1. Business
  2. Operations Management
  3. 1 consider the paper titled manufacturing network design with reliable...

Question: 1 consider the paper titled manufacturing network design with reliable...

Question details

Abstract This paper presents an optimization methodology for the design of manufacturing networks with reliable promising capabilities. The designed manufacturing network must have the capability to deliver products to customers within a promised combination of order-to- delivery time and service level. To cope with demand variability, the approach seeks an advantageous trade-off between service offered, standard parts inventory and manufacturing capacity deployment. Given the current state of the network, the facilities to open, close and reconfigure are optimally chosen from a set of potential facilities. Within each of these facilities, the selected production centers are configured so as to specify their assigned products, resources, capacity and lead time. The demand depends on the product- market offer selected. The model seeks to maximize profits related to the expected demand, revenues and costs of a selected set of product-market offers. A mixed integer programming model based on an engineering approach is presented. Results illustrating the practical use of the model as a tool for generating robust manufacturing networks are also presented1. Problem context This paper presents a methodology for the design of manufacturing networks with reliable promising capabilities. The enterprise makes order-to-delivery time and service level commitments for each of its product-market, while taking into account the stochastic nature of the demand. These commitments have an impact on the expected demand. When demand is stochastic and customers expect short lead times, a network capacitated for producing up to the average daily demand leads to many unsatisfied customers. On the other end, a network capacitated for producing up to the maximum forecasted daily demand leads to low resource utilization and high fixed costs. Therefore the enterprise needs to calibrate its network capacity in conjunction with its product-market offers development For example, if the business aims to offer a 99%-reliability two-day delivery, it must capacitate its network so as to be able to fulfill this offer in face of stochastic daily demand Different strategies may be used to achieve the capacity to cope with demand variability Capacity can be calibrated through the amount of resources used, work profiles in terms of allowed shifts and overtime, standard parts inventory at decoupling points, and the externalization of some activities through subcontracting and outsourcing. Therefore, the enterprise needs to evaluate the required capacity and the deployment of this capacity in its network at the same time as it develops its product-market offers. 2 The supply chain network design problem integrates many decisions that have evolved over the last 30 years. Good starting points for the study of this problem are the reviews of Verter and Dincer (1992) and Geoffrion and Powers (1995). The decisions considered in this problem are related to facilities location, capacity acquisition and technology selection. Erengüç et al 1999) discuss several of these decisions. Other relevant elements such as service levels and capacity planning related to the robustness of discussed by Martel (2005). However, the explicit consideration of many of these decisions in a single formulation leads to very complex mathematical models that are difficult to solve (Min and Zhou, 2002). The most comprehensive design models proposed in the literature include Cohen and Lee (1988), Cohen and Moon (1990), Cohen and Moon (1991) and Arntzen et al. (1995). These papers consider lead times, inventory decisions and economies of scale and scope in their proposed models. More recent papers integrate stochastic demand and lead time dependent demand with service levels (Rao et al., 2000 the supply chain areManufacturing network design with reliable promising capabilities 3 Keskinocak et al., 2001; So and Zheng, 2003; Candas and Kutanoglu, 2004; Jeet and Kutanoglu, 2004; Slotnick and Sobel, 2005; Santoso et al., 2005). Escudero et al. (1999) and Vidal and Goetschalckx (2000) discuss different types of uncertainty in the supply chain. Strategic positioning of the inventory in the network is another important aspect to consider. Through a case study Billington et al. (2004) present a robust way to design supply-chain networks that determines the location of stock when considering service levels and demand uncertainty. Vidal and Goetschalckx (1997) present a review on several of the mathematical formulations available in the terature. A modeling framework integrating most of the decisions is presented in Martel (2005). To the best of our knowledge, all these papers consider the design problem in make-to-stock environment. The methodology proposed in this make-to-order supply chains, including the positioning of decoupling inventory in the network. pape r addresses the design of assemble-to-order and Our approach also covers some layout decisions for each facility. Benjaafar et al. (2002) present a review of recent progress on layout methodologies. Montreuil and Lefrançois (1996) present a framework for layout design based on responsibilities and Montreuil et al. (1998) propose a framework to design faciliies using these types of organization. Key aspects for the design of layouts are presented by Montreuil (2000). The methodology also deals with capacity planning problems under demand uncertainty. Eppen et al. (1989), Karmarkar (1993), Verter and Dincer (1992), Verter and Dasci (2002), Benjaafar and Sheikhzadeh (2000), Orcun et al (2003), Venkatadri et al. (2004), Zhang et al. (2004) and Cochran and Marquez Uribe (2005) propose different approaches and models for this problem. The goal of our manufacturing network design methodology is to maximize the expected operating profits associated with the redeployment of an existing manu facturing network. The methodology proposed is appropriate for enterprises producing make-to-order and assemble-to-order products with high demand variability. Trade-offs between the expected demand and revenues stemming from the potential product-market offers on one side and the required network capacity and inventory on the other side are analyzed and the most profitable offer is identified for each