Page 119
6—
Information Infrastructure Issues
Introduction
Manufacturing information infrastructure refers to the computing and communications facilities and services needed to facilitate, manage, and enable efficient manufacturing; key elements include database and information management systems, data communications networks and associated services, and management of applications software. For the next generation of manufacturing, information infrastructure must have the high degree of connectivity, compatibility, and ease of use that already characterizes traditional physical infrastructure (electric power, water and sewage services, telephone service, and the like).
A vision for the manufacturing enterprise of the 21st century includes the ability to use an integrated information system to support the entire product cycle from product design to product delivery to the end user. Such a system would facilitate communications among groups, automate "corporate memory," and aid in decision making. Likely using the National Information Infrastructure (NII), this system would function at both the intrafirm (e.g., connecting branch offices dispersed all over the country) and interfirm (e.g., connecting the firm to suppliers and customers) levels. Such opportunities are particularly relevant to small manufacturers that will be able to find niches for their specialized capabilities more easily when the NII enables convenient electronic connections for firms small and large.
The use of networked infrastructures can diminish the role played by place, time, and hierarchy in the structure and management of organizations. A wide range of organizational possibilities remain to be fully explored, and thus investigation
Page 120
TABLE 6.1 Research to Advance Infrastructure Systems
| Subject Area | Example of Research Needed |
| Architectures and standards | Standard manufacturing control architectures Generic functionality within control architectures Cost-benefit criteria for potential standards |
| Intra-enterprise and inter-enterprise integration | Principles and architectures for coupling network-based applications |
| Automatic interpretation of transactions | |
| Automatic message routing and associated processing | |
| Support for multiple protocols and multiple speeds over a given medium | |
| Time-critical message delivery in interconnected networks | |
| Protocols and services that support specific demands of object-oriented applications | |
| Services for human- and machine-based browsing and searching of information and resources | |
| Tools and techniques supporting supply-chain dynamics and associated planning | |
| Mechanisms and systems to support information session management | |
| Architectures for autonomy and distributed intelligence | Autonomous agents to monitor and respond to production events |
| Knowledge agents for enterprise-wide management of models, names, transactions, and rules | |
| Architectures for manufacturing systems involving distributed intelligence | |
| Tools to find and distribute information | |
| Stable sets of rules for interacting agents | |
| Dynamic variations in agent autonomy | |
| Information retrieval systems | Modeling and prototyping functions for user interfaces |
| Next-generation data manipulation languages | |
| Maintenance of data consistency and integrity through database updates | |
| Network-based services for information browsing and searching | |
| Software engineering | Simplification of system designs, operation, and maintenance |
| Self-healing systems | |
| Support for new programming paradigms | |
| Tools for component-based architecture life-cycle approaches | |
| General tools unconstrained by the limitations of specific programming languages | |
| Tools with aspects of knowledge-based collaboration software |
Page 121
TABLE 6.1 continued
| Subject Area | Example of Research Needed |
| Software engineering (continued) | Rapid prototyping and other methodologies for faster development |
| Techniques for encapsulating legacy systems and developing mediator support | |
| Analysis methodologies, metrics, and selection techniques | |
| Reference architectures that cut across manufacturing domains to support software reusability | |
| Data representations that depict both spatial and temporal aspects | |
| Tools for better and more user-configurable user interfaces | |
| Dependable computing systems | Better technology to support changing software without removing a system from operation |
| Continuous availability of on-line services | |
| Fault-tolerant hardware and software | |
| Increased system security and trustworthiness | |
| Collaboration technology | Software, user interfaces, and hardware to support cooperative work |
into new information technology (IT)-enabled organizational forms is an area of research interest, as new organizational models will ultimately be needed, or earlier paradigms for understanding organizations revised, to better understand how activities might be reorganized to exploit information technology for better (or higher) performance.
