Asset Management and the Search for Actionable Intelligence

Mike Brooks, staff technologist for global refining at Chevron, has been tasked with planning the next generation IT infrastructure for Chevron’s worldwide operations. He’s wrestling with the technical aspects in order to engineer contextualized business value arising from combining data and information from multiple operational domains.

By Frank O Smith for Control Engineering December 1, 2007

Mike Brooks, staff technologist for global refining at Chevron, has been tasked with planning the next generation IT infrastructure for Chevron’s worldwide operations. He’s wrestling with the technical aspects in order to engineer contextualized business value arising from combining data and information from multiple operational domains. “As we move to the future, we’re looking for information richness that lies between separate functions,” he says. As you leverage more data and combine it with data from other functions, you get more contextual richness– hence business value.

Increasingly, the “other functions” that IT directors and manufacturing executives are seeking to tap are found on the plant floor. They are looking to review and improve machine asset availability and utilization to achieve strategic objectives. The challenge– and opportunity – for manufacturing engineers is not merely releasing control of data that has been part of their domain, but fully grasping its contextual significance to strategic business objectives.

Writes Bharat Nair, senior vice president and research director of global manufacturing markets for Boston-based Aberdeen Group, “Optimizing asset performance to maximize the economic value from your asset base offers the best path to increasing shareholder value.” The value of control data to the enterprise comes in contextualizing it. “Temperature, speeds and feeds, and flow without additional information doesn’t mean much to an executive,” says Jim Shaefer, product manager for SAP’s xMII toolset. The business implication needs to be made clear. “Customer satisfaction, brand management, and quality are all impacted by equipment making product,” says Maryanne Steidinger, marketing manager for Siemens SimaticIT.

Chevron’s efforts are emblematic of a global manufacturing trend. Machine and plant data is increasingly perceived as essential to enterprise performance and shareholder value. Consequently, it’s being leveraged to construct new, more dynamic and strategic key performance indicators (KPIs). Engineers who understand those equations can integrate business needs and plant floor needs in entirely new ways.

KPIs = actionable intelligence

Strategic KPIs are essentially equations. Borrowing on that, the move to elevate control data as integral to business can be represented by a conceptual equation of its own: abstraction + integration + aggregation + normalization + contextualization = actionable intelligence. That’s certainly a mouthful, not to mention a highly complex technical challenge. But it’s useful for grasping how control data drives business decisions. This equation breaks down thus:

  • Abstraction = based on a model for describing physical plant equipment and processes

  • Integration = interoperability facilitated most efficiently by standards

  • Aggregation = data elements summarized or joined in some fashion

  • Normalization = ensuring disparate elements of a type (e.g., compressor) and the data extracted from them have common– hence comparable – meaning or reference

  • Contextualization = putting aggregated, normalized data in business/operational context (e.g., line availability = throughput = profit margin contribution)

  • Actionable intelligence = information for making decisions.

Each is hugely important and represents thousands of hours of effort on the part of technology vendors, standards groups, and (too often) individual company project teams to take elements like temperature and pressure and convert them into strategic KPIs.

The vital KPIs Aberdeen used to determine best-in-class enterprises in its September report, “Ground Up Strategies for Asset Performance Management,” included: OEE, percent on-time delivery, throughput, and asset effectiveness. Though asset performance management (APM) is often narrowly viewed as maintenance-centric, “that’s a very short-sighted perspective,” Nair advises. The set of Aberdeen’s critical KPIs are equations, such as OEE = availability x quality x performance; with each individual element distilled into further equations like performance = ideal cycle time÷ the net of operating time ÷ total production.

There are countless permutations with significant business value, including, for example, highly strategic ones like: order delivery available-to-promise based on asset and resource availability; asset margin contribution; and lost profit opportunity.

It is into this maze of IT infrastructure and KPI development that data from the control layer is destined. As a consequence, it is imperative that control engineers begin to regard the individual data points they monitor as information with significant value in the business process calculations that drive other people’s work, as well as their own.

