Software taking control

Linking the factory floor with other parts of the enterprise will drive much development in 1997.

Interoperability, field-based, real-time, modularity, objectoriented technology, open architecture, integrated system, enterprise-wide-these are some of the buzzwords as the process control industry enters 1997.

Faced with downsized engineering and other technical staffs, plant and corporate managers see integrated enterprise-wide information systems as the key to improving productivity and profits. During 1997, the relative importance of hardware versus software will continue to slide inevitably and inexorably towards the soft end of the continuum.

Industry experts foresee increased user interest in field-based control systems and "smart" instruments, with analog devices losing some of their traditional popularity-especially for new process control installations. Vendors will continue to respond to long-time user demands for more open control architectures, resulting in a broader array of equipment choices and allowing smoother integration of control with the enterprise. Meanwhile, Internet- and intranet-based monitoring and control systems are rapidly appearing in the marketplace, shoving security and other new technical challenges to the forefront. Sales growth will rise to 13%

According to a recent report by market research firm FIND/SVP, New York, N.Y., 1996 global sales of process-control equipment were valued at about $29.4 billion. The study, The World Market for Process Controls, predicts total process control equipment sales growth will rise to 13.5% per year by 2000, when worldwide sales will reach $49 billion (see Table 1).

Computerized hardware and software will continue to enjoy the most rapid increases, reflecting the growing importance of software and advances in computer technology, network communications, and user interfaces. The research firm also sees advanced process control becoming more practical, and emergence of the PC and Windows 95 or NT, especially NT, as the preferred platform for new control systems.

The world process-control market is dominated by chemical and petrochemical end-use industries, which accounted for 31.5% of sales or $8.2 billion in 1995, FIND/SVP says. Chemicals are predicted to narrowly outpace the overall market, growing 14.2% a year versus 13.5% for petrochemicals, through 2000. Water and wastewater will emerge as the next largest segment, with 18.5% compound annual growth rate (CAGR) from 1995 to 2000. Oil and gas applications will rise from 1995's $3.4 billion, but at a somewhat lower rate, slipping to the numberthree end-use rank, the report contends.

Geographically, the 1996 process-controlequipment market was divided as North America (36.7%), western Europe (33.5%), Japan (15.3%), and rest of world (14.5%). FIND/SVP sees improved growth in all markets from 1995 to 2000, but forecasts that developing markets (India, China, etc.) will continue to gain global share, growing 18.7% per year, rising to $8.3 billion. Global demand will rise a comparatively modest 13.5%, with Europe lagging the other major markets, as shown in Table 2.

MES on front burner

A major driving force in 1997 will be integrating plant control systems with other enterprise computing systems, via manufacturing execution systems (MESs) and other business management systems, to improve productivity and to reduce costs. Users, seeking to keep software options open, are demanding standardized interfaces.

Several companies are now, or soon will be, marketing systems that go part way towards complete integration. At the business end of the spectrum, for example, SAP will release its IS-Oil industry-specific comprehensive solution that optimizes R/3 for the oil and gas industry supply chain and links crude supply to commercial and retail customers. SAP has also published specifications for 100 new business application programming interfaces (BAPIs) that will enable SAP customers to integrate R/3 business processes directly across the supply chain and the Internet.

Elsag Bailey (Symphony), Fisher-Rosemount (PlantWeb), Foxboro (I/A Series), Honeywell (TotalPlant Solution), and Smar are among suppliers offering wide-reaching, more open, management and control systems.

There is considerable confusion in industry with the term MES, and it would appear that suppliers need to educate their prospects on the concept, and on how their individual software products fit into an MES. These findings are detailed in a recent report prepared by Thomas Marketing Information Center (TMIC), New York, N.Y.

The study, Manufacturing Execution Systems User Study summarizes interviews with MES users and prospective users. Respondents were asked if they are using any MES software applications at their location and if they are planning to purchase more within the next 18 months. It was reported that 48% were not using MES applications and had no plans to do so, but the surveyors believe that some may have said that because they didn't know about, or didn't understand, MES.

Among those using and/or planning to implement MES, more than half are using a variety of applications falling into three of the four MES application categories shown in Table 3.

