SITY LlBRARIES llllll\lllllllll\\\\\\\\\\\\\\\\\\\ 3 1293 01029 8952 This is to certify that the dissertation entitled - 57:64 7567C g M L a H 7/ 0A) of” 1 Le Off/M44 ”67m; / mu ' C) 7C /\/9 Cd IDQUCYSS‘ flCLNC/tjy presented by fill/yum cybt Kill/14 has been accepted towards fulfillment of the requirements for Ph- D- degree in J18 niqgmsufr {471/644, #4LZ ‘/ Major professor Date Seotember 3, 1993 MS U is an Affirmative Action/Equal Opportunity Institution 0-12771 LIBRARY MIchigan State Unlverslty PLACE IN RETURN BOX to remove this checkout from your record. TO AVOID FINES return on or before date due. 4 DATE DUE DATE DUE DATE DUE I ml 1’ MSU Is An Affirmative Action/Equal Opportunity Institution Warns-9.1 STRATEGIC EVALUATION OF THE OPTIMAL ACQUISITION OF NEW PROCESS TECHNOLOGY BY Hyun Gyu Kim A DISSERTATION Submitted to Michigan State University in partial fialfillment of the requirements for the degree of DOCTOR OF PHILOSOPHY Department of Management 1993 ABSTRACT STRATEGIC EVALUATION OF THE OPTHVIAL ACQUISITION OF NEW PROCESS TECHNOLOGY By Hyun Gyu Kim This dissertation focuses on the strategic decisions involved in optimal investment policies for the acquisition of new process technology. Here, optimal investment policies include not only the acquisition of new process technology but also the attrition of existing manual production capacity. This study explores analytically why and how the technology decisions involving these optimal investment policies are related to corporate strategy and external market environment. This dissertation also presents a conceptual framework for investment decisions regarding the acquisition of new process technology. This framework explains the relationships among decision variables and other variables used in the development of a model. The model developed here is a dynamic optimal control model. It has an objective function which maximizes the long-term value of a firm's strategic business unit (SBU) and also minimizes both tangible and intangible costs associated with the use and acquisition of process technology during the planning horizon. These cost factors include not only the direct costs associated with the acquisition and attrition of process technology but also the cost associated with the flexibility attribute of process technology. It is assumed here that there are two components to process technology, capacity (automation capacity) and flexibility. Capacity refers to the total volume that can be produced by the system, whereas flexibility refers to the variety of product (part) types that the system can handle. Other cost factors in the objective function are related to the penalty costs associated with deviations between an SBU’s actual market demand and its goals, as well as deviations between its actual market demand and its production level. These deviations are measured in terms of both volume and variety. The constraints address the change in market potential for volume and variety, the change in unit production cost due to the acquisition of new process technology and flexibility, and the change in system capacity and flexibility. The optimal solution of the model provides decisions regarding both the acquisition of new process technology affecting the system capacity and flexibility, and the attrition of the conventional (existing) production capacity. A scenario analysis is conducted to examine several research issues. The results of scenario analysis lead to some insightful conclusions. First, in the long run for the optimal performance of a firm's manufacturing systems it should select a strategic priority which is congruent to external market conditions. Second, the acquisition policy of new process technology is affected by the type of product life cycle (external market conditions). Copynght by IIyun(3yul§hn 1993 Dedicated to my father ACKNOWLEDGMENTS My deepest gratitude goes to dissertation advisor, Dr. Soumen Ghosh, for his support and guidance throughout this dissertation research and the Ph.D. program. His kindness allowed me to finish this dissertation. I also wish to express my sincere appreciation to the members of the dissertation committee, Dr. Michael Moch, Dr. Paul Rubin, and Dr. Shawnee Vickery, for their valuable comments and suggestions. In addition, I want to acknowledge my indebtedness to Dean Gill Lim of International Studies and Program who always supported and encouraged me. I wish to cite the contribution of my mother, who has supported my education for more than thirty years. I also thank my wife and children. Without their patience and sacrifice, it would have been difficult to finish this dissertation. Finally, I wish to thank all the teachers and friends of my life whose encouragement has been always a prime motivation for my study. vi TABLE OF CONTENTS page LIST OF TABLES . . . . . . . . . xiv LIST OF FIGURES . . . . . . . . . xvii CHAPTER I. INTRODUCTION . . . . . . . . 1 1.1 OBJECTIVES OF RESEARCH 2 1.1 RESEARCH ISSUES . . . . . . 5 1.1.1 Flexibility and Cost Efficiency of New Process Technology . . . . 5 1.1.2 Acquisition of New Process Technology and Evolution of Technology Choice . . . 8 1.2 RESEARCH QUESTIONS . . . . . 10 1.2.1 Questions for The Acquisition of New Process Technology . . . . 10 1.2.2 Questions for The Evolution of Technology Choice . . . . . l 1 1.4 SUMMARY . . . . . . . . 11 II. LITERATURE REVIEW . . . . . . . 13 2.1 MANUFACTURING FLEXIBILITY . . . . 14 2.1.1 Definition of Flexibility . . . . 15 vii 2.2 Manufacturing Flexibility and Its Strategic Implications 2.1.2.1 Definition of Product Mix Flexibility and Volume Flexibility 2.1.2.2 Strategic Value of Manufacturing Flexibility. Measurement of Manufacturing Flexibility Trade-off Between Flexibility and Efficiency . Summary and Implications PROCESS TECHNOLOGY AND 2.2.1 2.2.2 PRODUCT LIFE CYCLE Process Technology in Manufacturing Strategy 2.2.1.1 Characteristics of Process Technology 2.2.1.2 Process Technology in Capacity Planning 2.2.1.3 Process Technology in Manufacturing Strategy Product Life Cycle in Manufacturing Strategy 2.2.2.1 Product Volume and Product Mix in the Product Life Cycle 2.2.2.2 Manufacturing Cost in PLC 2.2.2.3 Reversal of PLC Integration of Product and Process Decisions 2.2.3.1 Product-Process Matrix 2.2.3.2 Effect of Flexible Technology on Product-Process Matrix 2.2.4 Summary and Implications viii 17 17 18 20 22 24 25 25 27 29 2.3 NEW PROCESS TECHNOLOGY AND MANUFACTURING FLEXIBILITY . . . 39 2.3.1 New Process Technology . . . . . 39 2.3.1.1 Definition of New Process Technology . . 39 2.3.1.2 Rigid Automation and Flexible Automation . 39 2.3.1.3 Measurement for the Flexibility of New Process Technology . . . 41 2.3.2 Effects of New Process Technology . . . . 42 2.3.2.1 Effects on Manufacturing Lead Time . . 43 2.3.2.2 Effects on Cost Structure . . . . 43 2.3.2.3 Effects on Sales Growth . . . . 44 2.3.3 New Process Technology and Learning Effect . . 45 2.3.4 Analytic Studies for New Process Technology . . 47 2.3.4.1 The Model of Fine and Freund . . . 47 2.3.4.2 The Model of Gaimon . . . . 49 2.3.4.3 The Model ofRoth et al. . . . . 50 2.3.5 Summary and Implications . . . . 52 2.4 RESEARCH STATEMENT . . . . . 52 III. STRUCTURAL FRANIEWORK FOR THE STRATEGIC ACQUISITION OF NEW PROCESS TECHNOLOGY: A CONCEPTUAL FRAMEWORK . . . . 54 3.1 DECISION PROCESS OF NEW TECHNOLOGY ACQUISITION . . . . 55 3.2 TECHNOLOGY AND EXTERNAL ENVIRONMENT . . 59 3 .3 TECHNOLOGY AND MANUFACTURING COST STRUCTURE . . . 6O ix 3 .4 PRODUCT-TECHNOLOGY BOX THE MODEL 4. 1 INTRODUCTION 4.1.1 General Description of The Model 4.1.2 Assumptions . 4.1.3 Variables 4.1.3.1 Endogenous Variables 4.1.3.2 Exogenous Variables 4.2 THE OBIECTIVE FUNCTION 4.3 CONSTRAINTS 4.3.1 Change in Market Demand 4.3.2 Change in Variable Costs 4.3.3 Change in Production Capacity 4.3.4 Change in Cumulative Capacity of New Technology. 4.3.5 Change in Overall System Flexibility 4.3.6 Inequality Constraints . 4.3.7 Variable Boundaries 4.4 SUMIVIARY SOLUTION APPROACH 5.1 OPTIMAL POLICIES 5.1.1 Necessary Conditions 5.1.2 Optimal Policies for Acquiring New Technology 62 68 68 68 69 7O 7O 71 76 76 78 79 80 81 82 82 84 84 84 86 5.2 5.1.3 Optimal Policies for Reducing Conventional Capacity 5.1.4 Optimal Policies for Production Rates SOLUTION ALGORITHM 5.2.1 Discretized Non-Linear Version of the Model. 5.2.2 Limitations VI. SCENARIO ANALYSIS 6.1 6.2 OVERVIEW OF DESIGN FOR SCENARIO ANALYSIS . 6.1.1 Strategic Choices and the PLC 6.1.2 Positioning of Initial Process Technology 6.1.3 Settings of Variables . 6.1.4 Values of Exogenous Variables 6.1.5 Hypothetical PLC curves 6.1.6 Frameworks of Design 6.1.7 Performance Measures RESULTS OF ANALYSIS 6.2.1 Analysis with the Normal PLC 6.2.1.1 Effect of Initial Process Technology . 6.2.1.2 Effect of Strategic Choice 6.2.1.3 Conclusion 6.2.2 Analysis with the Reversed PLC 6.2.2.1 Effect of Initial Process Technology . 6.2.2.2 Effect of Strategic Choice 6.2.2.3 Conclusion xi 89 9O 92 92 93 95 95 95 97 99 114 114 114 116 118 119 119 120 122 6.2.3 Comparison Between the Normal PLC and the Reversed PLC. 6.2.3.1 Optimal Acquisition of New Process Technology 6.2.3.2 Evolution of Process Technology 6.2.3.3 Cost Performance 6.3 SUMMARY OF FINDINGS . 6.3.1 Acquisition of New Process Technology 6.3.2 Evolution of Process Technology 6.3.3 Best Strategic Choice 6.3.4 Conclusion VII. CONCLUSIONS 7.1 RESEARCH OVERVIEW 7.1.1 Research Scope and Issues 7.1.2 Research Model 7.1.3 Research Methodology 7.1.4 Summary of Findings . 7.1.5 Limitations of Study 7.2 CONTRIBUTIONS 7.3 POSSIBLE RESEARCH EXTENSIONS APPENDICES I. Derivation of Necessary Conditions 11. An Example of Model in Spreadsheet. xii 124 125 125 126 126 173 174 174 174 175 176 177 179 180 BIBLIOGRAPHY . . . . . . . . . 183 xiii Table 2.1 3.1 6.1 6.2 6.3 6.4 6.5 6.6 6.7 6.8 6.9 6.10 6.11 6.12 6.13 LIST OF TABLES Characteristics of Process Choice Objectives of the Acquisition of New Technology Strategic Choices and the Normal PLC Strategic Choices and the Reversed PLC Variable Settings for the Stages of the PLC Variable Settings for Strategic Priority Value Settings of Research Variables Related to the Strategic Choice Variable Settings for the Positioning of Initial Technology Choice Value Settings for the Research Variables Related to Initial Process Technology. Value Settings for Exogenous Variables Values of Variables of the Normal PLC Values of Variables of the Reversed PLC Design Framework with the Normal PLC Design Framework with the Reversed PLC Settings of Research Variables xiv page 26 58 96 97 99 100 101 102 102 104 106 107 110 110 111 6.14 6.15 6.16 6.17 6.18 6.19 6.20 6.21 6.22 6.23 6.24 6.25 6.26 6.27 6.28 6.29 6.30 6.31 6.32 6.33 Value Settings of Research Variables . Acquisition of Automation Capacity with the Normal PLC Acquisition of Flexibility with the Normal PLC Cumulative Automation Capacity with the Normal PLC Cumulative System's Flexibility with the Normal PLC Acquisition of Automation Capacity with the Normal PLC Acquisition of Flexibility with the Normal PLC Cumulative Automation Capacity with the Normal PLC Cumulative System's Flexibility with the Normal PLC Acquisition of Automation Capacity with the Reversed PLC. Acquisition of Flexibility with the Reversed PLC Cumulative Automation Capacity with the Reversed PLC Cumulative System's Flexibility with the Reversed PLC Acquisition of Automation with the Reversed PLC Acquisition of Flexibility with the Reversed PLC Cumulative Automation Capacity with the Reversed PLC Cumulative Flexibility with the Reversed PLC Cost Performance Acquisition of Automation Capacity with the Normal PLC (el = 10, e2 = 20) Acquisition of Flexibility with the Normal PLC (el = 10, e2 = 20) Cumulative Automation Capacity with the Normal PLC (el = 10, e2 = 20) 112 129 130 134 137 141 142 145 146 157 158 161 164 166 6.35 Cumulative System's Flexibility with the Normal PLC (el = 10, e2 = 20) . . . . . . . 167 xvi Figure 2.1 2.2 2.3 2.4 2.5 b.) p—J 3.2 3.3 3.4 3.5 3.6 3.7 6.1 6.2 6.3 6.4 LIST OF FIGURES Variety-Related Costs and Scale-Related Costs Product and Process Matrix The Effect of Flexible Automation on the Product-Process Matrix Integration of Automation and PLC Four Stages of Possible Evolution of the PLC Structural Diagram of the Decision Process for the Acquisition of New Advanced Technology Linkage Among Various Decisions Technology and Investment Cost Automation and Average Volume-Related Variable Costs Flexibility and Average Variety-Related Variable Costs Technology Matrix Product-Technology Box Four Types of Initial Technology Choice Positioning of Initial Technology Choice Changes in Total Market Volume of the Normal PLC Changes in Total Market Variety of the Normal PLC xvii 64 66 98 98 108 108 6.5 6.6 6.7 6.8 6.9 6.10 6.11 6.12 6.13 6.14 6.15 6.16 6.17 6.18 6.19 6.20 6.21 6.22 6.23 6.24 6.25 6.26 Changes in Total Market Volume of the Reversed PLC Changes in Total Market Variety of the Reversed PLC Acquisition of Automation with the Normal PLC Acquisition of Flexibility with the Normal PLC Cumulative Automation Capacity with the Normal PLC Cumulative System's Flexibility with the Normal PLC Acquisition of Automation Capacity with the Normal PLC Acquisition of Flexibility with the Normal PLC Cumulative Automation Capacity with "the Normal PLC Cumulative Flexibility with the Normal PLC Acquisition of Automation with the Reversed PLC Acquisition of Flexibility with the Reversed PLC Cumulative Automation Capacity with the Reversed PLC Cumulative system's Flexibility with the Reversed PLC Acquisition of Automation Capacity with the Reversed PLC. Acquisition of Flexibility with the Reversed PLC Cumulative Automation Capacity with the Reversed PLC Cumulative System's Flexibility with the Reversed PLC Evolution of Process Technology with the Normal PLC Evolution of Process Technology with the Reversed PLC Acquisition of Automation Capacity with the Normal PLC (el = 10. e2 = 20) Acquisition of Flexibility with the Normal PLC (e1=10,e2=20) . xviii 109 109 131 132 135 136 139 140 144 147 148 151 152 155 156 159 160 168 169 6.27 Cumulative Automation Capacity with the Normal PLC (e1= 10, e2=20) . . . . . . . 