy Bartels

Down on the plant floor is where the money is made—or lost. That's where the product is made, the quality is determined, power and raw materials are consumed, and the labor is expended. How the people and machines perform there will make or break a company. Thankfully, people have always had ways to get their say. Machines, not so much.

But paying attention to the machines is more important than ever. The logic is inescapable. "The cost of lost production is 10 times more than maintenance costs," says Bart Winters, senior manufacturing marketing manager for Honeywell (

On the other hand, when technicians spend time and money performing tasks or replacing parts before it's necessary "just in case," that's money that could have gone to the bottom line—a number that seems to grow more important every day.

Chris Rogers, Western Canada regional manager for reliability-centered maintenance software vendor Matrikon ( nails it. He says, "We're hearing from every single process customer that the competitive issues around maintenance are there, and we have to improve how we're getting return from these assets."

Fortunately, over time, the technology to improve that return has become more sophisticated and, at the same time, more available. For more than a decade now, computerized machine maintenance systems (CMMS) and enterprise asset management (EAM) applications have been creating work orders and tracking parts inventories and labor. Dashboards give us more tightly focused looks how the production numbers are adding up. Connectivity to the ERP system is much easier than it was—especially now that major ERP players such as SAP ( and Oracle ( have their own EAM modules.

Wireless sensor technology that enables the acquisition of more data from more assets and more and better analytical algorithms to squeeze meaning out of all that data make the idea of remote monitoring, well, less remote.

Finally the most critical link in the EAM chain—the asset itself—is finding its voice. Whether you call it condition-based monitoring, reliability-centered maintenance, plant asset management, asset life-cycle management or plain old maintenance, listening to critical capital assets to find out how they're feeling and what they need is where the excitement around EAM is now.

Power Generation Gets It

The power industry is at the leading edge of this trend in part because of the nature of its business. Honeywell's Winters explains: "They have no tanks. They can't store anything. Even in a continuous process, you have buffered storage." Not to mention that when a power plant goes down, it's not just that facility that's dead in the water, but every home or business that depends on the electricity it generates.

SmartSignal's Availability and Performance Center provides services, including deployment, model maintenance and monitoring for asset management.

That's why companies such as Baltimore-based Constellation Energy, a power generation company with a portfolio that includes, nuclear, coal, natural gas, oil and renewable and alternative fuels, are taking a hard look at the latest technology for keeping their systems running.

Frank Andracchi, managing director in Constellation's Power Generation (CPG) division took a page from the airline industry's maintenance playbook, of all places. "I'm involved in merger and acquisition activity. I see what the competition has, and I've noticed the performance centers they have, which led me to believe in their value. Rolls Royce monitors every jet engine in flight. Maintenance of those engines is done based on their performance while operating. They have truly reached condition-based monitoring."

Figuring that if a jet-engine company could do it, so could Constellation, Andracchi became a champion of building a performance maintenance and diagnostic (PM&D) program, one power plant at a time. To help get started, Andracchi brought in outside fire power—MRG Solutions (, an EAM and asset management services company, and predictive analytics specialist SmartSignal (

Constellation began with its Maryland coal-powered fleet. "Coal plants are old, and they're only as good as their assets," explains David Lillefloren, director, CPG business performance. "You want to make sure you're taking care of them. MRG is helping us with that. We look at a turbine or a boiler—everything about it. We look at predictive maintenance (PM) and cost of maintenance. What can we improve on to extend the life of the equipment? What critical spares do we need to keep this going? Are we doing the right amount of PM?"  

SmartSignal's task was to provide software and tools for early failure warning. It monitors two out of three of Constellation's coal plants now and expects to bring a third online next year, with more in the future.

Right now, SmartSignal is handling the asset monitoring, but the goal is to bring that expertise in-house. "We have a centralized, virtual PM&D center," says Lillefloren. "We have several individuals whose sole task is to monitor SmartSignal and other software. The control operators can see what we see. When they see an alert, they forward information to the individual plant and alert them. We have reliability engineers at each plant to focus on the asset health care, who are working with people in the plant, looking at each system and making assessments."

But MRG Solutions and SmartSignal are providing only part of the puzzle. Constellation has a MAXIMO ( system to handle the traditional functions of work order management. It also uses MAXIMO to keep track of critical spares, costs and labor for budgeting purposes. 

Constellation's foray into cutting-edge asset management is a work in progress—not all of Constellation's plants are connected to its EAM and ERP systems—but it's already starting to pay off.

