COMPUTER INTEGRATED MANUFACTURING:CIM DEFINITIONS AND CONCEPTS

INTRODUCTION

Joseph Harrington introduced the concept of computer integrated manufacturing (CIM) in 1979 (Har- rington 1979). Not until about 1984 did the potential benefits the concept promised begin to be appreciated. Since 1984, thousands of articles have been published on the subject. Thanks to the contributions of researchers and practitioners from industries, CIM has become a very challenging and fruitful research area. Researchers from different disciplines have contributed their own perspec- tives on CIM. They have used their knowledge to solve different problems in industry practice and contributed to the development of CIM methodologies and theories.

CIM DEFINITIONS AND CONCEPTS

Manufacturing Environment

From the name ‘‘computer integrated manufacturing,’’ it can be seen that the application area of CIM is manufacturing. Manufacturing companies today face intense market competition, and are experi- encing major changes with respect to resources, markets, manufacturing processes, and product strat- egies. Manufacturing companies must respond to the rapidly changing market and the new technologies being implemented by their competitors. Furthermore, manufacturing, which has been treated as an outcast by corporate planning and strategy, must become directly involved in these critical long-range decisions. Manufacturing can indeed be a formidable competitive weapon, but only if we plan for it and provide the necessary tools and technologies (Buffa 1984).

Besides the traditional competitive requirements of low cost and high quality, competitive pressure for today’s manufacturing companies means more complex products, shorter product life cycles, shorter delivery time, more customized products, and fewer skilled workers. The importance of these elements varies among industries and even among companies in the same industry, depending on each company’s strategy.

Today’s products are becoming much more complex and difficult to design and manufacture. One example is the automobile, which is becoming more complex, with computer-controlled ignition, braking, and maintenance systems. To avoid long design times for the more complex products, companies should develop tools and use new technologies, such as concurrent engineering, and at the same time improve their design and manufacturing processes.

Higher quality is the basic demand of customers, who want their money’s worth for the products they buy. This applies to both consumers and industrial customers. Improved quality can be achieved through better design and better quality control in the manufacturing operation. Besides demanding higher quality, customers are not satisfied with the basic products with no options. There is a com- petitive advantage in having a broad product line with many versions, or a few basic models that can be customized. A brand new concept in manufacturing is to involve users in the product design. With the aid of design tools or a modeling box, the company allows the users to design the products in their own favor.

In the past, once a product was designed, it had a long life over which to recover its development costs. Today many products, especially high-technology products, have a relatively short life cycle. This change has two implications. First, companies must design products and get them to the market faster. Second, a shorter product life provides less time over which to recover the development costs. Companies should therefore use new technologies to reduce both time and cost in product design. Concurrent engineering is one method for improving product design efficiency and reducing product costs. Another new paradigm is represented by agile manufacturing, in which the cost and risks of new product development are distributed to partners and benefits are shared among the partners. This requires changes to or reengineering of traditional organization structures.

Several demographic trends are seriously affecting manufacturing employment. The education level and expectations of people are changing. Fewer new workers are interested in manufacturing jobs, especially the unskilled and semiskilled ones. The lack of new employees for the skilled jobs that are essential for a factory is even more critical. On the other hand, many people may not have sufficient education to be qualified for these jobs (Bray 1988).

To win in the global market, manufacturing companies should improve their competitive ability. Key elements include creative new products, higher quality, better service, greater agility, and low environmental pollution. Creative new products are of vital importance to companies in the current ‘‘knowledge economy.’’

Figure 1 presents a market change graph. From this figure, it can be seen that the numbers for lot size and repetitive order are decreasing, product life cycle is shortening, and product variety is increasing rapidly.

End users or customers always need new products with advancements in function, operation, and energy consumption. The company can receive greater benefits through new products. A manufac- turing company without new products has little chance of surviving in the future market. Better services are needed by any kind of company. However, for manufacturing companies, better service means delivering products fast, making products easy to use, and satisfying customer needs with low prices and rapid response to customer maintenance requests.