170 6.28 Cumulative System's Flexibility with the Normal PLC (el = 10, e2=20) . . . . . . . 171 6.29 Evolution of Process Technology with the Normal PLC (el = 10, e2=20) . . . . . . . 172 xix CHAPTER I INTRODUCTION This dissertation focuses on the strategic decisions involved in optimal investment policies for the acquisition of new process technology. Here, optimal investment policies include not only the acquisition of new process technology1 but also the attrition of existing manual production capacity. This study explores analytically why and how the technology decisions involving these optimal investment policies are related to corporate strategy and external market environment. This dissertation also presents a conceptual framework for investment decisions regarding the acquisition of new process technology. This framework explains the relationships among decision variables and other variables2 used in the development of a model. The model develdped here is a dynamic optimal control model. It has an objective firnction which maximizes the long-term value of a firm's strategic business unit (SBU) and also minimizes both tangible and intangible costs associated with the use and acquisition of process technology during the planning horizon. These cost factors include not only the direct costs associated with the acquisition and attrition of process technology but also the cost associated with the flexibility attribute of process technology. It is assumed here that there are two components to process technolog , capacity (automation capacity) and flexibility. Capacity refers to the total volume that can be produced by the system, whereas flexibility refers to the variety of product (part) types that the system can handle. Other cost factors in the objective firnction are related to the penalty costs associated with 1 Hereafter, "new process technology" is sometimes referred to as "new technology." 2 In an optimal control model, decision variables are called "control variables." and the other variables are called "state variables." 2 deviations between an SBU's actual market demand and its goals, as well as deviations between its actual market demand and its production level. These deviations are measured in terms of both volume and variety. The constraints address the change in market potential for volume and variety, the change in unit production cost due to the acquisition of new process technology and flexibility, and the change in system capacity and flexibility. The optimal solution of the model provides decisions regarding both the acquisition of new process technology affecting the system capacity and flexibility, and the attrition of the conventional (existing) production capacity. From the scenario analysis using the model, this study attempts to evaluate two strategic issues found frequently in the literature. One is the effect of the strategic emphasis on flexibility over cost efficiency on optimal policies, and the other is the effect on the changes in the position of technology choice on the product - technology matrix3 due to technology acquisition. These two issues are evaluated within the framework of the product life cycle (PLC) theory. In doing so, this dissertation attempts to provide insights for both managers and researchers on how to make strategic acquisition decisions for new process technology. This is an important understanding given the current, rapid diffusion of new process technology. 1.1 OBJECTIVES OF RESEARCH Faced with fierce and ever-rising competition created by the proliferation of new products and shorter or truncated product life cycles, many US. manufacturing firms recently began to adopt new process technology, such as the Just-In-Time (JIT), Robots, Computer Integrated Manufacturing (CIM), or Flexible Manufacturing System (FMS), to 3 The product and technology matrix is a slightly different version of the product-process matrix suggested by Hayes and Wheelwright [1984]. In this matrix, a process technology choice is determined by the combination of the degree of automation capacity and the degree of flexibility. The stages of PLC are represented by the third dimension in the matrix. improve manufacturing flexibility. Although the need for investment in new process technology has been widely recognized by most manufacturing firms as well as academicians there have been few studies concerning the technology acquisition problem in the broad context of manufacturing strategy. This study attempts to respond to this need by exploring the strategic implication of new advanced technology acquisition. Results of this study are expected to offer managers sound guidance for optimizing manufacturing systems' performance potential in the long run (Kantrow 1980). It has been argued frequently that the relatively weak performance of the US. manufacturing industry during the last two decades has been due partly to an obsession with short-term performance and to the attention given solely to cost efficiency by most production managers (DeMeyer et al. 1989; Kaplan 1986). A similar phenomenon can be observed in past analytic studies. Most of the studies employ cost efficiency as the single decision criterion. However, the importance of manufacturing flexibility has become recognized by both practitioners and academicians. In fact, some empirical studies have revealed that manufacturing flexibility is a key ingredient for the survival of manufacturing firms in the fierce competition of the international market (DeMeyer, et al. 1989; see also Buffa 1984; Jaikumar 1986; Goldhar and Jelinek 1983). In addition, manufacturing flexibility is regarded as one of the competitive priorities as opposed to cost efficiency in terms of the capabilities of process technology (Hayes and Wheelwright 1984). A firm with the competitive priority of manufacturing flexibility may focus on improving the flexibility of its process technology to achieve the economies of scope. On the contrary, a firm with the competitive priority of cost efficiency may focus on increasing the automation capacity of its process technology to achieve the economies of scale. Hence, if the objective of technology acquisition is to enhance a firrn’s overall competitive position in the market, the decision criteria for optimal performance should reflect the trade-offs involved in these two conflicting priorities. This view is shared by many researchers (Kaplan 1983; Jaikumar 1986; Hayes and Jaikumar 1988; Hough 1989) who assert that 4 the evaluation for new technology acquisition should include its intangible benefits. As suggested by Kim (1989), the studies in strategy can be placed into three categories: conceptual, empirical, and analytical. Most studies in manufacturing strategy will be categorized as either conceptual or empirical. Few studies will be included in the analytical category. In fact, the analytical study of manufacturing strategy is still in its infancy. One of the important benefits of analytic study is the high controllability of variables. This controllability permits the exploration of many hypothetical scenarios in a given study, which would allow the development of new directions from existing theories. The above view was supported by Robert H. Hayes during his presentation at the 1988 Decision Sciences Institute National Meeting. Hayes asserted that the study of manufacturing strategy should move to a new stage in which transportable and practicable research tools are developed. Many have also argued that analytic tools should be developed to examine the interrelationships among the various components of manufacturing strategy (see also Kim 1989; Adam and Swamidass 1989). In fact, a too simplistic or static representation of a strategic problem such as technology acquisition may distort the analysis of its full impact on overall performance measures, such as ROI, market share, total production cost, etc., and cause the decision maker to reject the proposal. Due to the greater power of computing available with the advancement of personal computers, numerical solutions to the complex model of optimal control theory, such as the one used in this dissertation, can easily be attained. In summary, the objective of this study is threefold. First, it examines the problem of technology acquisition, taking a broad-based view founded on the conceptual framework identified from the body of literature in manufacturing strategy. Second, this study employs realistic settings in the model that consider the trade-offs between two strategic priorities - flexibility and cost efficiency. Finally, the study develops a dynamic optimal control model to capture the strategic impact of the acquisition of new advanced technology on a SBU's performance in the long run. The purpose of the model is to 5 determine the strategic mix of automation capacity and degree of flexibility so that not only the SBU's strategic objectives but also its tactical and operational objectives are fulfilled. 1.2 RESEARCH ISSUES Strategic justification for investment in new process technology is a complex problem. The investment in new process technology is related not only to capacity planning but also to a firm’s competitive position in the market. The introduction of new process technologies, such as FMS, CHM, and Robots, is argued to promise great changes in the performance of manufacturing firms and even changes in industry structure (Jaikumar 1986; Hayes and Jaikumar 1988). In particular, many have asserted that investment in new equipment and technology is indispensable for improving the flexibility of a firm (Drucker 1971; Frohman 1982; Goldhar and Jelinek 1983; Hayes and Wheelwright 1984). Accordingly, this dissertation identifies and examines two key issues. The first is the relationship between the relative emphasis on flexibility and cost efficiency and the optimal acquisition of new process technology, given market conditions of product variety and demand growth along the product life cycle. The second issue is the effect of the optimal acquisition of new process technology on changes in the process technology choice within the product-technology matrix. This dissertation explores the strategic implications of these two issues in a dynamic environment. 1.2.1 Flexibility and Cost Efficiency of New Process Technology Primarily, this dissertation examines the relationship between a firm's strategic priorities and the optimal acquisition of new process technology along the stages of the product life cycle (PLC). Based on the literature review, this dissertation identifies two major reasons for the lack of strategic perspectives in past analytic studies of the new 6 process technology investment. The first one is the improper definition of new process technology, and the second is the failure to address the dynamic interaction between technology investment and strategic competitiveness variables, such as degree of automation capacity, level of overall system flexibility, market conditions, and manufacturing cost structure. Most past analytic studies for technology acquisition problems define process technology in units of output (Bustein 1986; Fine and Freund 1986; Gaimon 1985, 1986, 1989, and 1992; Roth et al. 1991). These studies fail to consider the strategic aspect of new process technology —- flexibility -- and thereby narrow their scope by disregarding the product variety aspect of market demand. In addition, the models in these studies usually have an objective function with a single criterion of cost performance. In this dissertation, however, new process technology is defined as the combination of automation capacity and flexibility. Automation capacity is defined as production capacity in units of output substituted for conventional old capacity (see Groover 1980). Also, flexibility is defined as the capability of the production system to produce a variety of products (parts). Hence, process technology of a manufacturing system is defined here in terms of both the automated capacity (degree of automation) and the level of flexibility (degree of flexibility). The acquisition of new technology, therefore, entails two separate decisions: the acquisition of automation capacity and the acquisition of flexibility. More significantly, this new definition of process technology allows the model here to consider the product variety aspect of market demand as an environmental variable. Many argue that justifying new technology acquisition based solely on cost savings is incongruous with the surmounting trend of increasing