انعطاف پذیری سرمایه گذاری در سیستم تولید : روش تصمیم گیری چند هدفه

In today’s market, flexibility has received much concern of companies because of its important role in responding to the ever-changing environment. Most of the research in the literature on flexibility investment is focused on one objective, i.e. profit maximization. However, more and more company managers and customers pay much attention to another objective, i.e. production efficiency maximization. There is a need to develop flexibility investment strategy by considering these two objectives simultaneously. In this paper, we propose a multi-objective decision making method to derive the optimal flexibility investment strategy. Both efficiency and profit are taken into account as objectives when making flexibility investment decisions. The proposed method is a hierarchical method which composes of two-level models. Based on the characteristics of the models, a guideline is presented to help managers conveniently find out the optimal flexibility investment strategy. Simulation experiments are performed to verify the validity of the proposed models and guideline. The results of the simulation illustrate that the flexibility configuration obtained by following the guideline leads to benefit very close to total flexibility configuration, and much higher than 2-Skill Chain configuration; while the cost much less than total flexibility configuration, and even less than 2- Skill Chain configuration.

Global competition and shortened product life cycles bring to the diversification of customer demand and the increasing complexity of the production environment (Chan, Bhagwat, & Wadhwa, 2007). Faced with these pressures, more and more companies invest on flexible manufacturing systems (FMS). Typically, flexibility refers to the ability that machines can process various parts without requiring a prohibitive effort in switching from one part to another (Chen and Chung, 1996 and Jordan and Graves, 1995). Flexibility provides companies the ability to match production to market demand (Sethi & Sethi, 1990), thereby increasing profits, capacity utilization and customers’ satisfaction. Despite of its benefits, however, flexibility comes at the expense of increased cost of requiring flexible manufacturing capacity, as compared with dedicated or nonflexible capacity (Fine & Freund, 1990). Thus how to invest for flexibility in a manufacturing system is an important problem for decision makers. The extant literature on flexibility investment strategies can be approximately categorized into two groups. Researches in the first group start with building a flexibility measure, followed by finding out the configuration with satisfactory flexibility but not much investment cost. Examples in this kind include Jordan and Graves, 1995, Graves and Tomlin, 2003, Iravani et al., 2005, Hua and He, 2010a and Hua and He, 2010b, etc. Jordan and Graves (1995) propose an index to measure the process flexibility of a system configuration. Based on the index, they find that few and long chains can have nearly the same performance (expected sales quantity) as total flexibility configuration. Following this work, Graves and Tomlin (2003) find that the chaining structure is also effective for supply chains. Iravani et al. (2005) propose a set of measures for structural flexibility by solving maximal flow problems. They also verify that the chaining structure performs rather well. Considering bill of material (BOM) constraints, Hua and He (2010a) present a set of hierarchical measures for process flexibility of production line and manufacturing system. Based on these measures, Hua and He (2010b) propose some flexibility investment guidelines for decision makers. Researches in the second group identify the flexibility investment strategies by directly solving flexible capacity investment or production programs. For example, Fine and Freund (1990) present a model of the firm’s flexible manufacturing investment decision. With the aid of the model, they characterize the necessary and sufficient conditions for a firm to invest in flexible capacity to protect efficiently against uncertainty in demand for all of its products. Van Mieghem (1998) highlights the important role of price and cost mix differentials, which significantly affect the investment decision and the value of flexibility. Akşina and Karaesmen (2007) identify preferred flexibility structures in service or manufacturing systems, by solving a network flow type model. They find out general structural properties of flexibility design pertaining to the marginal values of flexibility and capacity. Gong and Hu (2008) develop a product mix flexibility model, in which the product mix flexibility is measured by an economic index. They identify the bottlenecks that affect the flexibility most. Chou, Chua, Teo, and Zheng (2009) give some analytical results on the performance of chaining strategies vis-á-vis the total flexibility system. The literature reviewed above investigates the flexibility investment problem with the only objective of profit maximization (or equivalently cost minimization). In the competitive market of today, there is another objective that companies also pay much attention to, i.e. production efficiency. Production efficiency is an important factor that affects customer satisfaction. In today’s market, time-based competition makes more and more companies focus on production efficiency (Koufteros, Vonderembse, & Doll, 1998). Production efficiency is mainly determined by the time spent to complete the production of all the demand. However, time and cost are often inversely related (Gupta & Goyal, 1989). In most cases, time and cost have their trade-off and time might be more important than cost (Chang, Whitehouse, Chang, & Hsieh, 2001). For non-homogeneous machines and parts (i.e. the unit processing cost as well as processing time is different for different machine-part pairs), different flexibility configurations lead to different system performance (in terms of profit-efficiency mix). Considering these two objectives, the flexibility investment methods reviewed above are not applicable. It is necessary to build a multi-objective method (Wang and Zeng, 2010 and Wu et al., 2009) on flexibility investment for companies. This paper aims to propose a multi-objective decision making method for guiding flexibility investment. The rest of the paper is organized as follows. In Section 2 we define the problem to be investigated, and develop a two-level multi-objective programming model. Section 3 presents a guideline for constructing flexibility configuration based on the proposed model. Section 4 discusses the way the weights of multiple objectives can be determined. A set of simulation experiments is performed in Section 5 to verify the proposed models and guideline. Section 6 concludes this paper.

In this paper, we investigate the flexibility investment problem for companies by proposing a multi-objective method. Both the efficiency and profit are taken into account as objectives when making flexibility investment decisions. Actually, the proposed method can be easily generated to considering more than two objectives. The proposed method is a hierarchical method which composes of two-level models. The lower level model aims to give the optimal production planning decisions upon given flexibility configuration and demand realizations. The higher level model aims to find the optimal flexibility strategy based on the results obtained by solving the lower level model. However, it is unrealistic or at least rather complicated to directly solve the higher level model. Therefore a guideline is presented based on the characteristics of the two level models. The guideline is built to help managers conveniently find out the optimal flexibility investment strategy. To verify the validity of the proposed models and guideline, a set of simulation experiments are performed. The results of the simulation illustrate that the flexibility configuration obtained by following the guideline leads to benefit very close to total flexibility configuration, and much higher than 2-SC configuration; while the cost much less than total flexibility configuration, even less than 2-SC configuration. Future research directions include but are not limited to the following two. First, in this paper, the expected improvement of system performance is approximated by the average value of simulation experiments. Effort can be devoted to mathematically compute the expectation for specific demand distribution, e.g. normal, uniform, etc. Second, many assumptions in this paper could be relaxed. For example, unfilled demand may be backordered rather than lost; penalty can be considered for unfilled demand; and so on.