تکامل یک سیستم تولید سلولی - یک مطالعه موردی طولی

This paper describes the evolution of a cellular manufacturing system in a medium-sized company over a 13-year period. The objective of this paper is to analyze the arguments that gave rise to the nearly continuous readjustment of the design of the cellular manufacturing system of this company and the direction in which these adjustments took place. The study indicates that two interrelated factors played an important role in the decision to change the system: the market and manufacturing technology. Analysis of these factors offers important insights into the aspects that need to be taken into account in cell formation. It is argued that a cellular system should reflect market characteristics. New technology, furthermore, demands specialized cells, producing in a multi-shift situation. These two developments point in the direction of market-oriented, reasonably sized, functionally organized manufacturing units. It is argued that market developments, new manufacturing technology and modern production control systems will probably constrain the application area of cellular manufacturing.

A cellular manufacturing (CM) system is defined here as the grouping of workers and machines into relatively independent cells, which are responsible for the complete manufacturing of a set of part types. Many firms have adopted CM as a strategy for improving performance. Case studies (e.g., [1], [2], [3], [4], [5], [6], [7], [8], [9], [10], [11], [12], [13] and [14]) and survey articles (e.g., [15], [16] and [17]) show important advantages of CM such as shorter throughput times, better product quality, and lower material handling costs. An important flaw of these studies is that they consider the success of CM only in the short term. Cell formation is likely one of the most researched topics in the field of cellular manufacturing [18]. A large number of cell formation techniques have been developed through the years. All techniques require nearly the same information: number and types of machines, volume, routings, and processing times of part types [19], [20] and [21]. A topic related to cell formation that has gained little attention in the literature concerns the stability of a CM system through the years. Some authors have investigated the effect of volume and mix changes on the performance of a cellular manufacturing system (e.g., [22] and [23]) and conclude that these changes may necessitate the redesign of a CM system. Others indicate the inferior ability of a CM system to cope with volume and mix changes (e.g., [24] and [25]). Marsh et al. [26], however, found in an empirical study that major changes in the design of a cell, such as its termination, are relatively infrequent because management has several alternative, less radical and less expensive interventions to cope with declining cell performance. These actions to cope with altering circumstances include, for instance, working overtime, cross-training, set-up time reduction, alternative routings, outsourcing, anticipating inventory, adding workers and allowing intercell movements. March et al. [26] report that 26% of the 185 cells they studied in 15 metal machining firms were dismantled in the average six-year existence of the CM system. Their study is focused on individual cells and does not deal with the impact of cell changes on the entire CM system. The study presented in this paper concerns a longitudinal case study covering 13 years of experience with a CM system. Our main objective is to gain insight to the reasons why parts of the CM system were redesigned in the course of this time span. Furthermore, we focus on the impact of redesign decisions on the entire CM system. More specifically, we discuss three factors which may drive a redesign of the CM system: (i) the market, (ii) manufacturing technology, and (iii) managerial choices. The market determines the demand for products of each cell and is therefore an important factor affecting the need for a cell's resources. New manufacturing technology is another factor that may provide an argument for redesign. A new and advanced technical system may be able to, for example, do more work, manage a larger variety of processes, cross the boundaries of cells or even make cells redundant. Finally, the functioning of a CM system is the continuous concern of management, and management policy may be an important determinant in the redesign of a CM system. These three factors are interrelated. Major elements of management strategy concern choices with respect to markets and technology investment. On the other hand, changes in markets and technological development may shape and mold strategy and force management to intervene to preserve or enhance performance. Interventions to improve performance may or may not involve the redesign of cells. Besides coping strategies that concern design issues, management may opt for interventions that focus on altering processes or environmental conditions which may improve cell performance. Although we suppose that management will have decision latitude, the organizational and physical environment of a CM system will direct the choices made by management. In this longitudinal case study we follow the gradual change of a CM system and address all the major actions taken by management to keep or improve the performance of the system. Our main interest is the stability of the CM structure and, therefore, our primary concern is changes to the design of the CM system. However, to gain a full understanding of the modifications and developments, we will also highlight non-design interventions. Section 2 of this paper gives an introduction to the firm in which we conducted our longitudinal study of the evolution of a CM system. Section 3 presents the major changes that took place in the cell design during the period 1987–2000. Section 4 relates these modifications to the factors presented above. In the final section, we discuss the impact of our findings in a more general way. In this section we also provide suggestions for further research.