Control engineers as business performance managers

Dr. Peter G. Martin, Invensys vice president of strategic ventures, argues persuasively that the root cause of lost profit opportunity in the plant is not all the islands of automation and information that exist, but rather the islands of organization . This circumstance sprouted from the increasing task/function specialization, seeded with the birth of the 19th-century factory model, and has only grown more extreme. Rigid emphasis on function has created a dysfunctional organizational family. What is required, Martin asserts, is that all employees view themselves as “business performance managers.”

While control data needs to be aggregated and contextualized for top-level executives, control engineers, production supervisors, line and equipment operators, maintenance personnel– in essence, everybody – needs to be provided with aggregated, business-contextualized data that enables them to better perform their jobs. Martin illustrates his presentation on the subject with a cartoon of a housewife turning down a thermostat, telling her husband, “I’m turningdown the heat to $325 a month.”

What exists to get us there? Quite a lot, actually. And the arsenal is growing monthly. Product offerings range from best-of-breed solutions in such areas as enterprise data historians and enterprise manufacturing intelligence (EMI) products, such as OSIsoft PI Historian, and Incuity EMI. They include intelligent middleware platforms and toolsets like Microsoft’s BizTalk and SharePoint portal products, and SAP’s NetWeaver and xMII products. Additionally, industrial-strength, integrated application frameworks such as Wonderware’s System Platform 3 and InTouch 10, announced in September; and Invensys’ Infusion platform announced last year. Other automation vendors are also ripe with new products and announcements, including Rockwell Software’s new FactoryTalk Historian; GE Fanuc’s pending new release of its Proficy product; and the Hyper-Historian from Iconics.

On the standards front, a new consortium of standards bodies, including ISA, MIMOSA, OPC, OAGi and the WBF are collaborating as Open O&M. The goal: to harmonize their individual process models and create a composite model for asset performance improvement that brings operations and maintenance (O&M) functions into closer unity via shared, aggregated and contextualized data. Though maintenance is only a piece of broadly-defined APM, its importance to strategic enterprise performance is seen in recent acquisitions by major players, with IBM adding Maximo and Infor adding the Datastream asset/maintenance capabilities that add touch equipment/controls into the breadth of their portfolio offerings.

Connectiv Energy is a merchant generating power company– meaning it both deals in the energy commodities market and generates electricity. Headquartered in Delaware, it sought to apply the capabilities of its existing OSIsoft PI historian product for storing equipment operations data, and RLINK, its historian-to-enterprise system connection, to leverage control data to enable condition-based monitoring/maintenance (CBM). CBM leverages operation controls data for improved maintenance – an emerging key initiative for many companies.

“PI and RLINK were key,” says George Muller, IT business manager. “We do equipment efficiency calculations in PI, and if it drops below a certain level, that triggers a message through RLINK to SAP to generate a maintenance work order.”

A major pharmaceutical firm implemented Incuity’s EMI solution and gained unanticipated benefits from the plant modeling, data extraction, contextualization, and query capabilities in the best-of-breed enterprise manufacturing intelligence product. “One of the essential elements they were calculating within Incuity EMI was the electronic power consumption for each batch run,” says Doug Lawson, president of Incuity. At a corporate level, electricity consumption was allocated as overhead to each department by a pre-defined factor.

Using Incuity to contextualize usage at the equipment level, “in an afternoon we helped engineer actual consumption calculations and pass them up to SAP,” Lawson says. While the plant had been struggling to optimize a quarter percent usage here and there, what they found was a massive error in the original corporate allocation model that was out of registry by 15 percent. “Incuity helps transform equipment into assets rich in information. Plants need to demand that their data be seen as having the same value as enterprise data,” continued Lawson.

“The biggest challenge companies are facing today is not lack of information, but knowing what information has value,” says Houghton Leroy, research director of Waltham, MA-based ARC Advisory Group. Good analytic tools based on historian repository data and EMI tools can aid that.