The TMIC report shows the most widely used application among current users was Production Reporting and Tracking, selected by 79%, followed by Resource Allocation and Monitoring, used by 55.3%. ISO 9000 Documentation and Controls, not in the top 10 applications currently being used, was tied for the number one application selected by companies planning to implement MES within 18 months, together with Production Reporting and Tracking, and Data Collection.

ISA standards committee ISA SP95 was recently organized to develop an industry-wide standard for this integration effort, and will endeavor to define a data-transfer methodology that allows control and business systems to communicate while maintaining security and integrity. (See related story by SP95 chairman on this page.)

Object technology emerging

A dominant effort in software development in 1997 will be object-oriented technology (OOT), improving interoperability among automation devices such as distributed control systems (DCSs), programmable logic controllers (PLCs), and smart field devices, and data sharing over networks between management and the plant floor.

The original Microsoft technology called object linking and embedding (OLE), which allows information to be moved and updated among different applications, has been superseded by Microsoft's COM (common object model) and DCOM (distributed component object model). These technologies form the basis of OOT-based products being developed by the recently incorporated OPC (OLE for process control) Foundation, and by individual vendors.

The new OPC Foundation, formed at ISA/96 Chicago, replaces the previous OPC Task Force. Organized along the lines of the Fieldbus Foundation, the OPC group's work is already well underway. (See related story by OPC Foundation president, p. 39.)

MES to become MES II

Although MES implementations have yielded benefits, these have been limited by cost, industryspecific requirements, and market fragmentation. Bill Swanton, director of research at Advanced Manufacturing Research, Inc. (AMR), Boston, Mass., believes that "object technology is changing the landscape of plant software." Even as corporations implement ERP systems, AMR sees networked business object (NBO) technology further restructuring worldwide commerce, and believes that NBOs will fundamentally change the relationship between the plant and the rest of the enterprise. Swanton believes Microsoft's COM, DCOM, and OLE, with both shop-floor and plant-floor systems moving quickly to Windows NT as the preferred platform, will significantly decrease integration costs.

"Current MES and control offerings are evolving to be the cornerstone of a new breed of plant software," said Swanton. "Their tracking and scheduling functions are being enhanced with select planning and inventory management capabilities. These manufacturing execution solutions (MES IIs) will plug into a corporate ERP backbone." MES Is will incorporate industry-specific plant-level planning and execution systems and will manage production control in conjunction with ERP "To be successful, however, the relationship of ERP systems to the plant must be redefined," opined Swanton. "Rather than trying to plan all details of production, ERP systems must take their cue from modern management techniques. They must delegate the production to the plant, telling them what to make, when it's needed, but not how to do it."

Using functional components (NBOs), OOT will ease cooperation between applications, and will "finally break the need for large-scale, singlevendor systems with a common data model by allowing integration to be accomplished after the fact," the researcher said. "Object technology, NBOs, and component software hold the promise of breaking vendors loose from the shackles that have limited growth over the past few years. By reducing end-user costs, AMR believes the volume in this market will increase significantly over the next two years," he predicts.

Open architecture boosts options

Manufacturers are gradually moving from proprietary to open-architecture factory-floor workstations and upgrading to operator interfaces (Ols) that allow easy transferal of plant-floor data to supervisory control and data acquisition (SCADA) environments. The resulting integration should improve access to real-time and accurate information and increase efficiency. Unfortunately, unlike control-room SCADA applications, which are already moving to open architecture and displacing vendor-specific DCSs, factory-floor operator workstations are still dominated by proprietary OI products that cannot easily be integrated. However, designers are now producing PC-based solutions for factory-floor applications in order to avoid proprietary architecture limitations. Compatibilities enabled by open architecture will allow a choice of vendors; will increase flexibility by granting linkage of workstations with other, most appropriate, vendors' I/O; and will improve cost-effectiveness.

Two-layer hierarchy emerges

In an article entitled "Open control systems must be open above and below," Andy Chatha, president of Automation Research Corp. (ARC), Dedham, Mass., a market research firm specializing in process industries worldwide, notes that "While connectivity is the basic goal, software interoperability is far more important and more difficult to accomplish. For software to be truly interoperable, many different aspects, including operating systems, networks, databases, languages, and user interfaces, must be taken into account." ARC believes that an open control system "must provide information integration with the business systems above and the numerous field devices below," and will be a significant component of an emerging two-layer automation hierarchy as shown in Figure 1.