The basic advantage of cellular manufacturing concerns its ability to connect all kinds of sequential relationships in the firm. Different manufacturing steps are grouped through cellular manufacturing. Moreover, the sequential support functions (e.g., process planning and production planning) can be more easily connected in the case of cellular manufacturing. However, these advantages have a negative side. Our case study indicates that a CM system is vulnerable to changes in markets, while functional layouts are less sensitive to such changes. Moreover, technological innovations – especially when they bring along complex tasks and cost a great deal – necessitate exploration of pooling aspects, such as the exchange of work among identical machines and the grouping of similarly specialized employees [37]. In addition, modern multifunctional machines, able to process parts completely, do not require cellular manufacturing systems to any great degree. There is no material flow to be dealt with by a cellular manufacturing system. In order to apply new machinery more efficiently, it may even be advantageous to cluster identical machines to profit from pooling synergy (multi-machine operating, same control software, etc.). The complexity of new machines is interesting enough for operators so, as we earlier argued, they will likely not perceive the assignment of control tasks not directly related to ‘their’ equipment as contributing to job enrichment. Generally, a major advantage of CM is the substantial reduction of intergroup relations, which makes the control structure of the work floor rather transparent and simple. Such a structure with rather independent cells makes it easy to create opportunities for local decision-making, trace problems and improve feedback. Modern production control systems, however, are able to deal properly with complex flows through various cells. This may further reduce the advantages of cellular manufacturing systems. Although there have been many changes, there are still elements of CM present that contribute to the current performance of the firm. The developments in HAT can be summarized as a move from a strict CM system to a mini-company concept. These mini-companies are functionally specialized and focused on a certain market segment. Elements, such as a rather flat organizational structure, the decentralization of many support functions to the clusters or to the cells, the feedback of performance and a positive attitude towards continuous improvement among workers, all part of the heritage of CM, still contribute to the healthy functioning of the present structure. The main business of the firm we studied is parts manufacturing in the mechanical engineering industry. Therefore, we believe that our conclusions will be especially valid in such a production environment. Although the following conclusion needs further empirical support, we believe our findings may be generalized to the statement that a CM system is especially unstable in dynamic markets and in production environments with rapidly advancing manufacturing and control technology. This implies that a CM system will be more durable and profitable in relatively stable markets with environments less dominated by advanced technology systems. However, even in such cases the advantages and disadvantages of a particular manufacturing design will shift depending on changes in the contingencies of organizations. The design process is not a static process with a well-defined result. On the contrary, a design is never finished and needs continuous improvements. The insights that follow from our 13 years of observation raise several questions that need to be addressed in future research. Most important are: (1) Do unplanned events play a significant role in the design and development of a CM system? As can be seen, some of the decisions on organizational and design changes were made after some indirectly-related incidents (e.g., OIC1, implemented after a manager left the firm). Noting the importance of indirectly-related incidents, an interesting empirical research issue concerns the impact of these types of incidents on manufacturing. (2) Does investment in advanced and expensive equipment lead to fewer but more specialized workers, a need for high machine utilization levels, more people working in a shift system, the integration of processing steps, and the pooling of similar capacities? This question concerns the impact of integrated manufacturing technology on the efficiency and effectiveness of a Cellular Manufacturing system. Integrated manufacturing systems reduce the number of manufacturing steps, which decreases the importance of short traveling distances and, therefore, reduces the advantage of CM systems. Furthermore, integrated manufacturing systems may demand a specialized workforce, which contradicts the multifunctionality condition of cellular manufacturing. On the other hand, integrated manufacturing technology may bring about enriched work situations, where operators are able to schedule the jobs to be produced on the integrated machine, the maintenance of the machine, and the quality control. These latter arrangements are in line with the cellular manufacturing concept. (3) Do perceptions of quality of working life change over time because of changes in technology and CM design? This question deals with the fundamentals of the quality of working life. This study suggests that quality of working life is more a social construct that acquires meaning and content depending on work, workers and context characteristics than an ‘objective’ concept that is independent of worker's perception [33]. More specifically, how do the perceptions of the quality of working life relate to the characteristics of manufacturing technology? Another important issue in the area of quality of working life is the impact of the design (e.g., CM versus functional) on interdependencies between team members, the urge for social-interaction and teamwork. The same is true for the effect of the shift system on these interpersonal aspects of quality of working life. (4) Do advances in information and production control systems reduce the advantages of the transparency of a CM design? Production orders can be easily traced by means of modern information and production control systems, which are based upon the premise that the availability of information is of major importance to gain an optimal control situation. Cell manufacturing (or better socio-technical design theory) assumes that major knowledge about the status of the manufacturing system is stored in the minds of workers and that control loops must be located as low as possible in the organization. It is questionable as to what extent advanced information systems match socio-technical design principles and support the functioning of CM environments. (5) Do CM systems evolve in the direction of a specialized, market-oriented production organization? This question focuses on the link between the extended enterprise concept, the mini-company concept, and the evolution of CM systems in firms. The extended enterprise concept suggests that firms need to focus on particular competencies, being the only way to remain competitive. All other, non-focused, activities may be subcontracted [38]. The firm must have close links with the entire supply chain (i.e., customers and suppliers). The extended enterprise concept likely supports the existence of a mini-company concept in which there is a clear customer focus. This development supports the development of a cellular manufacturing system towards more market-oriented cells. This statement indicates a need for market-oriented CM design methods. Cell formation is oriented mainly towards gaining autonomous manufacturing cells with respect to the routings of part types; the main objective is to minimize the number of intercell movements. The part types assigned to one cell, however, may refer to various markets. It may be useful if cell formation methods encapsulate some market considerations. This may help to adapt the cells towards the needs of particular markets. (6) Is there a variety of CM designs conceivable in a job shop, MTO environment? The initial CM design of HAT was a partly fractal CM design (see Section 2). The last CM design was a more functionally-oriented CM design. In a fractal design, cells are less unique (all cells are equal in a pure fractal design) than in a functional design. A functional ‘cell’ (unit) is much more focused on specific operations, while a fractal cell can handle a wide range of operations. Moreover, in a functional design, units are much more interdependent than cells in fractal design are. Several CM concepts are mentioned in literature, such as fractal manufacturing systems, holonic systems, virtual manufacturing cells, and dynamic cells. Recently, studies concerning these items have appeared [39] and [40]. It would be interesting to study the performance impact of the various concepts and to specify the environments in which these concepts are best suited.