the strategic importance of new technology (Meredith and Hill 1987; Hill 1989; see also Kaplan 1983). Some even assert that the introduction of new process technologies, such as FMS, CIM, or Robots, seems to promise great changes in the performance of manufacturing firms and even in industry structure (Jaikumar 1986; Hayes and Jaikumar 1988). In particular, many have asserted 7 that investment in new equipment and technology is indispensable for improving the flexibility of a firm (Drucker 1971; F rohman 1982; Goldhar and Jelinek 1983; Hayes and Wheelwright 1984). Accordingly, they argue that intangible benefits of new technology, such as improved flexibility, quality, and delivery dependability, should be considered in the evaluation process for the acquisition of new technology. This dissertation maintains that the cost-benefit analysis for new technology acquisition should be linked to the strategic priority determined by the corporate strategy. By doing so, the technology acquisition decision can be made consistent with the objectives of the corporate strategy (Frohman 1982; Garrett 1986; Hayes and Wheelwright 1984; Skinner 1974). In this dissertation, the acquisition of automation capacity is assumed to create benefits such as savings in variable costs due to the elimination of direct labor (Hough 1989; Jaikumar 1986) and improved product volume flexibility due to increased production volume capacity. The acquisition of flexibility is also assumed to yield benefits of improved product mix flexibility. However, new technology acquisition accompanies enormous fixed costs of capital investment (see Miller 1985; Smith et al. 1986).. Therefore, it can be conjectured that if a firm's strategic priority is cost efficiency, the firm would want to invest less in new process technology than a firm with a strategic priority of flexibility. Hence, the acquisition decisions for new technology have to consider the proper balance between the fixed investment costs of new technology and its long-term benefits in order to be consistent with the firm’s choice between cost efficiency and flexibility as its strategic priority. Accordingly, the model developed in this study considers both cost efficiency and flexibility in the objective fiinction as two conflicting criteria measuring the long-term performance of a manufacturing firm's SBU. Specifically, the strategic priority of a firm is reflected by the values of cost coefficients assigned to measure flexibility in the objective firnction. Through the scenario analysis using the model, this study attempts to explore the effects of different strategic priorities on the optimal acquisition of new technology by 8 changing the values of those cost coefficients. In summary, this dissertation extends the analytic models of past studies by defining new process technology with two separate variables —- automation capacity and flexibility -- and attempts to identify the relationship between the relative emphasis placed on two competitive priorities -- cost efficiency and flexibility -- and the optimal acquisition of new process technology. 1.2.2 Acquisition of New Process Technology and Evolution of Technology Choice According to the normative concept of the conventional product—process matrix theory (Hayes and Wheelwright 1979a), the best "fit" on the matrix is determined by identifying a technology choice which has the optimal operating characteristics for a particular market demand in volume and product (part) variety. One major drawback of this conventional matrix theory is the definition of technology choices identified along the vertical axis of the matrix. This drawback is due to the fact that both automation and flexibility are considered in a single dimension. This theory views each technology choice with a fixed setting, such as a highly flexible system with a low degree of automation (Job Shop) and, conversely, a highly automated system with a low degree of flexibility (Transfer Line). However, it does not consider process technologies with any combination of both flexibility and automation such as a highly automated system with a high degree of flexibility. As Hayes and Wheelwright (1984) admitted, the product-process matrix is no longer congruous with the current trend of rapid diffusion of new process technologies which have the characteristic of high degrees of both automation and flexibility. Recognizing the inadequacy of the conventional product-process matrix, this study proposes a new matrix called a "Product-Technology Box” in which a technology choice is defined in terms of both the degree of automation along the horizontal axis and the degree of flexibility along the vertical axis. In this box, an additional third dimension is included to represent the dynamic environment of the product life cycle (PLC), reflecting 9 the changing market requirements in terms of both product volume and variety as does the horizontal axis of the product-process matrix. This study also recognizes the recent findings of the "Reversal of the Traditional PLC" ( The Reversed PLC), in which demand for both product volume and variety continue to increase after the grth stage without making a transition to the maturity stage (Abernathy et a1. 1983; Ayres and Steger 1983). Accordingly, the third dimension of the new box can be defined by either the Normal PLC or the Reversed PLC. The premise linking a technology choice and a stage of the PLC in the new box is the same as in the conventional product-process matrix. The process technology choice determines both the manufacturing cost structure and the variety of either products or parts the production system can handle (Hayes and Wheelwright 1984; see also Hill 1989). Meanwhile, market demand for a particular combination of volume and product variety at a certain stage of the PLC requires the proper operating characteristics of process technology. Therefore, this dissertation posits that the best position for a technology choice in the new box is different, depending upon the type of the PLC. The analysis in this study also considers the interaction between the firm's strategic emphasis on a particular competitive priority and the actual changes in the future market environment, such as aggregate market demand changes and the introduction of new products. For example, Firm A, whose strategic priority is flexibility, may purchase new process technology to improve flexibility. On the contrary, Firm B, whose strategic priority is low cost, may be reluctant to purchase new process technology unless savings from new investment is greater than the fixed investment cost. Since improved flexibility due to the acquisition of new process technology by Firm A may reduce the operating costs by the efficient use of materials, elimination of manual activities, and so forth, its overall cost performance may surpass Firm B's cost performance in the long run. However, the performance of these two firms may be affected by changes in the market environment such as demand volume changes and product mix changes. If future market 10 demand in volume grows quickly with little introduction of new products, it may not be appropriate for Firm A to assume a burden of excess investment costs for unnecessary flexibility. On the other hand, if the future market demand for product variety increases steadily, Firm A will enjoy the advantage of its flexibility while Firm B may lose its competitiveness in the long run. In summary, using a dynamic optimal control model, this dissertation explores the difference in the evolution of technology choice under the different market environments encountered in two types of product life cycle. In addition, it explores how the technology choice evolves in the new product-technology box according to the interaction effect between changes in the firm's external market conditions and changes in its relative emphasis on two competitive priorities -- flexibility and cost efficiency. 1.3 RESEARCH QUESTIONS Based on the issues discussed in the previous section, this section sets forth research questions for this study. According to the discussion in the previous section, the research questions are divided into two categories: (1) the acquisition policy for new technology and (2) the evolution of process choice. There are also subordinate research questions related to the major questions in each category. 1.3.1 Questions for The Acquisition of New Process Technology The major question for the acquisition of new process technology is, "How do the optimal acquisition policies for the new technology of a firm differ according to the choice of its strategic priority ?" Related questions include: a) Does a firm purchase more new technology when its strategic priority is flexibility compared to other priorities? b) In the long run, can a firm with a competitive priority of cost efficiency 11 achieve a lower cost performance compared to a firm placing primary emphasis on flexibility? 1.3.2 Questions for The Evolution of Technology Choice 1. One major question concerns how the evolution of process technology choice through the acquisition of new technology is affected by changing market conditions. For example, for a product life cycle characterized by a sustained growth stage and an increasing number of new products instead of the transition from a growth stage to a mature stage in the normal product life cycle, how does the evolution of technology choice relative to this new product life cycle differ from that of the traditional product life cycle? 2. Another major question is, ”How do changes in market conditions interact with a firm's strategic (competitive) priority to affect decisions involving the evolution of process technology choice?" Related questions include: a) For each of the two PLC types described in section 1.2.2, how does the evolution of technology choice differ due to different choices of strategic priority? b) Which choice of strategic priority is most insensitive to changes in market conditions? 1.4 SUMMARY In this chapter, the fundamental theme of this dissertation is described. In this dissertation, the problem of new process technology acquisition is evaluated analytically within the context of manufacturing strategy. Two critical issues are identified. First, this study explores the effect of a firm's relative emphasis on two competitive priorities -— flexibility and cost efficiency -- on the optimal acquisition policy for new process technology. Second, the study explores the interaction effect of strategic priority and 12 external market environmental changes on the evolution of process technology choice along the stages of PLC. This study deve10ps a dynamic based on the optimal control theory, which considers two conflicting objectives -- maximizing flexibility and minimizing cost efficiency. With factor settings identified in the body of literature, a scenario analysis is conducted to examine the research questions. The results are discussed to find possible extension of and new directions from the existing theories of manufacturing strategy. CHAPTER II LITERATURE REVIEW This chapter reviews the current body of knowledge related to both manufacturing flexibility and new advanced process technology. The major purpose of the literature review is to identify the relevant issues to fill the gap between the current body of knowledge and the desired path of future researches in this field. It is also to identify the relevant variables and concepts built in the theories of manufacturing strategy to extend the researches of the past into a new direction. This chapter reviews the literature related to the research questions raised in Chapter I as well as the research methodologies for the acquisition of new advanced technology. Section 2.1 reviews the past researches for the definition and the conceptual works of manufacturing flexibility. Through the review, it is possible to identify the definition as well as the measurement of manufacturing flexibility relevant to the research context of this study. Section 2.2 is devoted to review the past researches on the process technology and the product life cycle theory related to the manufacturing strategy. This section allows to identify the gap between the conventional normative consensus applicable to the old operating environment of manufacturing firms and the new perspectives of theory building under the current trend of the operating environment faced by today's manufacturing firms. The major contribution of the literature review in this section is to provide the theoretical background to develop the conceptual frameworks in Chapter III, as the main focus of this dissertation, which links the corporate strategy and the acquisition decision of new advanced technology. Section 2.3 reviews the past researches for the manufacturing automation. Again, 13 14 it attempts to reveal the proper definition of new advanced technology based on the current body of knowledge. This section also provides contributions in developing the conceptual frameworks in Chapter III by identifying the interrelationships among the attributes of new advanced technology and the manufacturing firm's internal and external operating environment. At the end of this section, several analytical research approaches taken by the past studies are reviewed to help developing a new analytic model in Chapter IV based on the conceptual framework developed in Chapter III. 2.1 MANUFACTURING FLEXIBILITY Flexibility is one of the key objectives of any manufacturing system (Chatterjee et a1 1984). Hayes and Wheelwright (1984) also considers flexibility as one of the dimensions of the competitive strategy of a business along with price(cost), quality, and dependability. Some researchers contend that flexibility is the next battle ground in international industrial competition (DeMeyer et al. 1989; see also Stalk 1988). Baranson (1983) argues that the global view held by Japanese companies towards marketing and production explains why Japanese managers there take a long-term and comprehensive view towards capital investments that consider not only cost savings but more significantly the broader strategic implications of increased flexibility. There are many different versions of the definition of manufacturing flexibility. The dimensions of manufacturing flexibility defined by numerous studies include such factors as volume, mix, machine, design, operation, process, and design. This section reviews the literature on flexibility to identify the proper measure of flexibility as the firm's performance. This section focuses on two key dimensions of manufacturing flexibility —- product mix flexibility and volume flexibility -- which have been regarded as major competitive priorities closely related to manufacturing flexibility at the strategic level. 