ARC believes that all control companies will increasingly use PCs wherever they can over the next few years and that Windows NT will be an excellent platform for control once OPC enhancements become available, probably within the next two years.

Many control companies already consider that migration of their systems to the Windows NT platform will constitute de facto open control. This is demonstrated in an ARC study entitled Worldwide DCS Outlook:1995 Edition. ARC foresees PC-based control architectures having an impact on DCS hardware revenues and expects worldwide unbundled DCS application software revenues will see an average annual increase exceeding 21% through 2000. "The DCS business is fast becoming a software and services business. Declining prices and the emerging fieldbus standard will continue to drive the decline in bundled DCS hardware and software revenues," ARC says.

"More expensive workstations will be replaced with off-the-shelf PCs in the future, and adoption of Windows NT will allow for increased control at the PC level, translating to reduced control processor revenues," ARC predicts. "Increased competition from third-party suppliers in the area of control and OI software will also drive prices down." ARC observes that major PLC companies are also trying to get into the software business and are betting on the Windows NT platform rather than Unix, even though existing NT versions may not be ideal for real-time control.

Smart instruments gaining

The value of electronic pressure transmitters, transducers, and component level sensors manufactured in and imported into the U.S. in 1996 was about $1.6 billion, according to a market study, The MS. Pressure Transducer, Transmitter, and Component Level Sensor Industry: Seventh Edition, available from market research consulting firm, Venture Development Corp. (VDC), Natick, Mass. U.S. consumption of all three categories is forecast to increase over the next four years.

U.S. sales of smart electronic process pressure transmitters, which incorporate digital data communication capabilities, reached record levels in 1996 and have outpaced sales of traditional types (that generate analog data signals) during the past five years, according to the VDC study. The smart types often incorporate selfdiagnostics, electronic reranging, and temperature correction, features that make these devices relatively maintenance free and suitable for use in many remote process-control applications.

A total of about $362 million worth of process pressure transmitters were shipped by suppliers to U.S. markets in 1996, with 63% of this dollar value in smart types. Much of the increase in the use of smart transmitters has come at the expense of traditional types, which have been declining in shipments. VDC forecasts that the domestic consumption of process pressure transmitters will increase at 3. 1% CAGR over the next four years. Consumption of smart types will grow at 7.9% CAGR, while that of traditional analog types is expected to decrease at an 8.2% annual rate and that for low-cost, nonrepairable types is expected to decrease at a 2.6% rate-these forecasts are presented graphically in Figure 4.

VDC also forecasts 6.5% CAGR increases in domestic consumption for both nonprocess pressure transmitters/transducers and for component-level pressure sensors over the same period.

Sensor market shifting

In general, the sensor market is relatively well established, but according to a strategic market analysis conducted by international marketing consulting company Frost & Sullivan (F&S), Mountain View, Calif., there are several new factors influencing the market.

F&S's recent study World Sensor Technology Assessment: Pressure, Flow, and Level estimates the worldwide market for pressure sensors was about $2.59 billion in 1996 and will expand to $2.79 billion in 1997 (see Figure 5). The more recent developments within this category include silicon micromachined pressure sensors (especially captive and resonating types) and fiber-optic pressure sensors. Additionally there is some development in vacuum gauges, but at this point the technology is still limited.

F&S estimated the worldwide market for flow sensors at about $3.12 billion in 1996 and predicts it will grow to $3.24 billion in 1997 (see Figure 6). The area within this market that has experienced the most development is mass flow sensors, both Coriolis and thermal types. Other developments occurred in the vortex flowmeters market with the introduction of direct mass measurement capabilities. Future opportunities are also being explored in the field of ultrasonic technology.

The level sensors worldwide market was about $ 1.05 billion in 1996 with overall growth rates of 5% to 6%, F&S says. Developments in this market segment are mainly concentrated in the noncontacting technologies such as ultrasonic and microwave (radar).

F&S adds that key technologies contributing to sensor market growth include miniaturization of sensors and development of smart sensors that can offer multivariable capabilities, self-diagnostics, and self-calibration features. IT

Copyright Instrument Society of America Jan 1997
Provided by ProQuest Information and Learning Company. All rights Reserved