product in each demand zone. The methodology is an engineering approach to design robust manufacturing networks that respect the offers madeAfter the optimization, more precise capacity and inventory requirements can be computed with the current selection of center configurations, and the demand network can be simulated to ensure that the promised service levels are respected. These computations give new values for some of the selected center configuration parameters and the network must be re-optimized. The iterative phase (Steps 4 and 5) stops when no changes occur in the selection of center configurations and in the values of their parameters. The methodology leads to a reliable manufacturing network which can be used as an advanced basis for further investigation before its final deployment. propagation in the 5 Initial Phase Manufactured Products Potential Resources Finished Potential Processors Potential Worker Raw Materials and Manufactured Products a Set 2. Create a Set of 3. Create a Set of Potential Locations Demand Based on the CurrentZones -Potential Offers Networik Potential Suppliers Plants and Demand Zones Potential Supply and Manufacturing Network Potential Center Offers Legend Iterative Phase Step in the 4. Compute Center 5. Generate MP Model and Optimize Designed Supply and Network Figure 1: Manufacturing Network Design Methodology 3. Modeling Framework The modeling framework elaborated in this section describes the components of the potential manufacturing network build in the initial phase of the methodology 3.1. Products and Product-States A product graph with the sequence of manufacturing operations for each product is illustrated for a simple case in Figure 2. Products 1, 2 and 3 are raw materials, products 4,Manufact network design with reliable abilities The required inventory at the decoupling point is computed with Equation (37) and the inventory factor λ,.pg to be used in constraints (27) is computed with Equation (38) Xpip . λι.orA.. . g㎡λη,A, (according to the tumover cuve) λ,m μル./h(A允,Ep.«.) (default value set to 0.00) In order to respect the service level of the offer a the service level a of some center configurations can be modified during the simulation of the demand in the algorithm, leading to new values of the capacity inflation factor and the inventory factor 25 6. Experimental Investigation The objective of the experimental investigation is to demonstrate the feasibility of the approach and to develop managerial insights for typical uses of the model as a strategic decision making tool. The purpose of the methodology is to help enterprises in the redeployment of their existing manufacturing network. The methodology provides a robust feasible design which can be used as a good basis for further investigations. Therefore, the proposed methodology helps the designer in the testing of different supply chain strategies, under different scenarios, to generate efficient reliable networks in a what-if approach without necessarily redefining all aspects of the current network simultaneously. In our experiments, the models are solved with the CPLEX 10.0 solver (ILOG, 2006) on an AMD Athlon MP 2600 + processor. The simulator is built in VBA with Microsoft Excel 2003 The characteristics of the initial manufacturing network used to validate and illustrate the sign approach are given in Table 1 Table I: Starting Network Configuration Initial Network Configuration Plant 21: Centers 2101, 2102 Plant 22: Centers 2207, 2208, 2209 Plant 23: Center 2313 Plant 24: Center 2413 Plant 21: 181 processors (4 types) & 181 workers (4 types) Plant 22: 132 processors (4 types) & 136 workers (6 types) Plant 23: 16 processors (5 types) & 15 workers (5 types) Plant 24: 36 processors (6 t Plants Centers &36 workers (5t DT-2006-AM-2 25network design with reliable abilities The initial manufacturing network is presented in Figure 3 and in Figure 10. The product- state graph for the case considered is presented in Figure 2 and the potential center missions and resources are those presented in Figure 4 and Figure 5. In this network, plant 21 is configured for the manufacturing of products 4 & 5 with product centers while plant 22 produce parts 6 & 7 with function centers. Plants 23 & 24 produce finished products 8 &9 each with a single product group center Plant 21 26 A Demand Zones Text Figure 10: Initial Manufacturing Network Using this network as a basis, different types of strategies are tested to illustrate the effect of lead times, service levels, and decoupling points on the required capacity deployment and profitability of the network. Source-to-order (STO) strategies without inventory, make- to-order (MTO) strategies using raw material inventories, and assemble-to-order (ATO) strategies with raw material and standard part inventories are tested. Strategies are also defined by potential product-market offers (selling prices, OTD times and service levels) and by the potential network used (plants, centers and center configurations with potential re sources, lead times and service levels). For each offer, different sales response scenarios (values of the scale parameter p for each demand zone) are tested. For example, in a scenario, a demand zone can be interested by low prices, but another by short OTD times and high service levels. The business scenario parameters considered are presented in Table 2. Not combinations of these parameter values are tested; only those leading to meaningful what-if scenarios are considered. For the first sets of experiments, the potential DT-2006-AM-2 26

  1. 1.

Consider the paper titled: “Manufacturing network design with reliable promising capabilities”.

  1. Read the abstract and the first paragraph of first section and write a paragraph about the problem addressed in this paper.
  2. What is the goal of the paper? This can be found in “problem context” section of the paper.
  3. What is the methodology used to obtain the specified goal?
  4. How many products do they have and what are their types?  
  5. What are the components of potential manufacturing network?
  6. Pages 25-26 present the initial network configuration, and pages 31-32 present the optimal (best) network configuration. Compare the two configurations in terms of:
    1. Number of plants and centers
    2. Resources
    3. Inventory
    4. And profits.  
Solution by an expert tutor
Blurred Solution
This question has been solved
Subscribe to see this solution