15 2.1.1 Definition of Flexibility Although there are many different versions of the definition of flexibility, the term "manufacturing flexibility" is not well understood (Kumar 1986; Swamidass 1988). The general issue of flexibility can be traced as far back as the early 19205. In an economic context, flexibility was defined as "the risk arising from the immobility of invested resources" (Lavington 1921), "a relatively flat average cost curve" (Stigler 1939), and "the inverse relationship with the marginal cost" (Nelson 1962). Flexibility has also been examined in the Organizational Behavior literature, e. g, organizational flexibility (Feibleman and Friend, 1945), structural flexibility (Preece, 1986), or labor flexibility (Atkinson 1985).4 In the context of manufacturing, flexibility is defined as "the ability to respond effectively to a wide range of possible environments" (Gerwin 1987; Sethi and Sethi 1990; Slack 1983; Zelenovic 1982). Some define it in terms of resource capability, some in terms of constraints on manufacturing environment, and others in terms of both internal and external environmental uncertainties. In view of capability, Sethi et a1. (1990) defines manufacturing flexibility as the "ability to reconfigure manufacturing resources so as to produce efficiently different products of acceptable quality." Jaikumar (1984) asserts that manufacturing flexibility should be constrained within a domain defined by the portfolio of products, process, and procedures and the domain should be planned, managed, and with learning expanded. With regard to environmental uncertainties, Garrett (1986) suggests that manufacturing flexibility is required in order for a firm to cope with both internal changes and external forces. The internal disturbances for which flexibility is useful include equipment breakdown, variable task times, queuing delays, rejects, and rework (Buzacott and Mandelbaum 1985). External forces refer largely to the fiindamental uncertainties of the competitive environment (Behrbohm 1985; Zelenovic 1982; Garrett 1986; Maier 1982). 4 For the details of these definitions, see Sethi and Sethi (1990). 16 Kim (1989) defines manufacturing flexibility at three different levels, strategic, tactical, and operational. According to his typology, manufacturing flexibility at the strategic level reflects environmental uncertainties (see also Kamarkar and Kekre 1980; Zelenovic 1982; Slack 1983). At the tactical level, manufacturing flexibility is recognized as a concept conflicting with efficiency (see also DeMeyer et al. 1989). Here, manufacturing flexibility is viewed in terms of trade-offs between economies of scale and economies of scope (see also Cohen and Lee 1985 ; Goldhar and J elinek 1983). Kim points out that both product mix and volume flexibility are identified as two competitive priorities closely related with manufacturing flexibility in both the strategic and the tactical levels. Kim argues that these two flexibilities well reflect the strategic priorities imposed by a firm's external environments (see also Swamidass 1986). Two of the main external uncertainties faced by manufacturing firms will be the demand changes and the introduction of new products in the market. Accordingly, Kim asserts that the response to these uncertainties is properly captured by the definition of these two flexibilities. Finally, at the operational level, manufacturing flexibility is recognized as operational capabilities provided by a particular process technology (see also Browne et al. 1984). According to the literature, the definitions of manufacturing flexibility in the strategic context usually imply "responsiveness to uncertainties." Sethi and Sethi (1990) states "The probabilistic nature of these uncertainties may not always be known. Uncertainties may exist for level of demand, product prices, product mix, and availability of resources. Uncertainties may arise out of actions of competitors, changing consumer preferences, technological innovations, new regulations, etc.” Hence, manufacturing flexibility has major implications for a firm's competitive strength, and this significant role of manufacturing flexibility makes it a part of the firm's strategy. In this regard, by recognizing demand changes and the variety of new products as two major external uncertainties, this dissertation includes both product mix flexibility and volume flexibility as key competitive priorities of a manufacturing firm in coping with these uncertainties. 17 2.1.2 Manufacturing Flexibility and Its Strategic Implications As stated above, there are many dimensions of flexibility identified in the literature. Even the definition for each dimension of flexibility differs among authors. This section traces the proper definitions of product volume and mix flexibilities for the study in this dissertation. Furthermore, manufacturing flexibility entails not only the direct benefits of quick and economic response to environmental uncertainties but also other indirect benefits that might be overlooked. Thus, this section explores the strategic implications of flexibility for a firm's competitiveness in the long run. 2.1.2.1 Definition of Product Mix Flexibility and Volume Flexibility In this dissertation the main theme is characterized as the strategic analysis of process technology. Accordingly, product mix and volume flexibilities will be defined in the context of manufacturing strategy. In the literature of manufacturing strategy these two flexibilities have been frequently referred as primary types of flexibility at the strategic level (Cox 1989; Kim 1989; Schonberger 1986; Skinner 1974). Product mix flexibility is "the ability to respond inexpensively and rapidly to additional or substituted products" (see also Gupta, et a1. 1989). Sethi and Sethi (1990) defines product mix flexibility as the ease with which new parts can be added or substituted for existing parts. This definition is very similar to the definition of product flexibility of Browne et al. (1984), part flexibility of Gerwin (1982) and Falkner (1986), and design adequacy of Zelenovic (1982). According to these definitions, product mix flexibility is referred as "the capability of process technology to respond to part changes or part addition." It is important to note that in Lim's survey of FMSs in the United Kingdom (1987), 11 out of 12 reporting companies considered manufacturing flexibility to mean product mix flexibility. Product volume flexibility is defined as "the ability to operate profitably at 18 different overall output levels” (Gupta, et a1 1989, Sethi, et al. 1990). Similar definitions can be found in other studies (Browne et a1. 1984; Gerwin 1982; Maier 1982; Kegg 1984). This definition is also similar to the definition of demand flexibility by Son and Park (1987). According to Slack (1987), volume flexibility has two aspects: speed of response and range of variations. He suggests that the former is usefiil in the short term while the latter is useful in the long run. Buzacott (1982) defines flexibility with some strategic flavor. He separates flexibility into job and machine flexibility considering the nature of the change and disturbances that the system should be able to cope with. Job flexibility is defined as "the ability of the system to cope with external changes such as the type, mix, processing requirements, and quantity of jobs allocated to the system," while machine flexibility is related to the internal changes such as machine breakdowns, variability in processing times, and quality problems. According to his definition, both product volume flexibility and product mix flexibility reflect his definition of job flexibility. 2.1.2.2 Strategic Value of Manufacturing Flexibility According to a recent research, conducted by The Manufacturing Roundtable of Boston University (Miller and Roth 1988), flexibility was ranked from fourth to eighth in importance for future competitiveness, and first in the size of strategic gap (i.e., the difference between current capability and firture needs). However, flexibility did not appear at all in a list of 10 key performance measurements offered by the responding executives. The authors give two primary reasons for this discrepancy. First, in contrast to cost, delivery, and quality, which have been the cornerstones of manufacturing planning and control for many years, the idea of flexibility as a top priority has only recently come to the fore. Consequently, it tends to be treated, even on a conceptual level, less ofien and usually on a somewhat abstract as opposed to concrete basis. Second, and partly because of the first, the technology for measuring flexibility is poorly developed (see also Cox 19 1989) On the contrary, Baranson (1983) argues that the global view held by Japanese firms towards marketing and production explains why Japanese managers take a long-term and comprehensive view towards capital investments considering not only cost savings in labor, material, and space, but more significantly the broader strategic implications of increased flexibility and versatility in designing and producing of products (see also De Myer et al. 1989). Empirical evidence also supports the view that flexibility does not get its proper due at the time of decision making with regard to investment in manufacturing technology (Lim 1987). Hence, it is critical for the evaluation of technology investment to consider the long term strategic implications of enhanced flexibility due to the acquisition of new technology. The current trend of market and industry structure is being shaped by the adaptation of new advanced manufacturing technologies, rapid-response systems to environmental uncertainties and complexity, expanding variety and increasing innovation. The competitive values of manufacturing flexibility lie in its ability to neutralize the effects of demand uncertainties (Swamidass 1986; Swamidass and Newell 1987), to increase market share through the proliferation of new products, and to achieve low cost production with the economies of scope (Hill 1988). Stalk (1988) asserts that companies adopting strategies based on flexible manufacturing are reducing if not eliminating delays and using their response advantages to attract the most profitable customers. By characterizing the flexible firm as a time-based competitor, Stalk further asserts that the flexible firm achieves lead time reduction with small lot size and, hence, enjoys big advantages in both productivity and time: labor productivity in time—based factories can be as much as 200 percent higher than in conventional plants; time-based factories can respond eight to ten times faster than traditional factories. For example, Toyota had its suppliers reduce production lead time from 6 days to 1 day by reducing lot sizes and the number of inventory holding points (Stalk 1988). 20 In particular, a firm with high product mix flexibility can simultaneously pursue both a differentiation and a low cost strategy (Hill 1988). High product mix flexibility can allow the firm to introduce more products to the market than can its competitors. In the long run this will increase the brand loyalty of its customers and, consequently, increase the demand (see Roth et al. 1991). Hence, the firm can achieve the objective of the low cost strategy due to the economies of scale with the increased demand. An empirical study by White (1986) shows a significant and positive relationship between differentiation and low cost (see also Phillips et al. 1983). White found that 19 of 69 business units had a competitive advantage based on a combination of both differentiation and low cost. Moreover, his result suggests that business units that successfully combined both low cost and differentiation had the highest return on investment. The strategic values of both product volume flexibility and mix flexibility referred in the literature can be summarized as follows: first, the ability to neutralize the demand uncertainties; second, lead time reduction; third, low cost production; finally, increased market share. The latter two values explain how firms can pursue both a differentiation strategy and a low cost strategy simultaneously. 