Progress toward this goal involves development of both physical (e.g., wires, computers, gateways, and switches) and "soft" (e.g., protocols and information services) information infrastructure technologies. Areas with research opportunities are listed in Table 6.1 and are discussed in the remainder of this chapter. For convenience, these topics are divided into areas specific to manufacturing (architectures and standards, and integration of activities within and among enterprises) and not specific to manufacturing, although the dividing line between manufacturing-specific and nonmanufacturing-specific categories is particularly fuzzy for infrastructure elements.
Architectures And Standards
Architectures and standards provide structure for many levels of interconnection, from those among devices to those among businesses. There are architectures for systems supporting specific functions, for enterprise-wide information flow, for control of processes, and for management. As the need for
Page 122
exchanging data and knowledge among various manufacturing applications grows and the sophistication of individual applications increases, the interfaces between them become fuzzier and more complex and important. Although each interface may itself be easy to use, correct, and robust, the complex of interfaces in a large, diversified factory is fragile and is subject to failure for relatively insignificant cause.1 As a result, applications developers spend disproportionate time and effort addressing the interfaces, per se, rather than the application itself.
Standards support the passing of information between the various elements of a manufacturing enterprise's information system. For example, they facilitate connections from computer-aided design (CAD) to computer-augmented process planning to computer-aided manufacturing (CAM). Standardization provides the benefits of common systems (cost savings, improved integration ("plug and play" equipment and systems) and information flow, and so on).
The vision for 21st-century manufacturing presumes that interconnecting manufacturing applications will be as simple as connecting household appliances to a power grid—one need only know how to run the application (equivalent to using a microwave oven) and manage the interface (plug it in and press a few buttons). This ease of interconnection and interoperation extends from devices found on the factory floor to applications connecting the factory to the product design facility to applications connecting an enterprise to its suppliers and customers. An "application socket" for manufacturing would benefit both equipment vendors and customers and enhance factory performance.
The desirability of common standards notwithstanding, businesses have largely been unable to agree on protocols or standards of communication for different kinds of requests. One reason for this failure has been inadequate technology (e.g., the use of primitive fixed-format data representational formats) and excessive complexity. But a more important reason is that attempts to reach a common technical understanding have been hampered by epistemological differences reflected in the unstated assumptions of each participant in a technical discussion. Social factors relevant to the adoption of standards are discussed in Chapter 7.
Research is needed to develop better manufacturing architecture, standards, and interfaces, including research to develop standard equipment control architectures and generic functionality within the architectures, to support general manufacturing information standards, and to lower the cost of more open, less proprietary approaches. Especially desirable would be architectures whose standards accommodate some upgrade capability, so that technology vendors could worry less about premature freezing of technology and the locking out of potential
1 Wysk (1992) identified as an obstacle to implementing computer-integrated manufacturing the difficulty of integrating information system components (i.e., hardware and software) with information (both internal and external) into a smoothly running system.
Page 123
competitive advantages, while customers could worry less about intrinsic obsolescence.
The importance of standards for manufacturing interconnections raises the question of where technology research enters into the development of standards; the answer is not clear. For example, efforts to develop PDES/STEP (product data exchange using the standard for the exchange of product model data) are not research per se, but the kind of problem that PDES is intended to solve may require research.2
Because of the importance of international standards in the global marketplace, strong participation of U.S. interests in standards making is valuable, and greater involvement of U.S. academic researchers (who are less involved than researchers in other countries) in standards making may strengthen the U.S. technology base in manufacturing. Research is a foundation for the consensus building inherent in standardization, and researchers (from industry, academia, and government) should bring their insights to bear on that process.
Integration
Redesign of Business Practices
Today, workers talking on the telephone, typing on keyboards, obtaining parts, scheduling factory operations, and setting up delivery schedules contribute much more to manufacturing costs than do the workers performing "touch labor."3 Lessons learned in both service and manufacturing industries attest to the value of systematic redesign of business practices and to the key role played by information technology and infrastructure in such reengineering.4
In particular, workers in a factory may be working very hard on an inappropriate set of tasks. They may be doing things that are not necessary, doing them inefficiently or even redundantly, or doing their work in such a way that it is not usable or accessible elsewhere. Information technology can be used to help identify such problems. For example, when modeling techniques are used and
2 The PDES efforts are not standards making in the traditional sense. Like many other activities that have moved into the standards arena, they involve significant levels of development activity as opposed to simple harmonization of approaches. This is true in many other areas of high-technology standardization, such as the unified approach to conceptual data modeling.
