Manufacturing is the backbone of the United States economy, accounting for almost one-fifth of GDP over the last 40 years and 16 percent of all jobs. There are thousands of individual companies and plants but these plants and their equipment on average are more than 20 years old. Manufacturing operations will require a continual infusion of the newest information and control technologies if U.S. industry is to maintain its global ability to deliver products reliably that best serve the customer at the lowest cost. The agility and economy that can be provided by new integrated systems needs to be exploited to maintain and improve the competitiveness of U.S. industry.

One common factor contributing to the inability to realize these economic benefits is that the information technology originally was built into specific existing equipment and processes by developers who often invented the underlying software and interface concepts with a focus on specific markets (i.e., the chemical process industry or automotive assembly). This has resulted in a plethora of more-or-less proprietary tools, interfaces and data formats embodying incompatible semantics and computing technologies. These incompatibilities are costly because they lead to uncertainties in data, information, and knowledge as they are transmitted from process to process or between process and human. Uncertainty leads to slowed decision processes and errors in communication. Data, information and knowledge are frequently useful to only a small subset of the enterprise or not available in a timely or useful form, minimizing feed-forward capabilities and benefits. People who move from process to process or from machine to machine must unlearn the old tools and interfaces and learn new ones.

Dealing with these incompatibilities has been estimated to cost U.S. industry billions of dollars annually. ⁽²⁾ Evidence of this impact has been provided through a variety of studies including:

• An Electric Power Research Institute (EPRI) study estimate that improved process instrumentation can lead to improved fossil fuel power plant heat rates of up to 1 percent. This would result in an annual savings of $500 million per year, or improved production rates of up to 12 percent in nuclear power plants. ⁽³⁾
• A U.S. chemical processing industry study estimate that better modeling, monitoring and control would allow analysis of possible off design conditions directly improving product quality and addressing safety hazards.
• A General Motors study documenting that changing robotic assembly instructions costs as much as $5000 to have a vendor reprogram one step of a closed-architecture controller, but only costs $380 for an open-architecture, PC-based controller. ⁽⁴⁾

These three studies, as well as the many others, show that in general, industrial users of next generation manufacturing control systems will be able to realize the following business and economic benefits:

• Improved product quality through more precise control, on-line inspection and correction, and enhanced design for processing or manufacturing.
• Reduced operating costs through increased energy efficiency, increased productivity by way of better resource utilization, reduced downtime, cycle time, and inventories, reduced scrap and rework, and increased capacity by eliminating bottlenecks.
• Reduced costs for new product introductions through quicker response to customer demands via reduced cycle time, increased flexibility for existing facilities, reduced time and cost of starting-up new facilities, reduced new product prototyping costs, and the ability to serve smaller markets.
• Enhanced environmental protection through better emission/effluent control with lower energy consumption.
• Increased safety during processing as well as through reductions in product defects.

To realize these benefits, U.S. industry will need to address at least four items:

1. The technical challenges associated with high-risk problems of developing and integrating the next generation of controls.
2. The generic nature of infrastructural problems means no one end-user has the knowledge or resources to solve the technical problems. Vertical integration including sensor, actuator, and control vendors and industrial users is needed to overcome the technical challenges of this focused program. Also needed is horizontal integration to diffuse the technology across multiple industry sectors, successfully developing a technology infrastructure and leading to significant implementation cost reductions to the end-user.
3. The technical challenges of legacy systems provides some technical risk, but the actual cost of upgrading systems must be born by industry.
4. Standards will have to be agreed upon to address protocols, interfaces, data transmission, architectures, etc. These standards must be defined by industry, and in fact are being discussed and developed now, illustrating industry's desire and commitment to see the implementation of intelligent control systems.

Global economic considerations

In 1998, the world market for industrial and analytical instruments should reach $118 billion. ⁽⁵⁾ By 2002, the market is expected to grow to $140 billion, an increase of almost 19 percent. During this time, the U.S. market share is projected to fall from 37 percent in 1998 to 34 percent in 2002. One reason for this falling market share will be growth in newly industrialized economies. As their economies mature and are able to produce control system technologies, their need to import is expected to decrease.