2.1.3 Measurement of Manufacturing Flexibility Since the survival of manufacturing firms will depend on their ability to adapt to rapid change in the market and industry, flexibility plays a key role not only as a strategic goal but also as one of the important means for performance measurement (Gupta and Goyal 1989) reflecting the changing source of competition (Stalk 1988). However, since flexibility, as Slack (1983) points out, is an indication of potential it is difficult to measure (see also Son and Park 1987). This is why there are so many different definitions of flexibility. Before defining the measurement of flexibility it is important to identify the key elements that constitute it. Another important aspect of measurement is how to measure the manufacturing system's overall flexibility as a performance in the dynamic production 21 environment. There are two major elements in measuring manufacturing flexibility identified in the literature, time and cost. Gupta and Goyal (1989) states that manufacturing flexibility has been measured by cost, delivery speed, or lead time (see also Cox 1989). Particularly, product mix flexibility can be measured by time or cost required to switch from one part to another, not necessarily of the same part types (Browne et al. 1984; Buzacott 1982; Zelenovic 1982). Son and Park (1987) suggests a measurement of product mix flexibility as the ratio of the physical output of the system to the setup cost of the equipment, viewing the reduction in setup costs as the way to increase product mix flexibility. As a generalized measure developed from the definition of product volume flexibility of Browne et al. (1984), Gupta and Goyal (1989) suggests "the range of volumes in which the firm can run profitably. " Gerwin (1987) measures it by "the ratio of average volume fluctuations over a given period of time to the production capacity limit." Falkner (1986) suggests, as a measure of volume flexibility, "the stability of manufacturing costs over widely varying levels of total production volume." Son and Park (1987) suggest measuring it with "the ratio of the physical output to the inventory cost of finished product and raw material," indicating the difference between supply and demand, that is, response to internal and external demand. Although a single comprehensive measure of manufacturing flexibility takes into account its direct benefits such as quick and economic response, it should be carefirlly designed with a broad perspective since the benefits of flexibility are also reflected indirectly by the improvement in other performances, such as inventory cost reduction or improved market position. In the past literature, there are few examples of using a single quantified measure of manufacturing flexibility in researches on manufacturing strategy. Nagarur (1992) presents a simple approach to measuring the flexibility of FMS. He introduces an index, which is called "producibility," which measure the reroutability of any excess load of a machine center to other centers. This flexibility index may be useful in 22 measuring the overall operational flexibility of production system. A recent study by Ramasesh et al. (1992) employs the ratio of the net revenue generated by the system to the standard deviation of the net revenue distribution as a measure of aggregate flexibility. In their mixed integer stochastic program, the objective function generates ’optimal production levels for a set of products to maximize the net revenue considering costs associated with product mix and product volume flexibilities. Beside some conceptual drawbacks of this particular model, it is hard to expect that the solution of the model may guide any significant strategic implications, rather it may be suitable for a comparison purpose for different systems under a static environment. Relevant to the study of this dissertation, Gupta and Goyal (1989) suggests that a single measure for manufacturing flexibility should be concomitant with a given manufacturing strategy. Then, the manufacturing system could be designed and modified accordingly. Importantly, it is necessary to develop a measure of overall manufacturing flexibility under a specific production setting of process technology and production control system, considering the economic consequences of future environmental changes (see Buzacott 1982). As Gupta and Goyal (1989) asserts that flexibility is not a self-contained concept, it may be necessary to relate manufacturing flexibility as a performance measure to the other production objectives related to manufacturing cost and time such as operating cost and lead time. Hence, Gupta and Goyal imply that the system's flexibility will be reflected in the long term performance measured in both operating cost and lead time through constant changes in process technology and control system according to the environmental changes. 2.1.4 Trade-off between Flexibility and Efficiency The conflict between flexibility and cost efficiency primarily account for by the high initial investment of new technology which enhance flexibility. For example, a numerical controlled turning center costs $300,000 to $400,000, whereas an FMS 23 installation can balloon to $25 million (Bobrowski and Mabert 1988). For this reason, flexibility and cost efficiency have been considered as conflicting objectives (De Meyer et al. 1989). By defining flexibility with two conflicting factors, "Quick Response to Change" and "Economic Response to Change," Chung and Chen (1989) conceptualize flexibility as the compromise between these two factors. They provide a simple formula for conceptualizing a framework for flexibility evaluation, The Total System Flexibility = orQ + (1- or)E, with or for a weight, Q for a factor of the quickness, and E for a factor of the economic response. However, as discussed before, a firm with high flexibility may pursue successfully both differentiation and low cost strategies simultaneously. In particular, a firm with high product mix flexibility can introduce more new products into the market than can its competitors and, consequently, enjoy the economies of scale through the increased market share (Hill 1988). Hill (1988) suggests that, if the cost reduction due to the increased market share outweighs the increase in the cost related to product variety, the firm can pursue successfirlly both differentiation and low cost strategies at the same time. In fact, some empirical studies (Newell and Swamidass 1987; Meyer et al. 1989; Phillips, et al. 1983; White 1986) provide evidence that there is a positive relationship between MF and other manufacturing performance measures such as cost reduction and/or quality improvement. In summary, the cost implication of manufacturing flexibility entails two aspects. On one side, flexibility will increase the cost due to the huge initial capital investment and the potential increase in variety of products. On the other side, however, flexibility will reduce the cost due to the efficient use of labor and materials and the increased market share resulting from a firm's enhanced competitive position in the market. Hence, in deve10ping a model for the study of flexibility it is important to consider not only the fixed investment cost in improving flexibility but also the other benefits related to cost reduction 24 due to the improvement in flexibility. 2.1.5 Summary and Implications It seems that the flexibility of advanced process technologies can become easily submerged in technical requirements, thereby disregarding a holistic view (Sethi, et al. 1990). This View is consistent with the result of an empirical study by Slack (1987) in which the managers are found to focus more on flexibility of individual resources than on the flexibility of the production system as a whole. It is also found from the literature that manufacturing flexibility has a significant long-term strategic impact by enhancing a firm's competitive position in the market. However, the literature concerning manufacturing flexibility needs empirical studies with raw data from industries to identify the functional relationships among the means of achieving manufacturing flexibility such as investment in new process technologies and manufacturing parameters such as lead time, fixed and variable costs. These studies are indispensable as research background to aid the conceptual and theoretical studies in analyzing the strategic implications of investment in new process technologies on the firm's competitive position in the market. In summary, the implication for this dissertation can be summarized in three major points. First, both product mix flexibility and product volume flexibility, as performance measures for the analysis of strategic issues, are found properly related to and should reflect the contingencies of the corporate strategy. Second, manufacturing flexibility is found to have a significant impact on a firm's market position with potential market share growth and the cost structure of the firm's production system. Finally, it is also found that the variables related to both manufacturing time and cost representing the long-term effects of flexibility should be used for strategic evaluation of manufacturing flexibility. 25 2.2 PROCESS TECHNOLOGY AND PRODUCT LIFE CYCLE The first part of this section will be devoted to exploring the literature related to process technology. The purpose of the review is to find, first, how the past studies categorize different process technologies as defined in the product and process matrix theory (Hayes and Wheelwright 1984) and, second, how this categorization scheme can be applied as a theory of manufacturing strategy in the emergence of new advanced process technologies. The current trend of the international market, characterized by a great variety of custom-made products, shorter or truncated product life cycle, and the expectation of higher standard by customers, is quite different from the conditions in the past, which the theory has been constructed from and applied to. Furthermore, the greater emphasis on flexibility in every aspect of manufacturing process also steers the strategic priority from cost efficiency to flexibility. With the notion of these changing market trends and strategic perspectives, it would be necessary to revisit and reevaluate the conventional theory of manufacturing strategy. 2.2.1 Process Technology in Manufacturing Strategy Technological change is not important for its own sake, but it is important if it affects competitive advantage and changes industry structure as asserted by Porter (1980). It plays a key role in linking the manufacturing capabilities which it embodies and marketing requirements and, consequently, it determine the company's competitive position in the market. Simmonds (1981) also states, while pointing out the drawbacks of the conventional accounting approach of investment appraisal, that new production investment must imply a change in competitive position and it is this change that should be the focus of the investment review. 26 2.2.1.1 Characteristics of Process Technology There are many studies related to the definition of process technology (Buffa 1984; Hayes and Wheelwright 1984; Hill 1989). Buffa (1984) identified four different process technologies as physical systems in positioning strategy -- a product-focused system for high volume, a product-focused system for multiple products in moderate volume, a process-focused system for moderate to low volume, and a process-focused system for custom products. The definition of these processes is not much different from the conventional definition of job shop, batch, line, and continuous processes (Hayes and Wheelwright 1979a). The distinction among these four process technology has been made by the difference in two key aspects -- degree of flexibility and level of investment in hard automation. Buffa explains the rationale for the investment in hard automation as a mean of achieving low cost objectives with high volume products. Buffa also relates flexibility to the product mix to distinguish those four systems. Hence, Buffa identifies flexibility and automation as key dimensions of process technology. Table 2.1. Characteristics ofProcess Choice Process Choice Process Automa- Overhead Direct Direct flexibility tion cost labor Material Job Shop High Flexible Low High Low Batch Assembly Continuous Low Hard High Low High Source: Buffa (1984) and Hill (1989) A detailed description of the characteristics of these process choices is provided by Hill (1989). In Table 2.1, an anecdotal summary is presented for some of key characteristics of process choices extracted from two studies (Buffa 1984, Hill 1989). 