3 For example, Miller and Vollmann (1985) assert that "overhead costs as a percentage of value added in U.S. industries continue to increase and have become a major concern of manufacturing managers" and that "the growing implementation of factory automation increases overhead costs as a percentage of value added." They point out that ''most overhead costs are created by transactions, not physical products. Such hidden factory costs are generated by logistical, balancing, quality, and change transactions."
4 See Chapter 6 in CSTB (1994a). For a discussion of business process reengineering, see Hammer and Champy (1993).
Page 124
the model is compared with an actual process, steps that do not add value can often be identified. Such information is needed if management is to support a process of continuous change and improvement. Once such problems are identified, information technology can often play a key role in restructuring business processes for greater effectiveness and efficiency. IT can help to reduce the number of tasks involved in an activity, change the nature and sequence of tasks, and reduce the total time involved in a business process
Intra-enterprise Integration
Within a manufacturing company, activities such as design, production, marketing, finance, sales, use of human resources, and distribution have historically been linked to each other on the basis of issues such as ease of management and physical proximity. For example, using the manual and semiautomated manufacturing information systems of the past, integration was achieved through interactions between people where knowledge, data, and project status were shared through formal conversation in meetings and reviews and informal conversation in the hallways.
Today, electronic mail, shared and remotely accessible databases, "groupware," and other forms of electronic communication present new options for centralizing and dispersing activities and also for supporting cross-functional teams and processes. For example, enhanced communications options tend to be associated with greater decentralization, obviating the need for all tasks to flow through an office or headquarters. Electronic communications enable a much higher degree of geographic independence. While communication has always been fundamental to large organizations with distributed establishments, IT supports ever greater geographic dispersion of personnel, allowing firms to reap the benefits of lower-cost production labor or specialized expertise associated with specific geographic regions.
What has stood in the way of a greater degree of intra-enterprise integration? There are many causes, but one major reason is that the technical integration of various applications into a smoothly running system has been and continues to be quite problematic. Commercial software from one source does not integrate well with software from other commercial sources or with internally developed software, revisions come out slowly and hence frustrate users who expect a rapid response, and so on. System-level issues have received relatively little attention, both because of their difficulty and because they relate to problems that fall between the offerings of individual vendors. Customizing of general applications to specific customer needs and integrating applications and technologies from a variety of sources make development of a robust manufacturing system a formidable task. Such problems have plagued all kinds of organizations, but the combination of physical and information-based work in manufacturing adds further complications.
Page 125
For example, several manufacturing companies have invested many thousands of person-hours in developing and integrating factory information systems, a disproportionately large investment in comparison with their investment in other parts of the factory system. This imbalance is especially serious considering that these companies have often purchased "finished" software application systems from commercial sources, which when measured against the benefits received have proved to be less than satisfying. Since large-scale systems for manufacturing are often one of a kind, they present special problems because they are developed by teams with relatively limited experience with the kind of system needed, and they generally entail very high life-cycle costs.
A second illustration is the integration problems associated with CAD tools for mechanical products. Each vendor of such tools has its own database design, and exchange of models between different CAD programs is possible only for relatively simple constructs, despite the existence of a standard called the initial graphics exchange specification (IGES). Each CAD tool has its own unique interface that may take months to learn, and each comes with a different set of software tools for manufacturing support that may provide functions such as numerically controlled machining, sheet metal bending, or electrical wire routing.