In spite of this expected trend, U.S. equipment is still considered very competitive in North America, Europe and Asia. However, foreign companies are becoming increasingly strong, especially in higher-level, microprocessor-based systems. While Japan and Germany are leading suppliers of U.S. imports, there are at least 300 companies worldwide that produce process controls. At least half these companies are active in the U.S. This is expected to yield a $6 billion trade surplus in 1998 for the U.S., principally through its most important trading partners (Western Europe, Japan and the Chinese Economic Area, and NAFTA). ⁽⁶⁾ In terms of plant wide integration, use of distributed control systems and programmable logic controllers (PLC's) will continue to increase. ⁽⁷⁾ Next generation measurement and control instrumentation could easily add billions of dollars to domestic sales of this U.S. industry sector and create very substantial export opportunities. However, this growth will only be realized if the U.S. controls industry takes a leadership role in the development of the next generation of controls systems.

U.S. companies historically have not pursued opportunities for a global competitive advantage. For example, one strategic element in modern discrete part manufacturing operations is electronic tool control. Foreign companies foresaw an opportunity to establish market leadership in manufacturing and targeted machine tool and robot controllers as critical technology and invested heavily, including overt government aid, to capture this cornerstone of modern manufacturing. Dominating this controller market led to foreign dominance in the related fields of robotics, machine tools, and eventually, general manufacturing. As a result, foreign-owned or controlled companies monopolize this critical technology. The balance of the market is shared by dozens of small companies, largely targeting niche markets and lacking the resources to mount any serious effort to shift from current proprietary, incompatible architecture to a single open architecture industry that is cross platform compatible.

Intelligent Control's Incremental Economic Benefits

When ATP's Intelligent Control focused program has been successfully completed, a self-staining and growing infrastructure based on next generation integrated, intelligent control systems will be achieved.

The value of this legacy eventually must be measured by its social benefits. Social benefits are divided into private benefits (i.e., going to the innovators/project participants) and spillover benefits (i.e., going to industries, companies, and individuals that were not directly involved in the project).

Private benefits will accrue to developers of the control systems. It is anticipated that this group will include system developers and integrators as well as universities and national laboratories which contribute innovative new ideas to the development effort. They will benefit through greater sales in both traditional and new markets. Also, since a proposed project must apply its solution to a common industrial control problem that impacts more than one industrial group, there will also be private benefits accrued by the end-user whose problem was addressed by the proposed project.

The industrial end-users will also be spillover beneficiaries, and, in fact, this should represent a possibly greater portion of the incremental economic benefits. Industrial end-users will be able to purchase and implement control systems that will allow them greater product quality, flexibility and responsiveness, as well as a quicker returns on capital investment but which they would not have been able to develop on their own. This in turn benefits the consumers by offering lower costs, higher quality, increased product safety, and greater variety products.

There has been a realization by the controls community that customers are demanding open-architecture-based control systems. Further, if the controls community does not respond to this demand, they will be reduced from systems suppliers to component suppliers. There is thus considerable interest by the controls community in developing the next generation of controls. The end-users are defining their control system expectations, but have been hesitant to pursue the development of these new systems because they are skeptical that generic solutions can be developed. This concern can be overcome through appropriate teaming and technology demonstration.

Evidence that industrial is in fact committed to a research and development program whose goal is to develop these new control systems has been demonstrated through input from a variety of sources including individual companies from a broad variety of industries; industrial associations and study groups; suppliers of control systems, software, and systems integrators; U.S. laboratories; Government programs; and universities.

Prior History

ATP began gathering information concerning the development of a focused program area in sensors and controls between 1993 and 1995. During that period, ATP received approximately 40 white papers suggesting research topics for sensing devices, nondestructive evaluation, controls, materials and chemical processing. These papers were submitted by Fortune 500 companies as well as small and medium companies from various industries.

ATP also received numerous letters of interest and endorsement and support for government financial assistance in the area of sensors and control systems research and development. The need identified in these letters was based on the high risk, high cost nature of this development effort that had a payoff some time in the indefinite future.

