یک روش مشکل محور به رابط تجزیه و تحلیل استراتژی ساخت و طراحی سیستم های تولید

This paper introduces an interface between manufacturing strategy analysis (MSA) and manufacturing system design (MSD). MSA methods are not accurate enough to assess the manufacturing design choices. MSD requires functional-oriented scopes, and not only strategic initiatives resulting from MSA. That is why MSA and MSD must be interfaced. The proposed interface consists in the Model of Operational Manufacturing System, the evolution class framework, the model of problem and the problem handling procedure using these models. They point out the proposed detailed evolution classes (domains which have to be improved) adapted to a specific manufacturing system. Finally, the proposal is illustrated within an industrial case study, which underlines its efficiency.

This paper tackles the domain of the design process management of manufacturing systems (MS). According to Austin, Newton, Steele, and Waskett (2002), the early phases of MS design projects seem to be less studied. Even though the decisions made during this period have the most far-reaching effects on the remainder of the project. Indeed, before the design begins, user requirements must be consolidated as proved by Chen, Vallespir, and Dougmeints (1997). Moreover, as stated by Barad and Gien (2001), enterprises need support to define their specific technological and organisational needs and then to find the right way to fulfil these needs. The objective of this paper is to develop models and their exploitation procedure. They aim at enabling the identification of the main scopes of the project, i.e. the definition of domains where evolutions will be efficient to contribute to manufacturing process improvement. Several fields of research are concerned with this topic. The first field is manufacturing strategy analysis (MSA). Carrie, Durrani, Forbes, and Martowidjo (2000) define the results of MSA, as follows: • A technology portfolio (choices of alternative processes), completed by relative parameters (capacity, size, timing, location, investments). • The manufacturing infrastructure required to support production: function support, manufacturing planning and control systems, manufacturing system engineering, quality assurance and control, clerical procedures, work structuring, organisational structure,… The state of the art in MSA provides classes of methods to help a company analysing its products, market and operations. They allow to set objectives for relevant areas of concern (Wu and Ellis (2000)). In this domain, Chan and Spedding, 2003 and Zantek et al., 2002 propose some analyses based on Statistical Process Control. From data retrieved from manufacturing system monitoring, lacks of quality are identified and located in the system. These metrics oriented methods are completed by user-oriented methods. Thus, Barad and Gien, 2001 and Chen et al., 1997 interpret interviews with the stakeholders of the manufacturing system, in order to define the areas of concerns. No process monitoring is required for these approaches. Both methods result in a pattern of actions or in a set of areas of concern. The second field of research concerned by our topic is manufacturing system design (MSD). It aims at determining the best structure of a manufacturing system, in order to provide the skills required to support strategic objectives. This must be achieved within the allocated resources, and satisfy other constraints (Wu and Ellis (2000)). Some generic design methods exist, like the production system design of Cochran, Eversheim, Kubin, and Sesterhenn (2000) for implementing lean manufacturing, or GRAI–GIM methodology (GRAI Integrated Methodology) proposed by Doumeingts (1984). There are numerous applications of these methods in the literature. Among these, Kulak, Durmusoglu, and Tufekci (2005) propose a complete Cellular manufacturing system design. Such a choice of shop floor evolution should be based on a MSA, according to Carrie et al. (2000). Otherwise, as proved by Zantek et al. (2002), ignoring quality linkages can lead manufacturers to make suboptimal investments in quality improvement. However, MSA methods are not accurate enough to assess this manufacturing design choice. MSD requires functional-oriented scopes, and not only strategic initiatives resulting from MSA. That is why MSA and MSD must be interfaced. Existing design methods propose this interface through requirements consolidation. None of them provides formal models. Formal models are required to satisfy the consistency between strategic, organisational and operational viewpoints. This paper introduces an interface between MSA and MSD by suggesting a detailed manufacturing system evolution class framework, including the formal models and the exploitation procedure. They provide an accurate and consistent framework for setting MSD scopes. The next section introduces the generic model of manufacturing system (MOMS), the related taxonomy of evolution classes and the model of problem. They allow to classify the potential evolutions of a given manufacturing system. The third section details the computer based exploitation procedure of these models. The treatment of the problems through this procedure is the core of the proposed interface. This interface leads to the consolidated design requirements and their association to the particular evolution classes. So, our approach is based on a user-oriented problem analysis. At last, an industrial case study illustrates how problems encountered in a shop floor (manufacture of electrical gear-motors) have been processed to specify the objectives of the ‘High Speed Machining Implementation’ project. The case study outlines the use of the computer in the problem driven approach.

This paper introduces an interface between manufacturing strategy analysis and manufacturing system design, as consoli- dated requirements are expected. This interface consists in the Model of Operational Manufacturing System, the evolution class taxonomy, the model of problem, and the related exploitation pro- cedure. They point out the detailed evolution classes (domains which have to be improved), adapted to a specific manufacturing system. The method (including the models and the procedure) has been applied to industrial projects, which one of them is par- tially developed in this paper. The use of the computer supports the non-sequential design process. The potential improvements stand in the development of a specific computer based interface during the interaction with the workgroup. It should ease the com- mon understanding of the numerous concepts of the models. On the long range, the use of the computer may be extended to the problem handling (automatic association of problem, problem im- age and taxonomy), in order to enlarge the project scales.