Such incompatibilities are found throughout various manufacturing applications, and every phase of the design and manufacturing process requires its own distinct computer model and data representation. The representation of a part in a database may appear differently for a machining operation than it would for an inspection operation. The design representation necessary to support an engineering drawing is different from the design representation necessary to support a finite element analysis, and in most cases there are no automatic algorithms available to convert from one representation to another. For data other than data on the shapes of parts, few good computer representations exist at all. Even when the interfaces are perfect, a small design change requires a tedious and labor-intensive process of propagating the changes through all the CAD tools and representations. Box 6.1 elaborates on the fundamental knowledge gaps that underlie difficulties in integrating applications.
Databases and database management systems, equipment control systems, communications systems, and design support tools all present a multifaceted challenge to achieving interoperability. Meeting that challenge requires a deep understanding of underlying data constructs, the operations that constitute the relevant manipulations, and the assumptions behind various operations and calculations.5 While now readily available in certain domains (e.g., the representation
5 For example, Defense Department officials tell a Gulf War story regarding an inquiry about the number of MREs (Meals, Ready to Eat) available. The answer that came back was "–38,000." This ridiculous answer was the result of incompatibilities between two databases: one defined an "MRE" as a single individual package (one meal for one soldier), while the other defined an "MRE" as a pallet of 500 individual packages.
Page 126
| BOX 6.1 Knowledge Gaps Relevant to Intra-enterprise Integration | |
| • | The science and engineering of many phenomena underlying today's products are not well understood. Examples include the behavior of composite materials, failure modes of complex computer systems, turbulent fluid flow, failure modes such as crack growth and fatigue, and material and chemical incompatibilities. |
| • | The models that do exist capture one phenomenon, behavior, or mode of energy storage at a time, while most complex systems exhibit multiple modes at once. Integration of single-mode models is also problematic. |
| • | While the main dimensions of parts and parameter values of an electrical or chemical nature can be specified with reasonable certainty, less is known about how much variation about those nominal values is permissible. The current means of setting tolerances are historical or very conservative. An example is that materials specifications are often stated without indication of variances. The result is that costs are raised and materials and energy are wasted, both in production and in the over-engineering of products that will perform even if tolerances happen to reinforce rather than cancel each other. |
| • | The most efficient ways to carry out design processes to address all the interacting variables are not known. Right now an iterative and "concurrent" process is used, relying heavily on human memory and face-to-face communication. |
| • | It is not known how to construct and populate a database that will contain in an easily accessible form all the information that the designers of modern products will need. |
| • | It has been difficult to develop standards for data interchange to which all vendors will adhere. Indeed, a vendor seeking greater differentiation between its products and those of other vendors may well choose a proprietary data format or representation. At the same time, customers may not realize the waste and overhead incurred in retraining their personnel to use many different software packages. |
| • | Finally, we do not know how best to link people to a rising tide of increasingly complex information. For example, information systems must be able to provide information at appropriate levels of abstraction: a product designer may not need to know the details of how a partially finished part is held in a jig, although a production manager might. Information displayed for different individuals in the manufacturing system should reflect their different needs. The term "information ergonomics" has been coined to capture this challenge. |
of integer and real numbers), such an understanding is sorely lacking in many of the domains that manufacturing touches.
Commercial interests play a key role as well. Vendors have often chosen to develop products with proprietary interfaces and data representations. Suppliers of manufacturing equipment anxious to "differentiate" their products have obscured generic elements in their control computers, thwarting change and raising the cost of acquisition, operation, and maintenance. Such practices have tended to sacrifice interoperability to the desire to capture a set of customers. Not only
Page 127
do such proprietary designs inhibit communication with competitors' products, but they also block communication with adjacent machines and with older capital equipment and networks previously installed within a single plant. Communication is further confounded by network architectures that typically channel communication into hierarchies that mimic the power structure of an organization, inhibiting or complicating the exchange of simple messages between peers.