Features of a General Manufacturing System

The manufacturing company is a complex, dynamic, and stochastic entity consisting of a number of semi-independent subsystems interacting and intercommunicating in an attempt to make the overall system function profitably. The complexity comes from the heterogeneous environment (both hard- ware and software), huge quantity of data, and the uncertain external environment. The complex structure of the system and the complex relationships between the interacting semi-autonomous sub- systems are also factors making the system more complicated.

Year

A simple model of a manufacturing system can be a black box that takes as input materials, energy, and information and gives as output products. The internal details of the manufacturing system depend on the particular industry involved, but the key feature common to all manufacturing orga- nizations is that the system processes both materials and information. General manufacturing systems can be decomposed into seven levels of decision hierarchies (Rogers et al. 1992) (Figure 2). Decisions at the upper levels are made at less frequent intervals (but have implications for longer periods into the future) and are made on the basis of more abstract (and slower to change) information on the state of the system. Decisions at the lower levels are made more frequently using much more detailed information on the state of the system.

Three kinds of decisions should be made for any manufacturing company: (1) what kinds of products to make, (2) what resources will be needed to make the products, and (3) how to control the manufacturing systems. These decisions cannot be made separately. If the company wishes to make a decision at a certain level, such as at the business level, it should also get access to the information at other levels. In the whole process of decision making, the core concept is integration. This is the fundamental requirement for the research and development of computer integrated man- ufacturing.

CIM Definitions

There are many definitions for CIM, emphasizing different aspects of it as a philosophy, a strategic tool, a process, an organizational structure, a network of computer systems, or a stepwise integration of subsystems. These different definitions have been proposed by researchers working in different areas at different times from different viewpoints. Since the concept of CIM was put forward in 1973, it has been enriched by the contributions of many researchers and practitioners. One earlier definition of CIM is ‘‘the concept of a totally automated factory in which all manufacturing processes are integrated and controlled by a CAD / CAM system. CIM enables production planners and sched- ules, shopfloor foremen, and accountants to use the same database as product designers and engi- neers’’ (Kochan and Cowan 1986). This definition does not put much emphasis on the role of information.

Another definition is given by Digital Equipment Corporation (DEC): ‘‘CIM is the application of computer science technology to the enterprise of manufacturing in order to provide the right infor- mation to the right place at the right time, which enables the achievement of its product, process and business goals’’ (Ayres 1991). This definition points out the importance of information in manufac- turing enterprise, but unfortunately it does not give much emphasis to the very important concept of integration.

ther definitions have pointed out that CIM is a philosophy in operating a manufacturing com- pany. For example: ‘‘CIM is an operating philosophy aiming at greater efficiency across the whole cycle of product design, manufacturing, and marketing, thereby improving quality, productivity, and competitiveness’’ (Greenwood 1989).

To stress the importance of integration, the Computer and Automation Systems Association of the Society of Manufacturing Engineers gives the following definition: ‘‘CIM is the integration of the total manufacturing enterprise through the use of integrated systems and data communications coupled with new managerial philosophies that improve organizational and personnel efficiency’’ (Singh 1996).

CIM does not mean replacing people with machines or computers so as to create a totally auto- matic business and manufacturing processes. It is not necessary to build a fully automatic factory in order to implement a CIM system. It is especially unwise to put a huge investment into purchasing highly automation-flexible manufacturing systems to improve manufacturing standards if the bottle- neck in the company’s competitiveness is not in this area. In the current situation, the design standards for creative and customized products are more important than production ability in winning the market competition.

The importance of human factors should be emphasized. Humans play a very important role in CIM design, implementation, and operation. Although computer applications and artificial intelligence technologies have made much progress, even in the future, computers will not replace people. To stress the importance of the role of humans, the idea of human-centered CIM has been proposed.

