Deployment of standards-based, digital fieldbus networks has been steadily replacing combinations of proprietary local- and remote-I/O networks over the past decade, ushering in the plant floor information age, where flattened, streamlined organizations, offer the potential for everyone in the corporation to better understand the details of what does and doesn't make things work.

One of the downsides of the plant floor information age is that each time an "intelligent" instrument is connected, it provides the one or two values really wanted plus a hundred more values no one is quite sure are needed.

The result: businesses are drowning in data and thirsting for information.

We know how to link 50 devices. We can even link 50 networks of 50 devices each. The consequences of continuing to do that, however, have been ignored. It's simply not wise to continue to save more and more plant floor data without first establishing an information architecture designed to efficiently manage and make use of the mountains of data today's plant floor is capable of producing.

Performance measures (PMs) have long been used to indicate process/production performance. However, for PMs to positively influence a company's bottom-line, PMs must be dynamic and reported in time for operators to influence process/production cycles, thus the information architecture must meet the needs of dynamic process measures (DPMs).

Frequently, DPMs are given titles such as "yield" and "efficiency" and are composite values developed using operating, costing, and energy related values.

Being able to consistently apply DPMs to improve a company's bottom line requires an information architecture that:

• Supports data scalability and abstraction;
• Efficiently addresses the challenges of legacy systems; and
• Capitalizes on available standards.
• (See "CPAS information architecture" diagram.)

Lynn Linse, an engineer at Digi International, believes significant improvements in how a business manages and uses its data are feasible when "true scalability" and "abstraction" are provided.

Linse says, "True scalability avoids having a system with 25,000 data values be 10,000 times more complex than a system with 25 data values. True scalability helps users manage data complexities by allowing zoom-in and zoom-out."

"Data abstraction is also important," says Linse; "It's what allows the visualization of a production system as a large aggregate of many smaller, layered subsystems. For example, being able to easily integrate PLC-based controls provided by a skid-mounted equipment supplier into the main process automation system."

Object (function block) architectures, such as those used by FOUNDATION fieldbus, most process automated system (PAS) manufacturers, and ODVA (Open DeviceNet Vendor Association), eliminate the need to manage individual data values and provide the ability to zoom-in/zoom-out. However, thousands of objects of any type can easily prove difficult to manage and still not address the skid-mounted controls integration example.

A potential solution for addressing data scalability and abstraction is emerging in IEC 61499-Function Blocks.

Under the guidance of IEC (International Electrotechnical Commission) technical committee 56, Rockwell Automation's James Christensen is leading working group six to develop IEC 61499 as a wire-independent protocol designed to address the challenges of managing plant floor control systems including the data each produces. (See "IEC 61499 goals" sidebar.)

Christensen says, "IEC 61499 will first be adopted by the discrete industries as a way of integrating, configuring, and dynamically reconfiguring machines and systems built of intelligent, distributed mechatronic components. It's economics and open-architecture design will help it penetrate the process industries, but at a slower rate."

Several major suppliers are supporting the IEC 61499 development effort and Christensen reports its concepts have been tested in a European academic environment, but has yet to be applied to a real project.

In the mean time, users should continue to push PAS manufacturers for more robust third-party control system integration solutions.

Throughout the world, but especially in North America, large numbers of older (legacy) analog and digital instruments and control systems remain in-service with non-existent or loosely defined replacement plans.

Larry Komarek, Phoenix Contact automation manager, says, "Among the problems in managing plant floor data is that despite the availability of open system standards, control solution architectures remain essentially unchanged. Information from intelligent field devices, including transmitters, drives, motion control, vision inspection cameras, and micro PLCs is concentrated in a user-created database in the controlling PLC. From there it's sent to cell-supervising computers, where another database is formed. Data are further transferred to additional enterprise databases—MES, LIMS, etc.—with each new database introducing opportunities for configuration and communication errors."

What Linse says about true scalability and abstraction goes hand-in-hand with Komarek's statement about database architectures: The answer doesn't lie in storing more data. What's needed is an information architecture that accommodates legacy instruments and systems.

John Ditter, Wago product specialist says, "Legacy systems designed, built, and installed prior to 'open' system standards present unique challenges when designing ways and means to share plant floor data throughout an enterprise. Often legacy systems use proprietary operating systems, networks, and protocols; have no remaining processor time; and/or have no room to add more memory. For owners of these systems, the options are limited. One solution is to add a data concentrator or gateway."

Several manufacturers, such as Lighthammer, Phoenix Contact, and Wago, offer devices designed to connect to legacy systems and provide data in a variety of standard Internet protocols. Though each manufacturer's product offers unique features, each utilizes industrial Ethernet as its physical layer.

Lighthammer's Illuminator Manufacturing Intelligence Server uses standard Web browser technology to connect to the data sources and data types found on the plant-floor. Illuminator extracts, aggregates, and transforms data into plant-related information, such as asset utilization, production efficiencies and yields, inventory levels, order status, and quality and traceability data.

Phoenix Contact offers a range of serial data converters/isolators and Ethernet converters designed to extend, isolate, and protect industrial Ethernet and serial data networks. DIN-rail mounted, Phoenix Contact's interface devices are suitable for industrial plant floors including hazardous area (Class 1, Div 2) locations.

Wago's 750-841/842 Ethernet programmable fieldbus controller (PFC) provides a means of linking legacy plant floor data with enterprise systems. Each PFC serves as a smart gateway; connects serially; and uses Modbus/TCP, Ethernet I/P, or a custom Wago Ethernet-based protocol.

Obtaining and integrating data contained in legacy systems for inclusion into an information architecture doesn't require replacing the system. It may be as simple as adding a data concentrator or gateway device.

As previously noted, plant-floors typically use multiple control systems from a variety of manufacturers, further complicating integration issues. Also, the greater variety of control systems deployed, the greater the implementation and maintenance expenses.

The family of ANSI/ISA S95 specifications provides a standard terminology and a consistent set of concepts and models for integrating manufacturing operational functions with enterprise functions.

For users, S95 specifications provide a means of identifying needs, reducing costs of automating manufacturing processes, reducing lifecycle engineering efforts, and enabling vendors to supply appropriate integration tools.

ARC Advisory Group VP of emerging technologies, Robert Mick, says, "Users new to ISA S95 are surprised by the lack of something to build. Rather than being a specification for defining messages and/or programming interfaces, S95 is an alignment process, where product manufacturers incorporate the S95 terminology and select models as part of their products, architectures, and activities. The specification includes several models and structures, and no supplier is likely to use them all." (See "S95 supplier alignment matrix" table.)

Though the standard does not define a formalized protocol for information exchange, several ISA S95 committee members recognized that the creation of XML schemas could play an important role in making the standard "user friendly."

Under the stewardship of the World Batch Forum (WBF), a working group developed the "Business to Manufacturing Markup Language" (B2MML) schemas. Released as a royalty-free implementation of the ISA S95 specification, WBF's B2MML provides a set of XML schemas developed using the XML schema definition language.

Adding credibility to the ISA S95 specification and WBF B2MML schemas is recent indication that SAP AG has begun to address concerns about a lack of standards for efficiently connecting the plant floor to the enterprise business layer. SAP has begun addressing this situation by developing a roadmap that supports the ISA S95 specifications.

Businesses desiring to achieve maximum benefit from plant-floor generated data will most benefit from an information architecture that supports data scalability and abstraction, efficiently addresses the challenges of legacy systems, and capitalizes on available standards.