In May 1994, 430 participants attended a workshop entitled Sensing and Control. One outcome of that workshop was a draft proposal in August 1995 for an ATP focused program area titled Sensors and Control Systems in Bulk Processing Industries. ⁽⁸⁾The 1995 program focused on low cost/high accuracy sensors (pressure, temperature, volumetric flow, level, etc.) and advanced control systems for electrical utilities, pulp/paper, chemical, petrochemical, food, metal, and other industries. For various reasons, including ATP budgetary constraints and issues of program scope, the 1995 proposal was not implemented.

Discussions were held with interested parties and the program plan revised. This revision was presented at an ATP workshop for Intelligent Control in Chicago on November 20-21, 1997. Participation included 28 small/medium companies, 32 large companies, 11 associations, 19 national labs, and 26 universities for a total attendance of 108 interested parties. Additional interest has taken the form of phone and e-mail inquiries and comments. Several outcomes resulted from this recent workshop:

• A consensus definition of an "Intelligent Control System" as it pertains to this focused program:

"An intelligent system has knowledge of the goal and the environment based on process and product models and verifiable data from sensors. It is able to interact with the equipment to control the environment to achieve the goal. The system has the ability to learn from system/environment interaction and is fault tolerant."

• Fine tuning of the program's technical scope. Defining a technical scope that is not too broad (i.e., a scope that does not end with a collection of disjointed technical solutions incapable of encouraging and supporting infrastructure growth) has been difficult. The key to tightening the technical scope, while still encouraging infrastructure building, has been the requirement that project solutions address common classes of control problems. According to the Chicago workshop participants, the technical challenges encompassed by the revised scope are still hurdles that have yet to be overcome.

Domestic activities

A number of other significant inputs have been received and incorporated into Intelligent Control's program plan. These included the Next-Generation Manufacturing (NGM) Project ⁽⁹⁾ and numerous strategic plans and road maps from individual industries including aluminum, steel,metal-casting, heat treating, chemical processing, glass, and forest, wood, and paper. ⁽¹⁰⁾This information identified manufacturing control improvements as a priority need to foster U.S. competitiveness in the next century. ⁽¹¹⁾

Other initiatives from industry and industrial trade associations included General Motors, Ford and Chrysler initiative for Open Modular Architecture Controller (OMAC) ⁽¹²⁾ to establish specifications to be used by vendors that sell controllers to the automotive and aerospace industries; the Department of Energy's (DOE) Technologies Enabling Agile Manufacturing (TEAM) program to develop specifications defining an intelligent closed-loop controller environment to support open architecture concepts; and SEMATECH creating an application framework for computer integrated manufacturing within semiconductor factories. ⁽¹³⁾

Government activities

A number of U.S. government programs support industry efforts to improve manufacturing process controls. These activities are being supported through at least seven different agencies and programs. Factors that distinguish the proposed Intelligent Control focused program from existing work by other federal agencies are ATP's emphasis on integrated systems solutions for manufacturing process control that are broadly applicable across multiple industry sectors. Other government efforts are typically focusing on the development of manufacturing or process controls for a single industry sector, or as part of the implementation of a specific process, or on component elements of a control system such as improved sensors, all of which are agency mission related. It is important to note, however, that the work of these other government programs provides synergy and a tremendous foundation for the Intelligent Control program that greatly contributes to this program's timeliness.

Industry participation in ATP's general and focused competitions

Another indicator of a steady, high level of industry interest in a program for Intelligent Control is participation in ATP's general and focused competitions. To date, ATP has made 12 awards totally $57 million in funding for proposals that address elements of Intelligent Control while industry has provided matching contributions of $63 million for a total of more than $120 million in related research.

The technologies this focused program aims to foster are primarily infrastructural with technology demonstrations that establish feasibility. They constitute an underlying foundation required to enable and support important applications of control, information and related technologies to industrial processes. Like other types of infrastructure, these technologies are recognized as the means to realizing widely shared benefits. Yet, these benefits are often slow to be realized by the originators. This leads to insufficient incentive to develop infrastructure technology on its own. In the case of industrial control, individual companies do not have the capabilities to develop the full collection of underlying technologies.

Many industries have adopted a product-oriented outlook and have concentrated their research and development resources in areas that customers can clearly identify and use for product differentiation. The fraction of spending allocated to process oriented research and development tends to focus on a shorter term, incremental improvements, in contrast with the major long term gains in performance that this ATP focused program will promote. Individual manufacturers will pay for functional capabilities. They do not invest in infrastructural technology that will yield benefits of equal-and perhaps greater-value to competitors who did not share in the costs.