The result has been that information systems products from one vendor are not designed to be (and sometimes cannot be) integrated with similar products from a different vendor. (In some cases, even products from a single vendor are incompatible with each other.) Interfaces between these systems and equipment are often crude, causing many problems in their use and, according to many manufacturing executives, discrediting the entire information technology effort. Representations for product designs, process plans, and factory resources are quite often company- or even site-specific.6
Architectures (functional, control, communication, and management) also vary excessively. Furthermore, these architectures are difficult to describe and model. Defining architectural linkages and creating integration "hooks" for sensors, machine control, process planning, scheduling of maintenance and repair, and machine interfaces are as critical as creating faster and better devices for each of these areas. To a large degree, the presence of such hooks would give users the flexibility to assemble their own solutions from off-the-shelf software and hardware components.
The top-level goal is the effective and rapid exchange of data and knowledge between all operational systems and subsystems,7 although such a goal applied to research systems might well freeze technologies prematurely and inhibit improvement. Achieving it will require common languages, operating systems, and networks. Beyond common formats for data representation, a common understanding of data, that is, a common semantics, is also needed for complete knowledge interchange. Three important areas of research on the interconnection of applications include the following:
| 1. | Research is needed on organizing principles and architectures for connecting different network-based applications into a seamless environment. Such connection is necessary, for example, to link flexible manufacturing |
6 Internally developed integrated information technology solutions are always possible in principle. But as a practical matter, only the very largest firms have the resources to develop even partial information technology solutions in-house, and essentially no firm can do it all alone.
7 The need for commonality and interoperability among manufacturing applications of information technology is in some ways analogous to the problems faced by European engineers 250 years ago as they struggled to integrate mechanical systems that utilized a puzzling array of hole sizes, bearing sizes, and thread sizes. The crucial difference is that standardization of the latter took place over decades, while today U.S. manufacturing cannot afford to wait decades for these standards to arrive.
Page 128
| cells to a plant's scheduling function and to link the scheduling function to the enterprise order, delivery, and financial systems. Enterprise integration also implies a need for research to enable the automatic interpretation of the type of transaction being executed, the routing of the message to the right location for processing, and the processing that must occur when the message for the transaction reaches the correct system. | |
| 2. | Research is needed to support communications networks that can support the real-time control of manufacturing processes (e.g., a cutting tool shaping a piece of metal). Such networks transmit control messages, each of which must be received and acted on within some critical time, ▵t. While the process being controlled should be able to tolerate some variation in ▵t, the network must be able to guarantee delivery of that message unconditionally within that time. Many common protocol suites such as today's TCP/IP or Open Systems Interconnection deliver messages on a "best-effort" basis and cannot reliably meet such delivery-time requirements, while other protocols assure timely delivery by reserving communications bandwidth, usually on private lines and usually through restrictions on the type and number of messages carried. Since there is increasing demand for the ability to transmit control messages on existing open communications networks, research is needed to formulate the principles for construction and operation of networks that support time-critical message delivery in a context of interconnecting, multipurpose networks. If possible, such systems should be compatible and interoperable with other time-critical and non-time-critical applications and communications; in this event, they could be made compatible with a wide range of TCP/IP products. |
| 3. | Research is needed to develop services tailored for human- and machine-based information and resource browsing and searching, using knowledge-based assistance agents for semantic interpretations, translations, and relationships. The availability of such services will be essential to ensuring greater information infrastructure performance and efficiency. The higher-level network directory services of today will not meet the requirements of near-future applications.8 |
Inter-enterprise Integration
The notion of a supply chain (Figure 6.1) will be integral to manufacturing for the foreseeable future, and suppliers will be passing partially finished components or materials to those closer to the point of final assembly. Increasingly,
8 Some relevant work has been undertaken in the context of the X.500 directory standard, and further work on implementing that standard is necessary.