27 These studies are related to the study of this dissertation. As shown in Table 2.1, compared to continuous process, job shop process has higher process flexibility. In general, job shop process employs flexible technology, whereas continuous process employs hard automation to lower the variable production cost. In terms of cost structure, job shop tends to have a relatively low fixed cost compared to high variable cost, whereas continuous process tends to have high fixed cost compared to low variable cost. The different cost structures of the four process types are mainly due to the natures of the process technologies they adopt. Hill (1989) maintains that job shop process involves the high content of direct labor, thus making it the highest portion of total costs. On the contrary, due to the high process investment, direct labor costs for continuous process are small and its site/plant overheads are high owing to the need to support the process and handle the high output levels involved. However, an empirical study by Raffr and Swamidass (1987) shows that the ratio of manufacturing overhead costs (MOHC)5 to total manufacturing cost is not significantly different among process choices. There are not many studies dealing with the process choices in the past. Nevertheless, it is generally found that the process choices are usually characterized through the relative comparison in terms of cost and flexibility (Buffa 1984; Hayes and Wheelwright 1979a; Hill 1989; Raffr and Swamidass 1987). 2.2.1.2 Process Technology in Capacity Planning One of the critical issues in the studies of the acquisition of new technology is the pattern of change, frequently called as timing -- gradual transition, incremental, from old to new technology, or upheaval, radical, of old technology. Yet, there is no clear—cut conclusion on this issue drawn from the literature. Gaimon(l985) in a study of the 3 The major component of MOHCs include, indirect supplies, indirect labor, supervisory salaries, social security taxes, pension payment, health care costs, overtime premiums, idle time costs. vacation pay, depreciation, property taxes, property insurance, repairs and maintenance, power costs. and material handling. 28 optimal acquisition of automation supports the incremental changes of technology. Gaimon used this view as an assumption in developing her model in the study of the optimal acquisition of automation (see also Roth et al. 1991). On the contrary, Vickson(1985) found in his study that the optimal transition from old to new technology has to be instantaneous. He drew his View from the conclusion in the study of the investment choice between two production technologies. Unfortunately, this conclusion is almost predictable from one of the constraints in his model. Skinner (1974) also points out that the incremental change in production system and organization structure as one of the major factors causing confusion and inconsistency in making strategic decisions and leading to poor performance. Since both Vickson's and Gaimon's models are too simple, missing the dynamic relationships among many important aspects of manufacturing system, it may not be appropriate to draw any conclusion from their studies. Even though Skinner's assertion is based on broader perspective than Vickson's, it does not reveal the details of strategic implication of investment timing in the process of manufacturing strategy formulation. The other critical issue is related to the evaluation criteria of technology investment. As Hill(1989) points out, the excessive use of ROI distorts strategy building (see also Madu and Georgantzas 1991). He suggests that only when a company reviews investment in light of its corporate strategy, marketing strategy, and manufacturing strategy, the essential cohesion will be established. Until this happens, the company will be in danger of investing in ways that will not give the necessary synergetic gains of strategic coherence. Hence, the evaluation criteria for technology investment should be concomitant with the congruent objectives of those strategies. Particularly, evaluation criteria should be associated with the firm's competitive priority based on the forecast of the contingencies of the firm's operating environment. 29 2.2.1.3 Process Technology in Manufacturing Strategy Good process technology strategy can be characterized by fit or consistency, which ensures that the firm's process technology evolves in a directed fashion, so that as technological capabilities are renewed and augmented, they reinforce and expand the firm's competitive position (Hayes and Wheelwright, 1984). This fit can be understood by identifying general capabilities of process technology not only with existing products but also with new products to be introduced and the changes to the existing products in the dynamic environment. (Hayes and Wheelwright, 1984). Hence, the analysis of process choice for a firm should be carried out with the understanding of the dynamic changes of the firm's internal and external operating environment. As discussed in the previous section, another important observation in past studies of the evaluation of process technology is criticism of the use of a single performance criterion (Kaplan 1983; Simmonds 1981). Many of those earlier studies analyze their problems with a single criterion of cost minimization or profit maximization by including variables defined with narrow and abstract perspectives.6 As a result, most of these studies fail to characterize adequately the nature of different process choices that are frequently referred in manufacturing strategy literature and, consequently, fail to address the relevant strategic issues related to the acquisition of new advanced process technology. In particular, more researches is called for in the area of identifying key strategic variables affected by the adaptation of new advanced technology. An effort by Ettlie et al. (1984) shows empirical results identifying the organizational causality of two different innovation processes, radical and incremental. What is needed for the strategic evaluation of new advanced technology is empirical study such as this to identify the critical organizational factors in the causal relationships between decision variables related to the acquisition of new advanced technology -- timing and size -- and a firm's overall 6 The detailed review of these studies will be presented later in section 2.3.3. 30 performance including not only the strategic but the operational. 2.2.2 Product Life Cycle in Manufacturing Strategy 2.2.2.1 Product Volume and Product Mix in the Product Life Cycle Key elements of the conventional product life cycle relevant in manufacturing strategy are product volume and product mix. Four distinguished stages -- introduction, growth, maturity, and decline -- are characterized by the different ranges of product volume and product mix. Hayes and Wheelwright (1984) summarize the framework of Wasson (1978) with four dimensions that are directly linked to manufacturing: production volume, product variety, industry structure, and the dominant form of competition. They suggest that while the product life cycle is usually primarily in planning a firm's marketing strategy, it also has great implication to the firm's manufacturing strategy. For example, manufacturing has a major stake in decisions that may affect such variables as product customization (versus standardization), volume per model, and the average time before obsolescence or replacement. Given this perspective, the product life cycle can be used to summarize the customer and product requirements that must be satisfied by the manufacturing function and its production technology. They also assert that, by identifying the relevant range of product volume and variety for each stage of the product life cycle, the stage can be matched to the corresponding stage of the process life cycle according to the capability the production system can provide. As a result, they developed the famous "product-process matrix." Hill (1984), in his conceptual framework called "product profiling," also related the relevant range of volume and variety to process choice as one of the many operational aspects as a best fit between the marketing strategy and the manufacturing strategy. Even though the concept of the PLC theory has been well accepted in the area of POM until now, the concept of the traditional product life cycle has been challenged by some researchers in marketing area (Ayres et al. 1983; Dhalla et a1 1976). Particularly, Ayres and Steger ( 1983) asserts that, with the rapid evolution of technology, the product life 31 cycle can be reversed. 2.2.2.2 Manufacturing Cost in PLC Stalk (1988) suggests that manufacturing costs fall into two categories: those that respond to volume or scale and those that are driven by variety. He reports that scale- related costs decline as volume increases; usually falling 15 percent to 25 percent per unit each time volume doubles. Variety-related costs, on the other hand, reflect the costs of complexity in manufacturing: setup, material handling, inventory, and many of the overhead costs of a factory. In most cases, as variety increases, costs increase, usually at a rate of 20 percent to 35 percent per unit each time variety doubles. Stalk suggests that the optimum cost point for factories can be decided by the combination of volume and variety (see figure 2.1). + Scale-related Costs +Variety-related Costs +Total Cost of Operation Product Cost Low Mediu High m Volume, Variety Figure 2.1 Variety-Related Costs and Scale-Related Costs An empirical study of metalworking industries by Ayres and Miller (1981) shows evidence that the unit production costs increases as the batch size increases. The study 32 estimates that the unit costs are about ten times higher for custom production in a job shop environment than they are for mass-production operations. It also estimates that the unit costs in the intermediate batch processes range from three to five times higher per pound of material processed than in the most extreme mass production case. 2.2.2.3 Reversal of PLC In contrast to the traditional product life cycle concept, Ayres and Steger(1983) suggest the reversibility of the biological evolution of the product life cycle. They argue that a "product" may not move through the cycle from birth to death in one direction only and there are a number of historical examples of reversals in the sequence, attributable to technological changes that altered the product significantly but did not replace it. Ayres and Steger also suggest three conditions for the reversal of the product life cycle are potential for accelerated technological change, management flexibility, and manufacturing flexibility. The authors also argue that the evolution process can be accelerated or decelerated depending upon a firm's competitive advantage. If a firm is competent in mass production, it will be beneficial to accelerate the product standardization and maturity. On the contrary, if a firm has competence in technological flexibility, then it will be safe to adopt the deceleration strategy (see also Tombak 1988). Abernathy, et al. (1983) call this reversal of the life cycle "dematurity" in their framework of the product life cycle. They argue that major technology changes can throw an industry back into a growth stage. After close examination of the US auto industry, the authors found that the auto industry served as a prime example of this process of "de- maturation." The immature standardization efforts by the firms lock them into the stage of "rigidity," in which they will ultimately lose their competitiveness in the face of rapid evolution of technology. The "dematuration" process will deter the usual process of product standardization from capturing the advantage of the economies of scale. Rather, it will proliferate the introduction of new products in the market and make the corporate 33 managers rethink their competitive strategies. Another interesting empirical study by Dhalla and Yuspeh (1976) suggests that the universal application of the PLC concept may be misleading. They found that, while it is difficult to reverse the trend for the products in a final declining stage, no firm brand behavior can be found with the first three stages of the PLC. The authors give a list of examples that had been for a long time but were still full of vitality and whose sales went up and down but stayed in the growth phase. These anecdotal and empirical evidences suggest that the evolution of the product life cycle is not given as conventional PLC theory. Instead, the product volume and mix are truly the strategic variables whose behaviors are unpredictable. One interesting inference from the literature is the difference in the width of product variety between the growth stage of the reversed life cycle and that of the usual life cycle. The width of product variety in the growth stage of the reversed life cycle will be much greater than that of the usual life cycle. Hence, as the reversal of the product life cycle will generate more uncertainty into the game of competition a firm must recognize the importance of flexibility in enhancing its strategic competitiveness in the market. Potential reversals will occur with substantial benefits to those who carefully prepare through decisions regarding technology and management planning in a broad strategic fashion. 