"We convinced upper management that this is the way we should be controlling costs going forward. We used to automatically overhaul generators and turbines every five years. Now, we do some checks yearly, but we wait for testing and diagnostics to tell us it's time," says Andracchi. "We had to get $400,000 worth of saves this year. We've already met that goal."

A Bit Farther Down the Road

Houston-based RRI Energy Inc. is a bit farther down the road than Constellation with its predictive maintenance program. It has been using SmartSignal since 2005, and has deployed over 13,000 MW of generation that are monitored by its SmartSignal system. In RRI’s case, SmartSignal receives 30,000 data points from OSIsoft’s PI systems ( and analyzes the data every five minutes.

The SmartSignal system processes information for 67 coal and natural-gas units. "These are a combination of the most critical assets and the ones that have proper instrumentation to provide information for the model," says David Thomason, manager of IT/plant applications at RRI Energy. "The system monitors both processes and equipment—groupings that make sense to monitor together." 

Using the SmartSignal/OSIsoft combo, RRI is establishing trending data around machine performance. "We can tell when the rate of change is different and approaching a critical threshold. It gives us earlier predictive analysis and can catch real-time or near-real-time issues," says Thomason.

RRI is planning to use Smart Signal's xConnector to take data from the model and push it back into the PI screens. "We will be able to look at a bearing temperature in real time and where its expected value should be,  based on the statistical model," explains Thomason.  

Much of this complex analysis goes on behind the scenes at RRI. "The process analysis team and SmartSignal jointly create the models. The users don't see them; they only see the results. We have an email system that people can subscribe to, and SmartSignal sends an alert or puts an entry into the WatchList. That prompts an email outlining what the alert is about."

Like Constellation, RRI does not have a direct feed to its EAM or ERP system. However, it uses the EAM module in its SAP enterprise system ( That takes care of work orders, materials management, scheduling and resource planning.

Again, like Constellation, RRI's journey to predictive maintenance is very much in progress. The next step, says Thomason is a one-plant pilot of a Microsoft SharePoint ( -based dashboard. It will use a hierarchical database structured on same basis as it is in SAP.

RRI's Seward Generating Station is a coal-fired 525-megawatt plant located 80 miles east of Pittsburgh,  Pa."The content of the portal will be based on navigating through the equipment hierarchy. When a user selects the pump, for example, a query will be run against the EAM system and will retrieve completed work orders, planned work and notifications relative to that piece of equipment. At the same time, it will bring up associated SmartSignal information and PI screens, explains Thomason. "It puts all the relevant information and data  associated with that piece of equipment together. We will then have a  more dynamic way to maintain and monitor critical equipment condition and their assessments. This will also be helpful in outage planning."

And that critical bottom line that is the driver for much of this implementation? Thomason estimates that the predictive and proactive systems saved RRI $7.4 million last year—"a very positive ROI."

Allentown, Pa.-based PPL Generation has launched itself on the same kind of condition-based asset monitoring journey. It is doing a pilot project with Siemens Energy, Instrumentation, Controls & Electrical ( using Siemens new SPPA-D3000 Plant Monitor at its Brunner Island, Pa., coal-fired facility. There it is testing the D3000 by feeding it OSIsoft PI data on previously known faults, says Frank Lyter, the manager of PPL's Equipment Reliability & Performance Center. So far, it's passed with flying colors, catching the faults based on the PI information alone.

According to Lyter, the Siemens installation is just a part of a suite of systems from multiple vendors that listens to assets, analyzes what it finds and gets the maintenance process organized and on track. PPL is also using Matrikon's Operational Insight optimization software, SmartSignal monitoring software, IKS Plant IQ ( asset condition tracking solution, Ventyx eSOMS suite for operator rounds scheduling ( and DocKit equipment assessment software from ATI ( and OSIsoft's PI, among others.

This laundry list of applications targets one of the biggest challenges of doing this kind of asset management. As Lyter explains, "You really need a suite. One app won't solve all your issues. We have tried to get more stuff in one package because we don't like having three dozen apps, but no one company has the whole thing."

But sorting through the complexity of multiple systems has been worth it for Lyter. It's cut down on the number of people having to go out to facilities. "A significant amount of information is available via computer wherever the people are. They can troubleshoot right from home if they need to. No time is lost getting to the facility."


