An expert system to improve the functioning of the clothing industry: A development methodology

摘要

Purpose - Today the garment industry in Mexico is vulnerable to complex problems. This type of industry is subject to influences which over time, modify the perceptions of those involved in the design process due to the fact that they face problems that have both objective and subjective characteristics. In this study the authors used interviews, direct observation methodology, and theoretical argumentation to obtain the experts knowledge as they describe the problems that arise in the process of garment design for Mexican markets. The purpose of this paper is to generate a methodology so that the expert in this field will become highly specialized, resulting in heightened abilities and reinforcing them with the methodology of soft systems and the design management model (DMM). The results suggest that they are applicable to any area of design. Design/methodology/approach - The success or failure of expert system (ES) depends directly on the acquisition of knowledge (Mendez-Gurrola, 2007, 2012), to accomplish that, three large groups of techniques are used to extract that knowledge: manual, semiautomatic, and automatic. Within the group of manual techniques, interviews, protocol analysis, questionnaires, direct on site observation, and the extraction of closed curves are emphasized. This technique is chosen for its ability to extract the particular type of knowledge being sought. The knowledge of the dominion expert in the design process applied to the clothing industry in Mexico is based on processes and at the same time is episodic, meaning that part of the procedure is automated and each step of the process triggers the next. This knowledge is also based on experience which is of an autobiographical nature. Findings - First, when one simulates human behavior, the hardest thing is to choose a knowledge representation that conforms as closely as possible to its emulation. According to, choosing a given type of knowledge representation is an art that is discovered little by little. And this is true as one designs and assembles a representation, when one realizes how far the authors are from imitating the design of the human brain processes, and discovers or invents methodologies to achieve it, combining the results of investigations into knowledge representation, cognitive psychology, and cognitive engineering. Second, solving any complex situation in the design process function of the clothing industry is no easy task. It requires plenty of experience in the manufacturing process. One needs the ability to identify the signals emitted by complex situations, and being able to stop them in time before they create irreversible damage. By merging the soft system methodology (SSM) and DMM with the experts' abilities and knowledge (the result being the EXITUS model (EM)), makes knowledge modeling possible. A problem cannot be solved if it is unknown, if the problem persists and grows it becomes more complex. By describing a problem, based on: its origin, its relationship, and its effects, it also confers the ability to solve it. Thereby, an SBC with the characteristics presented in this paper, not only improves the design process function as a whole, it also contributes to achieving corporate success. Finally, it influences directly on: a quality product; market positioning; and good economic results. First, the SBC-EXITUS was tested and endorsed by expert management designers. When a designer identifies a complex problem using the SBC-EXITUS system, he is capable of verifying its existence with facts and real life situations. This enables quicker decision making decisions, which saves time and money, due to the fact that a non-desirable state of affairs may be contemplated in advance. Fourth, in this project an SBC named SBC-EXITUS has been implemented using the SSM and the DMM, with the purpose of detecting possible problems in the design process of the clothing industry. Its implementation is developed by the use of production rules. Fifth, utilizing the methodology and the production rules like knowledge representation technique, make possible to acquire dominion knowledge in complex problems as in the study case clothing industry in Mexico. This approach is also applicable to other areas of design. Practical implications - Utilizing the methodology and the production rules like knowledge representation technique, make possible to acquire dominion knowledge in complex problems as in the study case clothing industry in Mexico. This approach is also applicable to the garment industry in the international context, as well as to other areas of design such as architecture, furniture, and others. The EM is a generic methodology. In this research and case, it has been applied in a design process within the garment industry, specifically in a case in Mexico. Yet this does not limit its use in a different context and problem situation. Having in mind the difference between countries in aspects such as sponsoring, technology, worker skills, marketing, etc. the tools for diagnostics in the Appendix 2, may be adapted to each context with enriched questions directed to specific aspects involved. Every new application allows facts and production rules that make the use of ES more efficient. Social implications - The ES-EXITUS was tested and endorsed by expert management designers. When a designer identifies a complex problem using the ES-EXITUS, he is capable of verifying its existence with facts and real life situations. This enables quicker decision making decisions, which saves time and money, due to the fact that a non-desirable state of affairs may be contemplated in advance. Originality/value - Solving any complex situation in the design process function of the clothing industry is no easy task. It requires plenty of experience in the manufacturing process. One needs the ability to identify the signals emitted by complex situations, and being able to stop them in time before they create irreversible damage. By merging the SSM and DMM with the experts' abilities and knowledge (the result being the EM), makes knowledge modeling possible. A problem cannot be solved if it is unknown, if the problem persists and grows it becomes more complex. By describing a problem, based on: its origin, its relationship and its effects, it also confers the ability to solve it. Thereby, an SBC with the characteristics presented in this paper, not only improves the design process function as a whole, it also contributes to achieving corporate success. Finally, it influences directly on a quality product, market positioning, and good economic results.