Two views of CIM can be drawn from these definitions: the system view and the information view. The system view looks at all the activities of a company. The different functions and activities cannot be analyzed and improved separately. The company can operate in an efficient and profitable way only if these different functions and activities are running in an integrated and coordinated environment and are optimized in a global system range. The SME CIM wheel (Figure 3) provides a clear portrayal of relationships among all parts of an enterprise. It illustrates a three-layered inte- gration structure of an enterprise.

The outer layer represents general management and human resources management. The middle layer has three process segments: product and process definition, manufacturing planning and control, and factory automation. These segments represent all the activities in the design and manufacturing phases of a product life cycle, from concept to assembly. The center of the wheel represents the third layer, which includes information resources management and the common database.

The other view of CIM is the information view. As stated in the definition given by Digital Equipment Corporation, the objective of CIM implementation is to enable the right information to be sent to the right person at the right time. The information system plays a vital role in the operation of CIM. Although many kinds of activities are involved in managing a manufacturing company, each activity has a different function in business management and production control. The associated function unit for the information system of CIM normally can be classified into three kinds of tasks: information collection, information processing, and information transfer.

Information collection is the basic function of an information system. The information collected forms the basis of decision making at different levels from business management to device control. There are many methods of information collection, depending on the information sources and tech- nologies used. Device sensors may provide data regarding device status; barcode scanners may pro- vide data about the production status of online products; and form scanners and database table view interfaces may provide data about order, raw material purchasing, and user requirements. Some data may also come from e-mail systems. The data collected can be stored in different data formats and different repositories.

Information processing is closely related to the business functions of a company. Business func- tions range from strategy planning, process planning, product design, warehouse management, and material supply to production management and control. The upper-stream process data are processed by algorithms or human intervention and the instructions produced are used for the downstream process. In data processing, different decisions will be made. The decisions can be used to optimize the production processes or satisfy user requirements such as delivery time and quality requirements.

Information transfer between different function units has three main functions: data output from application software in a certain data format to a certain kind of data repository; data format trans- formation, and data transfer from one application to another application within the same computer or in a network environment.

Integration: The Core of CIM

The core of CIM is usually seen to be integration. In our opinion, computer technology is the basis of CIM, manufacturing is the aim, and integration is the key technology. Why should integration be considered the core of CIM? This can be seen from different aspects. The system view of CIM was described above. System means the whole company, including people, business, and technology. In order to form a coordinated system, these elements must be integrated. Material flow, information flow, and capital flow must also be integrated. Although those aims seem clear, the technology for realizing this integration is far from mature.

CIMOSA (Esprit Consortium AMICE 1993) identifies enterprise integration as an ongoing pro- cess. Enterprises will evolve over time according to both internal needs and external challenges and opportunities. The level of integration should remain a managerial decision and should be open to change over a period of time. Hence, one may find a set of tightly coupled systems in one part of a CIM enterprise and in another a set of loosely coupled systems according to choices made by the enterprise. The implementation of multivendor systems in terms of both hardware and software and easy reconfiguration requires the provision of standard interfaces. To solve the many problems of the industry, integration has to proceed on more than one operational aspect. The AMICE (European Computer Integrated Manufacturing Architecture) project identifies three levels of integration cov- ering physical systems, application and business integration (see Figure 4).

Business integration is concerned with the integration of those functions that manage, control, and monitor business processes. It provides supervisory control of the operational processes and coordinates the day-to-day execution of activities at the application level.

Application integration is concerned with the control and integration of applications. Integration at this level means providing a sufficient information technology infrastructure to permit the system wide access to all relevant information regardless of where the data reside.

Physical system integration is concerned with the interconnection of manufacturing automation and data-processing facilities to permit interchange of information between the so-called islands of automation (intersystem communications). The interconnection of physical systems was the first integration requirement to be recognized and fulfilled.