Power generation is not the only process industry working on ways to listen to its assets more carefully. Simmons Foods is a privately held company that sells processed chicken to restaurants and mass merchandisers. It has 20 different plants spread all over Arkansas, Oklahoma and Missouri.

Simmons uses the EAM module of its Oracle ( ERP system. "They're all on the same instance of Oracle," explains Bill Henderson, application developer, EAM/MRO inventory at Simmons. "Each maintenance department with a separate inventory is a separate ‘organization' within the EAM system. The data is visible either as a whole or individually. Corporate has visibility into all of it."

Like the other users we talked with, Simmons' journey to asset monitoring is a work in progress. The system has been fully implemented at four plants, with the others using various pieces of it.

Simmons is also working with Oracle to develop a prototype project using condition-based monitoring. At one of its plants, there are seven refrigerator compressors, each equipped with Opto22 ( sensors. The sensor information is sent directly to the Oracle system, and work orders are based on those reports. "We used to have weekly PM. Now it's based on run time because all those compressors don't run all the time," says Henderson. 

Simmons' decision to use the Oracle EAM module was based on the fact that it has implemented Lean practices. "We need to eliminate waste wherever we find it. Earlier we had a mixed culture. Some of our companies had paper; some had EAM; some were run-to-failure. Through Lean and EAM, we're starting to emphasize preventative maintenance," he adds.

Operations such as Constellation, RRI, PPL and Simmons are on the leading edge of this movement. It's still a tough sell, and it's still complicated. Matrikon's Chris Rogers says, "There is capability to do this top floor to shop floor integration. The real issue is that nobody has the focus or understanding to do it all by one company. You really have to fundamentally rethink how you're going to do your business. But outside forces will push the change. Innovators are honestly looking at what we can do differently. As soon as the masses start to see that there are real dollars to be made here, they will change."

Let's be clear. Advances in the asset-monitoring space have not brought about world peace, universal harmony and/or free beer for all. And there is still no such thing as a one-size-fits-all solution. Getting your critical assets to talk to you and then using the data they give you to drive production improvements or reduced downtime is no trivial pursuit.

But it is getting easier.

Nancy Bartels is Control's managing editor.

Familiar Names Get with the Program

Meanwhile, folks whose eyes have always been clearly focused on the plant—the process automation system vendors—are using their expertise to develop applications to gather data and monitor asset health. Some are also offering their software and expertise as a service to customers who want to outsource the complexities of managing their asset data.

At September's 2009 Emerson Exchange, Emerson (  unveiled its Syncade Operations Management Suite. Billed as a replacement for older manufacturing execution systems, Syncade connects to the PlantWeb architecture and integrates real-time plant-floor data with procedural, off-line and transactional plant business processes and asset management.

It also announced a partnership with asset performance management vendor Meridium ( Meridium will provide advanced analytics and decision-support technology to Emerson's AMS Suite: Asset Portal 4.0 powered by Meridium. Asset Portal 4.0 will offer a real-time connection to other AMS Suite applications to enable analysis of diagnostic alerts and other asset health information. It also offers integration with SAP's Plant Maintenance module and IBM's MAXIMO.

ABB ( is offering its asset management system on the software-as-a-service (SaaS) model. In this model, which the company describes as "more than SaaS," ABB installs the technology, does the system integration, maintains the solution and has remote access to the system to deliver the product. It provides the domain expertise and relieves companies of much of the responsibility for maintaining the systems. The company is also doing proof of concept on using Web 2.0 for reporting system. In the trials, an agent that does condition-based monitoring and is remotely connected to the ABB facility is placed on a specific device. When some threshold has been crossed, the agent issues a message through a secure pipe to a Web site. That site maintains a blog that reports on the device. Users can access it and communicate information via text, email, even Twitter. The device can be on a user's buddy list. Group instant messages allow end users to communicate with the domain experts to solve problems.

Invensys Operations Management ( has just released Avantis PRO 5.0, the latest iteration of its EAM solution. Based on Microsoft.NET architecture, it enables real-time monitoring of the maintenance health status of the entire organization. Users can see at a glance what areas of the plant need attention.

Honeywell ( has also announced a partnership, this one with Australia-based Synengco Pty Ltd. Under the partnership, Honeywell will license and sell the SentientSystem suite of applications for optimizing operations and asset management for power utility customers. The suite integrates with Honeywell's Experion PKS and other distributed control systems to improve operations efficiency and asset management. The asset management applications provide process monitoring, early warning systems, creep and fatigue monitoring and root cause analysis.