Page 129
FIGURE 6.1 The supply chain, a basic element of manufacturing that will become increasingly information intensive.
along with the movement of goods, information must move up and down the supply chain. Information technology can facilitate the passage of information within a manufacturing enterprise, but as importantly, it can enable different organizational structures and relationships among various elements in the supply chain, and networked infrastructures can extend well beyond a firm's boundaries. IT can be used to link value chains across firms and shift work involved in various functions between organizations. Already, electronic data interchange (EDI) connects companies to their suppliers, shippers, buyers, and even their banks. Providing a shared body of information and a common set of standardized forms, EDI transforms the ordering, invoicing, shipping, tracking, and customer service functions, among others, for a growing number of enterprises. Through these electronic linkages, IT enables companies to emphasize their core competencies by "outsourcing" activities that are mostly common across industries (e.g., payroll, purchasing, or accounting) to other providers that can perform these activities more efficiently. In addition, electronically linked firms can group and regroup as the need arises in alliances of convenience.
At present, much of the material-related cost of production stems from the
Page 130
purchasing, packaging, shipping, unpacking, and moving around of component materials. Better communication between supplier and customer enterprises would reduce such costs; although not new, EDI and other tools to support streamlined processes are obvious points of departure and candidates for supporting technology that will eventually culminate in the electronic communication of financial information and even electronic funds transfers, involving corporate-level systems and systems at third-party financial institutions.
Mechanisms and systems are required to support information session management. These include the means for connecting to and coordinating the delivery of information between multiple sources using multiple streams, as well as the mechanisms for collaboration using this information. In addition, tools and techniques are needed to facilitate real-time planning through the entire supply chain; in certain factories, this is already a partially achieved goal, but for a variety of reasons the techniques are not widely applicable across all manufacturers. Intelligent agents (computer automata) may be able to perform most of the hand shaking between customers and suppliers without requiring human intervention; if realized, these agents might provide human users with new, flexible, easy-to-use human-oriented protocols and standards that are as obvious as the current paper-based solutions. This level of analysis is generally impossible today; given the typical operation of manufacturing resource planning systems, it can take weeks. Research will be required in terms of software architectures, protocols for communication, and access and security mechanisms (see "Dependable Computing Systems" below). These needs are above and beyond the transaction-based processing and management that will clearly be required.
Non-Manufacturing-Specific Research
Given that manufacturing operations are typically large and complex, it is not surprising that the software for controlling and managing such operations is also large and complex. Research in many traditional areas of computer science, including information retrieval, software engineering, and reliable computing, as well as in collaboration technology, will be relevant to manufacturing. Work by computer scientists in these areas, but framed in a manufacturing context, will both enrich computer science and result in important benefits to manufacturing.9
9 An approach calling for research work at the intersection of computer science and engineering (CS&E) and other problem domains was contained in the CSTB report Computing the Future (CSTB, 1992). In particular, this report called for academic CS&E to increase its contact and intellectual interchange with other disciplines (of which manufacturing could be one); to increase traffic in CS&E-related knowledge and problems among academia, industry, and society at large; and to enhance the cross-fertilization of ideas in CS&E between theoretical underpinnings and experimental experience.
Page 131
Information Retrieval Systems
Manufacturing applications are data-intensive. Enormous amounts of data are needed to support manufacturing activities (e.g., to represent the millions of parts on a Boeing 777 or the millions of transactions that occur on a factory floor), and the relationships among the data are complex, sophisticated, and changeable. For example, a very large scale integrated chip component has a different description for each activity; functional, geometric, behavioral, temporal, and other views of the part must be supported. Any changes in one view must ripple through all of the other views and extend to invalidate or update usage of the component throughout a design process. The movement toward multimedia databases and applications compounds the problems presented by simple text or numerical databases.
