Volume_Flex

Raturi: OM 585 1 Volume Flexibility in a Network This exercise demonstrates the concepts of volume flexibility in a multiple product, multiple plant network. It allows the participants to 1. Understand the relationship between capacity utilization and service levels as measured by probability of a stock out and fill rate. 2. Understand how multiple plant networks create flexibility 3. Develop a working definition of flexibility (specifically volume flexibility) 4. Understand the concept of chaining and how limited flexibility gives almost the same results as complete flexibility. Source: Jordan, W. C. and Graves S.C. (1995). ‘Principles on the Benefits of Manufacturing Process Flexibility’, Management Science’, 41(4), pp. 577-594. Your firm makes 2 products A and B in two plants, 1 and 2. The demand for products A and B is independ and identically distributed (i.i.d.) and could be 5, 10 or 15 with probability 1/3 each. Each plant has 10 units of capacity. Product A Plant 1 Product B Plant 2 Case 1: The plants have no product flexibility (that is plant 1 makes product A and plant 2 makes product B). What is the probability of meeting demand of products A and B' What is the fill rate of products A and B' (The fill rate is defined as the amount of demand met divided by the expected demand). Case 2: The plants have complete flexibility (that is plant 1 and plant 2 can make both products A and B). Thus when demand for product A is 15 and for product B is 5, plant 1 makes 10 A and plant 2 makes 5B and 5A. What is the probability of meeting demand of products A and B' What is the fill rate of products A and B' (The fill rate is defined as the amount of demand met divided by the expected demand). This exercise demonstrates the concepts of volume flexibility in a multiple product, multiple plant network. Specifically, you conclude the following: 1. Designing multiple plant networks is a challenge: "plants" here refers to processing centers whether they are in manufacturing or services. 2. Flexibility (in this case volume flexibility) of processes can be measured in terms of the simultaneous achievement of two objectives. In this case, networks, which simultaneously increase resource utilization and service levels, are more flexible. Similarly, inventory methods that simultaneously increase service level and decrease Raturi: OM 585 2 inventory levels are more flexible (as in Barilla). Processes such as multi-tasking or cross training simultaneously increase efficiency and variety and increase flexibility. 3. Flexibility is the ability of a system to cope better with environmental uncertainties. In this case network 2 is more flexible as it can cope with demand uncertainty better. Flexibility is a measure of the process (technology); not the environment. 4. "Chaining" provides many of the benefits of "full" flexibility. It may not be worthwhile to pursue the agenda of full flexibility (as in complete cross training of every operator in line E in Andreas Stihl). An interesting measure of volume flexibility is the capability of a firm to supply demand (have lower stock out) and simultaneously minimize its inventory or maximize its capacity utilization. Harvey is able to do this through product knowledge in the sales force. Consider a firm that makes two products in two facilities: products A and B in plants, 1 and 2. The firm has a choice of dedicating plants to products (making product A in plant 1 and product B) or making the plants flexible (both products A and B are made in both plants. The demand for products A and B is unpredictable: it could be could be 5, 10 or 15 with probability 1/3 each. Each plant has 10 units of capacity. Figure 1 shows a pictorial representation of the two alternatives. Figure 1: Example of a network with volume flexibility In the top figure, the plants do not have the flexibility to produce both the products. When the demand for product A or B is 15, the system cannot supply the product. Thus it will meet demand for product A 2/3 of the time and for product B 2/3 of the time. Of the nine scenarios for product demand, in only four cases will the plants be able to meet demand - when demand for A and B is (5,5) or (5,10) or (10,5) or (10,10). Table 1 calculates the fill rate and capacity utilization for both the plants and for the whole system under this scenario. The fill rate for the product accounts for the 1/3 of the time demand is 15 and there are 5 Product B Plant 1 Plant 2 Plant 1 makes Product A Plant 2 makes Product B Product B Plant 1 10 units/ day Plant 2 Plant 1 makes both A and B Plant 2 makes both A and B Product A Product A Plant 2 makes both A and B Plant 1 makes both A and B Raturi: OM 585 3 units of stock out. The capacity utilization for each plant is based on operation at a production level of 5 (when demand is 5) 1/3 of the time and at a production level of 10 (when demand is 10 or 15) 2/3 of the time. Table 1: Performance in an Inflexible Plant Network Probability of stock out Fill rate Capacity Utilization Plant 1 - - 83.33% [=1/3* (5+10+10)/10] Plant 2 - - 83.33% [=1/3* (5+10+10)/10] Product A 1/3 83.33% [=1- 1/3*5/10] - Product B 1/3 83.33% [=1-1/3*5/10] - System 5/9 (=1-2/3*2/3) 83.33% 83.33% In the flexible alternative in Figure 10, plant 1 and plant 2 can make both products A and B. Thus when demand for product A is 15 and for product B is 5, plant 1 makes 10 A and plant 2 makes 5B and 5A. Also when demand for product A is 5 and for product B is 15, plant 1 makes 5 A and 5B and plant 2 makes 10B. This changes the system performance as followsi: Table 2: Performance with a Flexible network Probability of stock out Fill rate Capacity Utilization Plant 1 - - 88.89% [=1/9* (2*5+7*10)/10] Plant 2 - - 88.89% [=1/9* (2*5+7*10)/10] Product A 1/3 88.89% [=1-2/9*5/10] - Product B 1/3 88.89% [=1-2/9* 5/10] - System 1/3 88.89% 88.893% [=1/9* (10+2*15+6*20)/20] As can be seen by comparing Tables 1and 2, volume flexibility in the network has generated a simultaneous increase in the capacity utilization at the plants as well as an increased fill rate in the market. Thus inflexibility can be thought of the situation where a firm as underutilized capacity, as well as stock outs in the market as in the first scenario. Bottom Line: Flexibility, which allows a process to respond to changes in its environment, is the least precise and hardest to define of the characteristics we have considered thus far. Flexibility must often be described in qualitative terms; doing so, however, does not make it any less important to managers.ii i To complete the calculations, generate the 9 scenarios for demand. Now when demand for A and B is (5, 15) or (15, 5), the flexible network does not have any stock outs as 5 units of the production is shifted to the other plant. It still has stock outs when the demands are (10, 15), (15, 10) and (15, 15). ii Upton, David, “The Management of Manufacturing Flexibility” California Management Review, Winter, 1994 or Upton, David, "What really makes factories flexible'" Harvard Business Review, July-August, 1995.