Even when business integration has been achieved at a given time, business opportunities, new technologies, and modified legislation will make integration a vision rather than an achievable goal. However, this vision will drive the management of the required changes in the enterprise operation.

The classification of integration can also be given in another method, which is different from that given by CIMOSA. Regarding integration objectives and methods, integration can be classified as information integration, process integration, and enterprise-wide integration.

Information integration enables data to be shared between different applications. Transparent data access and data consistency maintenance under heterogeneous computing environments is the aim of information integration. Information integration needs the support of communication systems, data- representation standards, and data-transfer interfaces. Communication systems provide a data transfer mechanism and channel between applications located at different computer nodes. Data-representation standards serve as common structures for data used by different applications. Data-transfer interfaces are used to transfer data from one application to another. They fulfill two kinds of functions: data format transfer (from application-specified data structure to common structure and vice versa) and data transfer from application to interface module and vice versa. The traditional information- integration methods include database data integration, file integration, and compound data integration. The most efficient support tool for information integration is the integration platform (Fan and Wu 1997).

Process integration is concerned with the collaboration between different applications in order to fulfill business functions, such as product design and process control. The need to implement process integration arises from companies’ pursuit of shorter product design time, higher product quality, shorter delivery time, and high business process efficiency. Business process reengineering (BPR) (Jacobson 1995) and concurrent engineering (CE) (Prasad 1996) have promoted the research and application of process integration. Business process modeling, business process simulation, and busi- ness process execution are three important research topics related to process integration.

A number of methods can be used in modeling business processes: CIMOSA business process modeling, IDEF3 (Mayer et al. 1992), Petri nets (Zhou 1995), event driven process chain (Keller 1995), and workflow (Georgakopoulos et al. 1995). The modeling objective is to define the activities within a business process and the relationships between these activities. The activity is a basic func-

tion unit within a business process. The control and data flow between these activities form the business process, which fulfils the business task of a company. Optimizing the flow path and short- ening the flow time can help the company increase its working efficiency and reduce cost.

The third type of integration is called enterprise-wide integration. With the advent of agile man- ufacturing, virtual organization is ever more important than before. In agile manufacturing mode, a number of companies collaborate in a virtual company to gain new opportunities in the market. Enterprise-wide integration is required to enhance the exchange of information between the compa- nies. The success of virtual organizations is predicated on the empowerment of people within the enterprise with the aid of computer technology including communication networks, database man- agement systems, and groupware. These allow team members in the virtual organization to make more effective and faster group decisions. Such interaction lays the foundation for enterprise-wide integration, encompassing various plants and offices of an enterprise, possibly located in different cities, as well as customers and suppliers worldwide. Therefore, enterprise-wide integration is much broader than factory automation integration. It is the integration of people, technology, and the busi- ness processes throughout the enterprise.

Enterprise-wide integration is required to ensure that all the technical and administrative units can work in unison. This requires a great deal of information about a large number of activities, from product conception through manufacturing, customer delivery, and in-field support. All these life- cycle steps require a large volume of data. The transformation process from one stage to another yields volumes of new data. Furthermore, many of these design, manufacturing, distribution, and service activities responsible for generating and using volumes of data are scattered across a wide spectrum of physical locations. The information is generated by a diverse set of highly specialized software tools on heterogeneous computing hardware systems. Often, incompatible storage media with divergent data structures and formats are used for data storage. This is due to the peculiarities of the tools and systems that generate data without any regard to the needs of the tools or systems that will eventually use the data.

The main idea of enterprise-wide integration is the integration of all the processes necessary for meeting the enterprise goals. Three major tools for integration that are required for overcoming the local and structural peculiarities of an enterprise’s data processing applications are network com- munications, database management systems, and groupware. A number of methods for enterprise- wide integration have been proposed; supply chain management, global manufacturing, and a virtual information system supporting dynamic collaboration of companies. The Web and CORBA (Otte et al. 1996) technologies are playing important roles in the realization of enterprise-wide integration.