Modern information storage and retrieval systems are designed to handle all aspects of data and information management except generating or collecting the data. These systems serve as the primary interface to data for the user through powerful query languages, for bulk storage devices through the file system, for the rest of the computer through the operating system, and for the rest of the world through communications networks. In each of these areas, advances have been rapid, but progress has only generated more ambitious goals. Information retrieval for manufacturing will stress existing technology and will require a major research effort. The following three problems illustrate the type of work that is needed:
| 1. | Incorporation of modeling and prototyping functions into a manufacturing database management system. The language used for data manipulation within the system should contain the basic syntax and semantics that describe products and processes; in other words, the product and process description language should be a sublanguage of the data manipulation language. However, current data manipulation languages were developed to handle financial and other types of data that can easily be cast into a relational database model. Manufacturing data, especially data related to geometry, are instead associative. Consequently, using current data manipulation languages for manufacturing data is counter-intuitive and difficult. Research is needed to develop the next generation of data manipulation languages, perhaps based on the object-oriented data model. |
| 2. | Maintenance of data consistency and integrity in a rapidly changing database that may be distributed. In the design phase, the database must serve not only as the primary repository of ongoing work but also as the principal medium of communication among the participants (who may be using simulation, analysis, and design tools, as well as manufacturing execution systems); these participants may also be geographically distributed |
Page 132
| across many time zones. Any disparity in the copies of the database seen by different users as changes are made will wreak havoc. In the processing phase, consistency is essential if real-time control of the process is to be achieved. A real-time production schedule is realistic only if the data about equipment status and work location are correct when the schedule is generated. Here the performance (i.e., speed) issue that is always present in database consistency becomes particularly difficult to resolve. Propagating changes throughout the entire system imposes additional constraints on real-time performance. | |
| 3. | Information browsing and searching in a distributed manufacturing environment. Vendors and suppliers connected through the National Information Infrastructure will want to make information about their products available electronically to customers that may not know about them. Design and production engineers will need to browse various information archives in search of a part or a process that they may need. Thus, while network tools such as gophers, MOSAIC, and Wide-Area Information Search systems are promising starts to the general problem, research is needed to develop more advanced automated tools that accept information on needed parts or processes as input, search or make appropriate inquiries at appropriate sites on the NII, and return to the user a list of those sites and the information found there. Such an effort would also depend on new indexing schemes and data representations that would allow semantically driven searches of both text and nontext information. |
Software Engineering
In view of the experiences of the major manufacturers represented by members of the committee and those who briefed the committee, the current state of the art in software engineering is barely adequate to meet current manufacturing needs. The software problem has a direct impact on a manufacturer's ability to respond quickly to a changing marketplace with new products. The reason is that new product development often calls for rapid responses and agile manufacturing, which themselves require advanced information technologies, open systems concepts, and component reuse.
The following list describes some of the most important areas of software engineering research for manufacturing:
| • | Tools to support the software engineering process in manufacturing. Automating complex, intelligent manufacturing systems requires both solid engineering life-cycle approaches and supporting tools. Unfortunately, in many cases, tool development lags significantly behind the use of new methodologies, as in the case of tools to support the recent component-based architecture |
Page 133
| life-cycle approaches. Companies using such approaches for component development and reuse are suffering from the lack of useful tools for system analysis, requirements management, design, configuration and integration, and simulation. Until such tools become available, the building of large, complex manufacturing systems will remain a costly exercise. For example, more general tools are needed that are not constrained by the limitations of specific programming languages, as well as tools that have aspects of knowledge-based collaboration software. | |
| • | Programming paradigms that enable manufacturing personnel, rather than software development experts, to develop and change application systems for the shop floor or the design laboratory. This capability will require better methods of developing software and better human-machine interfaces to enable domain-specific software specification. Object-oriented programming is a particular paradigm of interest. |
| • | Support for faster development of easier-to-use and more effective systems. Simplification of designs, operation, and maintenance is desired, as are increased predictability of systems, self-healing systems, and system extensibility. Research is needed in system optimization and enhanced system operation and maintenance; better capabilities for rapid prototyping are a particular need. Visualization and human-computer interaction techniques will be key. |