Encompassing breadth

Microsoft’s BizTalk and SharePoint products, and SAP’s NetWeaver and xMII are being applied to deliver greater breadth via aggregated, contextualized plant floor data, joined with information from widely disparate systems. New industrial application platforms from Wonderware and Invensys typify plant-centric system vendors pushing the envelope in comprehensive breadth for creating enterprise-class plant systems.

Microsoft’s BizTalk is a system collaboration tool for connecting disparate systems and enabling new enterprise applications; and SharePoint is a web-services portal enabling personell collaboration. “To get data from the controls layer, BizTalk taps into Microsoft’s partners solutions, like ActivePlant, AspenTech, and Iconics, which they collect and expose in Excel,” says Jon Brusseau, Microsoft industry technology strategist for automotive. “If a supervisor needs to be alerted to a control layer issue, BizTalk can take data from a partner’s system and send a message to a line supervisor or plant manager with a dashboard on his desk, or by email or pager.” Plant data can also be made available to multiple decision makers in SharePoint’s common workspace via web browser.

“SAP’s xMII product is a set of application tools to aggregate and develop business context. Connectivity is part of the solution. It connects the ‘last mile’ to the plant,” says Schaefer of SAP. “It’s good at extracting, transforming, aggregating and messaging so information can be consumed by the enterprise. It spans the whole spectrum, and supports drilling down into the details. It’s also a rapid application development tool for creating composite applications that sit atop xMII,” he says.

Invensys is working with SAP to create an integrated, real-time, plant process accounting/financials composite application that draws on the combined strength, flexibility and utility of Invensys’ Infusion platform and SAP’s NetWeaver/xMII capabilities. It is one of the best examples of the conceptual model: abstraction + integration + aggregation + normalization + contextualization = actionable intelligence. It also strikes at the heart of giving every plant employee the contextualized information needed to transform them into business performance managers. It does this by putting tasks and processes into financial terms– be it changing a value setting, running a piece of equipment longer or taking it off line for maintenance.

Part of the power of Invensys-class solutions is that the model for abstracting the plant and driving how the system works is conceived with a top-down, enterprise business mindset, but draws data and makes information available from the bottom up. A services oriented architecture (SOA) structure empowers the designing and building of small discrete applications services (functions) that can be easily coupled with other services to create larger applications. These can also be reused repeatedly to create other, new applications as real time data-supported analytics expose new areas of opportunity.

“Many executives look down into their plants and see little value in physical equipment and control systems, wondering why should I invest more money there– why not ship production overseas,” says Alison Smith of AMR. “They tend to view assets as sunk costs – not a competitive advantage.” They need to start looking at assets in the purest sense, she asserts, “not as a liability, but as the means to make products faster, better, cheaper.”

Control engineers are in a prime position to enable this. “They need to start seeing what they do in terms of return on assets. Start by dollarizing the data in the domain,” Smith encourages. “Get the information out there. Expose it to decisions makers.”

In essence, they need to become business performance managers. Make the controls performance the new strategic, enterprise KPI.

Definition of Maturity Class Mean Class Performance
Source: AberdeenGroup and Control Engineering
Best-in-Class: • 92% OEE
Top 20% of aggregate performance scorers • 97% On Time Delivery
• 96% Throughput
Industry Average: • 80% OEE
Middle 50% of aggregate performance scorers • 93% On Time Delivery
• 88% Throughput
Laggard: • 61% OEE
Bottom 30% of aggregate performance scorers • 76% On Time Delivery
• 73% Throughput

Online Extras

Links to Companies and Resources

Aberdeen Group


ARC Advisory Group



Connectiv Energy

GE Fanuc Proficy iFix, Asset Management

IBM Maximo


Incuity EMI


Infor Datastream


Microsoft BizTalk, SharePoint

OSIsoft PI Historian

Rockwell Software FactoryTalk Historian

SAP xMII toolset

Siemens SimaticIT

Wonderware System Platform 3, InTouch 10

Author Information
Frank O Smith is a contributing writer for Control Engineering.