2.2.3 Integration of Product and Process Decisions 2.2.3.1 Product-Process Matrix Integration of the product life cycle and the process life cycle has been best described by "fit" (Hayes and Wheelwright 1984) or "alignment" (Pearson et a1 1991) based on the market requirements imposed by a stage in the product life cycle and its corresponding process capabilities to generate the optimal performance in meeting those requirements (Pearson et a1 1991). Empirical studies (Abernathy 1978b; Hambrick et al. 34 1982a and b; Shipper 1981) have found that the production process life cycle begins with a fluid, uncoordinated configuration and evolves first into a segmented and finally into a specific, systematic configuration. This transition of the process life cycle corresponds to the evolution from job shop, to batch, to assembly line, and finally to continuous flow production. A conceptual framework to match properly the evolution of the process life cycle by the evolution of the product life cycle is the product and process matrix (Hayes and Wheelwright 1979a and b) as depicted in Figure 2.2.7 Process Choice Project 0 Job Shop 0 Batch 0 0 Continuous: 0 Process ow Volume High Volume Unique, Variety Standardized Product Volume and Variety Figure 2.2. Product and Process Matrix Hill (1989) presented almost the same integration scheme with a different approach, which is called "Product Profiling." He identifies important elements of both marketing strategy and manufacturing capabilities and characterized the process choices according to those elements. The process capabilities are matched with the market requirements prescribed by the business chosen, mainly product range and volume. In this 7 For detailed reasonings for the match, refer to Hayes and Wheelwright [1989]. 35 respect, this method is based on the conceptual framework of the product-process matrix. However, he attempts to extend the product and process matrix by linking market requirements and process capabilities through a concept called "order winning criteria," which is similar to the concept of competitive priority. He also asserts that the transition described in the product-process matrix does not appear as a liner sequence (Hill 1989). Firms may follow an iterative, repetitive course, jumping forwards and backwards at different times according to their perceptions of market demands in the future and the corresponding process design issues (Pearson et al 1991). For example, a firm will have a job shop type of manufacturing process for one of its products. As volume increases a transition to a batch type of process may occur. When volume decreases, towards the end of the product life-cycle, the transition back to job-shop may take place (Hill 1989). According to Hill's argument and those supporting the reversal of PLC, the process choice may be very much dependent on the future market situation projected by the firm. Based on past studies, it can be presumed that there may be two types of the evolution of process technology. First, given the usual evolution of a product's life cycle as being true and the theory of the product-process matrix as being accepted, the firm's process technology will cycle with the transition from the flexible system, to the dedicated and inflexible system, back to the flexible system, and so on. Second, if the product life cycle stays in the growth stage by the reintroduction or design updates of the product (Smith 1980; Ayres and Steger 1983), the firm may always focus its investment more on the flexible manufacturing technology instead of blindly switching to the dedicated system. According to the current trend of market conditions and technology development, the former is not as persuasive as the latter. Thus, the investment in process technology should be carefully planned according to the dynamic changes of the market requirements in the firture. 36 2.2.3.2 Effect of Flexible Technology on Product-Process Matrix The recent development of flexible technology allows firms to produce a wider range of product variety and volume at the same or lower cost than does conventional technology(Goldhar and Jelinek 1983; J elinek and Goldhar 1984; see also Suresh and Meredith 1985). Hence, flexible automation has challenged the managerial wisdom of the past decades that was captive to in the paradigm of product-process matrix theory (Adler 1988). It seems that the matrix's normative implication is no longer applicable to any process choice with the higher level of flexible automation. Process Choice Job Shop Batch Assembly Line Continuous K Flow \ w Unique Standardized Low Volume High Volume Figure 2.3. Effect of Flexible Automation on the Product-Process Matrix In a dynamic perspective, it is necessary to consider the implications of recent flexible automation, which make it possible to view the production of less standardized products in a "quasi-continuous process" (Adler 1988). Hence, a new matrix's diagonal would be flattened out, as in Figure 2.3 (Boothroyd 1982; see Hayes and Wheelwright 1984). In fact, Hayes and Wheelwright (1984) discussed the impact of flexible technology as one of the limitations for the application of the matrix. They states that "such improvements in production flexibility, without movement along the diagonal, might be thought of as a third dimension to the matrix." Hence, if we separate this new dimension 37 from their product-process matrix, it will be possible to draw a new matrix as shown in Figure 2.4. In this new matrix, the vertical axis represents the degree of flexibility of the new process technology instead of the conventional category of process choice. Figure 2.4 depicts the wide range of variety and volume covered by flexible technology, whereas rigid technology covers only standardized and high volume products. Flexible technology Rigid technology Low volume High volume Unique Standardized Figure 2.4. Integration of Technology and PLC One of the major implications of such a revision is the undermining of a widely felt "intuition of a corollary between efficiency and rigidity” (Adler 1988). It can denounce the proposition, as suggested in an empirical study by Abernathy (1978b), that there is a fundamental dilemma between innovation and efficiency. The capability to produce a wide variety of products at low cost is a major advantage of flexible technology. Hence, as discussed in section 2.1.4, the most significant implication of the introduction of flexible technology like FMS, CHM, and CAM is the coexistence of differentiation and low cost as a firm's strategic focus. 2.2.4 Summary and Implications According to the review of literature, the potential paths for the market/product 38 evolution may not be given as the main propositions of the PLC theory. As shown in Figure 2.5, there may be four possible states of the market/product requirements a firm at any stage of the PLC may encounter in the future. Each of these four possible states may be reflected by a firm's overall competitive strategy as a combination between differentiation and cost leadership (Porter 1980). Hence, in this dissertation, the four states given in Figure 2.5 will be distinguished by two exogenous variables -- volume and variety. Another finding by the literature review in this section is the possible coexistence of differentiation and low cost strategies, which means the simultaneous pursuit of both economies of scope and economies of settle by the adaptation of flexible process technology. In this dissertation, flexibility and cost efficiency are chosen as competitive priorities representing differentiation strategy and cost leadership strategy, respectively. Through the scenario analysis with a dynamic control model, the study will attempt to find which strategy is suitable for each of the four possible states in Figure 2.5. The study will also explore whether the coexistence of differentiation strategy and cost leadership strategy is possible and under what condition it is possible. Volume Volume Increasing Decreasing Variety Increasing A 8 Variety . C D Decreasung Figure 2.5. Four Stages of the Possible Evolution of the PLC 39 2.3 NEW PROCESS TECHNOLOGY AND MANUFACTURING FLEXIBILITY 2.3.1 New Process Technology 2.3.1.1 Automation of New Process Technology Automation is defined as "the technology concerned with the application of complex mechanical, electronic, and computer—based systems in the operation and control of production (Groover 1980). A very similar definition of automation can be found in the Encyclopedia Britannica; " a significant substitution of mechanical, electrical, or computerized action for human effort and intelligence a technology concerned with carrying out a process by means of programmed commands capable of operating without human intervention." Groover (1980) states that automation includes: (1) automatic machine tools for processing work parts; (2) automatic materials handling systems; (3) automatic assembly machines; (4) continuous-flow processes; (5) feedback control systems; (6) computer process control systems; and (7) information systems to support manufacturing activities. Groover also lists the economic and social reasons for automating which include the following: (1) increased productivity; (2) high cost of labor; (3) labor shortages; (4) trend of labor toward the service sector; (5) safety; (6) high cost of raw materials; (7) improved product quality; (8) reduced manufacturing lead time; and (9) reduction of in-process inventory. 2.3.1.2 Rigid Automation and Flexible Automation Traditionally, the manufacturing system has been represented by two kinds of equipment. The first one -- dedicated machinery, such as transfer lines -- is best suited for mass production of a single part. This process specialization permits low unit costs, but it inhibits flexibility. The second kind of equipment -- nonintegrated general purpose 40 machine tools -- is best suited for the very small batch production of different parts (Gupta and Goyal 1989). Due to the high degree of flexibility, costs per unit tend to be high, but the flexibility of the process can accommodate design changes, demand fluctuations, and shifts in product mix. These definitions -- dedicated and general purpose -- are applicable to the concepts separating process choices used in Hayes and Wheelwright's product-process matrix. Today, advanced manufacturing systems such as flexible manufacturing systems (FMSS), computer aided design/manufacturing (CAD/CAM), and Just-in-Time (HT), offer a new dimension for the definition of process technology -- flexibility (Groover 1980, Gupta and Goyal 1989). With the combination of the concept of automation and flexible system and dedicated system, automated process technology can be categorized into two groups —- flexible automation and rigid automation (Groover 1980). Whereas the rigid automation is related to cost efficiency, the flexible automation is adopted to improve the firm's flexibility. Flexible automation like FMS and CIM has brought about significant advantages over traditional batch manufacturing by reducing the level of WIP inventory (Ranky 1983), manufacturing lead time, and improving utilization of resources (J aikumar 1986). As compared to conventional systems, the greater flexibility provided by such automation also allows it to be used for successive generations of products and gives it a longer useful life than traditional process investments (Gupta and Goyal 1989). Cox (1981) refer to an experience of the Ford Motor Company to illustrate the advantage of flexible automation over rigid automation; "to convert one plant to manufacture six-cylinder engines instead of eight-cylinder engines, it was necessary to remove and replace all the tooling in the plant It also suggests why, in an increasingly uncertain and volatile worldwide market, a more flexible manufacturing technology would offer economic advantages to mass producers." In addition, flexible technology also provides many additional benefits such as better product quality, reduced inventory and 41 floor space, lower throughput and lead times, and faster response to market shifts (Groover 1985; Hayes and Jaikumar 1988; Kaplan 1986). In their review paper for new manufacturing technology, Naik and Chakravarty (1992) summarizes the characteristics of the flexible systems with high fixed costs, low variable costs, dynamic reallocation and coordination of resources, small lot sizes, and shift from the economies of scale to the economies of scope. 