Playing the role of catalyst, ATP can help the large number of specialized process control hardware and software vendors and manufacturers to overcome this impasse and focus their collective expertise and resources on surmounting barriers to developing broadly useful control applications. Eliminating these barriers requires the constructive participation of multiple industries to insure cross-fertilization and widespread adoption of the technologies to be developed. On the basis of input from a variety of industries, ATP believes that both process control vendors and manufacturers see it in their best interests to participate in such efforts. Manufacturers realize that, with today's customized manufacturing applications, they cannot reconfigure their operations quickly enough to react to changing markets and new business opportunities. ATP's cost-shared support mitigates the "free rider" problem that has stalled efforts to correct the well-recognized limitations of current systems. Control equipment and software vendors realize that, for their industry to thrive, they must follow the lead of the personal computer industry, developing the bases for architecture, interfaces, and other core technologies essential for broad-based system applications rather than the specialized and incompatible approaches of the past.

ATP's cost sharing leverages industry's limited research and development funds for high risk technology development, fosters cooperative and constructive partnerships between manufacturers and their suppliers, and encourages companies that would not otherwise have resources to carry out manufacturing process control technology on their own to move out aggressively and ambitiously.

Convergence of technologies and timing issues are the focus of "Why Now." Computer hardware, software and control algorithms based upon neural networks, autonomous agents, etc. are developing rapidly. New sensor designs, actuators, process models and analytical equipment are also being developed. What is lacking are the integration technologies of open architecture interfaces and modularity of hardware and software components that are necessary to support the development of intelligent control. This integration technology is just beginning to influence the marketplace for machine tool control where technical feasibility has been demonstrated, but significant inroads have not been made into other areas of manufacturing or process control.

Timing issues focus on the rate of developments of new control technologies overseas and the age and state of control technologies upon which domestic industries is now relying. The technical challenges are real, as is the need for support, thus providing ATP an opportunity to make a significant difference.

1. Current state-of-the-art relies on a variety of vendors contributing to a system, each using possibly different protocol, data link and interface, making modification, trouble shooting and maintenance extremely difficult.

2. Next Generation Manufacturing Project: A Framework for Action

3. Mechanical Engineering, September 1994, page 55

4. "PCs Undercut Top Factory-Robot Maker," The Wall Street Journal, Tuesday, May 20, 1997, p. A14

5. U.S. Industrial and Trade Outlook 1998: Industrial and Analytical Instruments, U.S. Department of Commerce/McGraw Hill, p. 23-1, 1998

6. U.S. Industrial and Trade Outlook 1998: Industrial and Analytical Instruments, U.S. Department of Commerce/McGraw Hill, p. 23-5, 1998

7. U.S. Industrial and Trade Outlook 1998: Industrial and Analytical Instruments, U.S. Department of Commerce/McGraw Hill, p. 23-5, 1998

8. Sensors & Control Systems: Their Application in Bulk Processing Industries, Ronald J. Repka

9. The Next Generation Manufacturing Project was initiated in 1995 to develop a framework for action that U.S. manufacturers, individually and collectively, can use as a guide to chart a course in an increasingly complex and competitive business environment. Individuals from more than 100 companies, industry associations, government agencies, and academic institutions worked together to develop a broadly acceptable framework for next-generation manufacturing enterprises. They identified key competitive drivers; defined required attributes and imperatives responding to these drivers; and developed recommendations for action that industry, government, and academia can take to help American manufacturers thrive in the 21st century.

10. The U.S. Department of Energy, Office of Industrial Technology facilitated these industry assessment and planning efforts. These road maps identified many research topics that need to be addressed, with only those related to this proposed focused program being cited here. It should also be noted that DOE's budget only allows it to support a fraction of the identified needs and is primarily directed at energy-related considerations.

11. Next-Generation Manufacturing Project: A Framework for Action, Vol. III, Digest of U.S. Industry Road Maps.

12. Requirements of Open Modular Architecture Controllers for Applications in the Automotive Industry

13. http://www.sematech.org/member/division/fi/cimhome.htm

Date created: November 1997
Last updated: April 12, 2005