| • | Accommodation of legacy systems. Manufacturers have a great deal of investment that they will be loath to give up even when new and improved technology is available. Thus, research on how to facilitate transitions to new technology is necessary. One direction for such research is the development of techniques for encapsulation of legacy systems and of mediator support. |
| • | Tools for systems analysis. Although there are many different analysis methodologies, such as responsibility-driven, data-driven, and activity-based methodologies, system analysis remains more an art than a science. How do we compare types of analyses? Are some techniques simply better than others, or do different types of system development or applications dictate a choice? Are the techniques sufficient, or are new ones needed? Research is needed to help answer these questions. |
| • | Better metrics. A number of metrics have been developed for the testing phase of the engineering process to measure dimensions of system performance such as speed and timing. There are no metrics that specify how well a problem in manufacturing is understood and formulated. What are the |
Page 134
| metrics for analysis, and how do we measure our successes and failures? When do we know that analysis is complete or at least good enough? Can these questions be addressed with metrics? Given the frequent need for rapid system development, metrics specifying temporal aspects of analysis should be considered. Are there temporal metrics for analysis techniques (e.g., state-transition diagrams, object models, event-stimulus diagrams) that could be related to system characteristics? | |
| • | Software reuse. The ability to reuse software and system artifacts and results from associated analyses in the development of new system or applications would leverage previous investments in expertise, effort, money, and time. Examples of software artifacts from object-oriented analysis and design include object models (including representations of data and behavior) and object interface definitions. One type of research supporting reusability could lead to reference architectures that cut across manufacturing domains. For example, there could be families of reference architectures for continuous processes and for discrete processes. Goals would include minimizing the number of reference architectures for which third-party suppliers must develop software or systems and increasing the ability to leverage previous expertise across projects. |
| • | Better representations of spatial and temporal dimensions in software. Software representation currently focuses on declarative knowledge. Research is needed to achieve representations that depict both spatial and temporal aspects of associated data, enriching a system's repository of knowledge and facilitating the visualization, design, specification, monitoring, and analysis of manufacturing processes. Multimedia visualization of these new representations may include ''walking through" models or providing simulations of system behaviors in one window while the activated models employed are presented in another window. Such elements would contribute to the realization of virtual factories, which involve enhanced modeling and simulation of processes and products. New programming interfaces that capture the spatial and physical abstractions of manufacturing are essential to allow end users to program their own processes and work flows. |
Dependable Computing Systems
Manufacturing plants require continuous operation, creating a need for dependable computing systems. Better technology to support "hot swaps" of software (i.e., changing software without removing the system from operation), continuous availability of on-line services, and fault-tolerant hardware and software are among the technologies needed for dependable manufacturing systems.
Manufacturing provides an application arena for a wide variety of research
Page 135
relating to increased system security and trustworthiness, including access control and authentication. The need to determine and verify the user of a computer system or network is becoming increasingly important as a way of preserving system and data integrity, ensuring that sensitive data and systems are accessed only by those who are authorized, and ensuring the highest levels of system availability. These concerns affect both intra-enterprise and inter-enterprise communications.10 The manufacturing environment, especially the factory environment, calls for economical and robust technology to address these needs.
Finally, as manufacturing tools become more autonomous, it becomes imperative to ensure the correctness of controlling software, as control software with errors may be a major contributor to factory downtime due to physical damage.
Collaboration Technology and Computer-supported Cooperative Work
The trend toward organizing workers of all kinds into teams with significant levels of decision-making authority gives rise to a need for technology to support collaborative activity.11 For example, intelligent systems are needed to support collaborative efforts in the design of complex products; they can also facilitate collaboration among factory and other, nonproduction personnel. Research needs include information technology to support empowered work teams of various kinds of personnel and tools for total quality management.
Research relating to technical tools for computer-supported cooperative work (including software, user interfaces, and supporting hardware) should be complemented by research examining relevant aspects of human behavior, education and training requirements, and other similar issues, to ensure both that optimal tools are developed and that they can be used easily.
10 Electronic contract documents and order qualifications, which are envisaged as part of our future enterprises, will require methods for identification, authentication, and non-repudiation.
11 Aside from the emerging, largely message-based groupware products, today's tools are aimed at supporting individual professional performers rather than collaborating teams.