2.3.1.3 Measurement for the Flexibility of New Process Technology There are not many attempts to find the differences between rigid and flexible automation. Groover (1980) separates these two automation with the fixed sequence of the operations in the rigid automation versus the programmable sequence of the operations in the flexible automation. This difference in programmability separates two automation with the number of parts or products each automation can handle economically. Groover lists the examples of two types of automation; (1) transfer lines, automatic assembly, and so forth as rigid automation and (2) numerically controlled machine tool as flexible automation. The most frequently used measure for the flexibility of the production system in past research (Burstein 1986; Groover 1980; Fine and Freund 1986) is the number of parts or products produced. It is also consistent with the definition of production flexibility by Browne (1984). In particular, Gaimon (1986) uses a flexibility index indicating a firm's combined capability for product—mix, volume, quality of output, and customer service acquired by the flexible technology. She considers the flexibility index as positively related to the acquisition rate of the flexible technology acquired and to the growth of the firm’s market demand. These examples of both automation suggested by Groover (1980) and the anecdotal evidence in the literature typify the measurement for the distinction between two automation. The variety of parts or products that each automation can handle 42 economically provides the conceptual foundation for the measurement. In addition, there is no clear border line between two automation. For example, we will call the automation that can handle two products more rigid than the automation that can handle four products. Hence, the relative distinction between two types of automation can be made according to the degree of flexibility, defined as the number of products or parts producible (see also Meredith and Suresh 1986). Given the definition of a system's flexibility as being the number of producible product (part) types the most critical issue in the measurement of a system's flexibility is whether the system's flexibility is related to the system's capacity level in units of output. It may not be appropriate to conjecture that the system's flexibility will increase as more machines are installed, which has been a widely used assumption in the analytical studies for new process technology. A good illustration is presented in the study by Nakarur (1992). According to his measurement index for the system‘s flexibility, the system's flexibility is independent on the number of machines in a system. Rather, the system's flexibility is dependent on the flexibility of each machine, such as producibility of excess capacity of other machines. Hence, the overall system's flexibility may have to be considered independently of the capacity of process technology in the development of the model in this dissertation. 2.3.2 Effects of New Process Technology Trade-offs against the costs of the capital investment and the human resources are a wide range of benefits attributed to new process technology. These benefits include lower direct manufacturing costs resulting from reductions in setup time, processing time, labor requirements, lead time, inventory, factory space, and so on (Jaikumar 1986). The effects of new process technology can be grouped into two categories - internal effects and external effects. Internal effects include mainly the effects on manufacturing lead time and the cost structure of production system. External effects include the effect on the 43 firm's sales, usually the result of the internal effect. 2.3.2.1 Effects on Manufacturing Lead Time Manufacturing lead time is the combination of processing time, setup time, and waiting time. New process technology can reduce the significant portion of time a part or product spends in the shOp (Jaikumar 1986) like throughput time, waiting time (Goldhar and Jelinek 1983; Groover 1980), and setup time (Groover 1980). For example, one major computer integrated facility at Messerschmitt-Bolkow-Blohm in Augsburg, West Germany, reduced production lead time for the Tornado fighter plane from 30 months to 18 months(Jelinek and Goldhar 1983). A survey result by J aikumar (1986) also shows the drastic cut in processing time to one third on average after the installation of FMSs. One significant advantage of the lead time reduction is the reduction of WIP inventory. Since the setup time becomes negligible in new process technology, the EOQ can approach one as in JIT system. More importantly, the range of EOQ the system can handle at reasonable costs may be much wider than the traditional system can handle. Another important advantage is faster market response. The results of faster market response include less safety stock, fast response to inaccurate forecast, and widened product line (Meredith 1987). 2.3.2.2 Effects on Cost Structure Total manufacturing costs include fixed cost and variable costs. Fixed cost is related to capital investment in machinery and equipment in the factory. A survey result shows initial investment cost in FMS ranges between $10 and $25 millions (Smith et al. 1986; see also J elinek and Goldhar 1984). The investment costs are higher from traditional job-shop, automated transfer line, to flexible manufacturing system, like FMS and CIM. Hence, each additional automated capacity increases the fixed cost of capital investment. The flexibility embedded in the acquired new process technology may 44 increase fithher the fixed cost of capital investment. This proposition might be inferred from the fact that the investment cost for the flexible production systems like FMS or CIM is enormous compared to the cost for the stand alone NC machines. Variable costs can be measured in terms of the order quantity and the variety of products or parts processed. As discussed above, there are two types of variable costs -- the scale-related costs and the variety-related costs. Stalk (1988) suggests that, in a flexible manufacturing system, the variety-related costs start lower and increase more slowly as the product or part variety grows than in a conventional manufacturing system. Hence, a flexible system enjoys more variety with lower total costs than do traditional factories, which are still forced to make the trade-off between scale and variety. Hough (1989) emphasizes the reduction in direct labor (see also Gaimon 1985; and J aikumar 1986) and a probable flattening of the learning curve as the effects of automation. He also points out the reduction in indirect labor and other elements of overhead. Depending upon the degree of automation, he asserts, there should be a diminished opportunity for workers to learn better ways of accomplishing a task, primarily due to the reduction of direct labor. In addition, elements of learning due to improvements in tool coordination, shop organization, and inventory systems would be eliminated by the ability to simulate production processes and layouts prior to setting up a plant (see also Levy 1960 and Wild and Port 1987). 2.3.2.3 Effects on Sales Growth The effect on sales growth of new process technology is an indirect result of faster market response and/or widened product line. As discussed above, new process technology allows the firm to update products continuously and thus tends to increase the range of product line, its complexity, and its rate of change. It also creates an entry barrier against competition due to increased technological content of products, closer market links, and improved responsiveness (Goldhar and Jelinek 1894). These overall 45 effects may increase a firm's market share. Roth et al. (1991) suggest that new process technology improves the firms' market share through the anticipated improvements in outputs and allows the firms to pursue a broader marketing strategy, thereby enabling them to capture a portion of their competitors' demand. The result of an empirical study by Thietart and Vivas (1984) shows that market share increases as investment and assets increase for all kinds of industry involved in their study. This result implies that the possible investment in new process technology has a positive relationship with the sales growth of the firms. In summary, the acquisition of automation of new process technology seems to be viewed usually as the substitution of the new automated capacity for the manual output capacity to improve manufacturing performance. Furthermore, the flexibility of new process technology is found to provide the manufacturing firms with additional cost and benefits. Hence, there should be clear distinction between the effects of the acquisition of mere "automation" of new process technology and the effects of the acquisition of "flexibility" of new process technology. This distinction should be considered in the process of the model development in this study. 2.3.3 New Process Technology and Learning Effect Learning effect is an widely discussed in many studies of developing or applying the production fimction. Productivity improvement due to the learning effect was first identified by Wright (193 6). Yelle (1979) found this effect existing in a number of industries from early in this century. Levy (1965) develops a learning function based on actual examples in which the rate of learning is related to the experience of workers which are exogenous to the cumulative production volume. Levy claims that the productivity improvement is realized by two factors, "the initial efficiency of the process" and "the autonomous learning effect.” He assumes that "the initial efficiency of the process" depends on the firm's preplanning efforts including testing prototype process, training or 46 education of workers, and so on. On the other hand, "the autonomous learning effect" is assumed to depend on the cumulative experience of workers as the production system runs. Hence, his learning fiinction implicitly assumes that the automated production system with a low level of labor force may show a slow learning effect with high initial efficiency compared to the conventional production system with a high level of labor force. Hence, his learning fimction assumes "the flattening learning curve" as discussed in section 2.3.2.2 (see also Hough 1989; Wild and Port 1987). Recognizing the shortfall of conventional studies of the learning effect that dealt only producing homogeneous products, Meredith and Cam (1989) introduce a new term -- synergy -- to include the learning effect in a situation of adapting several new advanced technologies to produce a variety of products. The authors present a unit production cost fiinction incorporating this synergy effect. The fiinction assumes that the high synergy effects among new advanced technologies adOpted by the manufacturing firm result in low unit production cost. They contend this function can be used in determining the order of acquisition of several new advanced technologies. The significance of their study is that they provide an approach to finding the learning effect due to the synergy of new advanced technologies. This is a significant development in view of the fact that the conventional studies of the learning effect have focused only on the improvement in labor productivity or labor cost. However, a major drawback of this study is that the synergy effect can only be estimated after physical running of new advanced technologies which would be almost prohibitive due to the huge amount of costs and time involved. Roth et al. (1991) applied this synergy concept in developing their analytic model to study the optimal acquisition of new flexible technology. They views the synergy effect as causing a structural shift in the level of "technological progress" of the production System, which they use as a new term for the learning effect reflecting gains in system utilization and productivity. This learning effect is assumed to result from improvements in layouts, machine loadings, machine speeds, yields, use and integration of system 47 components, and management methods. Like Levy's learning function, their model implicitly assumes high system synergy afforded by new flexible technology results in a flattening learning curve. This is the same as the view of others like Levy (1965), Hugh (1989), and Wild and Port (1987) who argue that more automation leads to less room for the learning effect. 2.3.4 Analytic Studies for New Technology Acquisition Analytic studies in new technology acquisition have many different approaches in terms of methodology. Most studies have used either mathematical programming (Burstein 1986; Fine and Freund 1986; Li and Tirupati 1992) or optimal control theory (Amit and Ilan 1990; Gaimon 1985, 1986, 1989, 1992; Roth et al. 1991 . Addressed in these studies are the issues such as trade-offs between scale and scope, the timing of acquisition, the size of acquisition, and the effect of learning on optimal acquisition. In this section, several key models are presented and discussed with issues related to this dissertation. 2.3.4.1 The Model of Fine and Freund Fine and Freund (1990) presents a two-stage stochastic model for optimally choosing a portfolio of flexible and nonflexible manufacturing capacity. Their model is very similar to Burstein's (1986) except their model includes stochastic demand and multiple products. A similar study by Li and Tirupati (1992) presents a nonlinear model for the technology choice with concave cost functions. The focus of Li and Tirupati's study is the development of an efficient heuristic algorithm. Fine and F reund's model focuses on the economic tradeoffs between the acquisition costs of flexible capacity and a firm's ability to respond flexibly to future uncertain demand. They claim that the model captures a technology's flexibility characteristics with greater revenue resulting from broader product mix capabilities and lower operating costs due to economies of scope. 48 Fine an Freund formulate a single-period, multi—product, dual-technology problem to maximize the total profit by netting the technology acquisition costs and the manufacturing costs out of the total revenue. Below is the formulation of their model. 1 max —rFKF —§jer-KJ- +Ziipi§{Rij-(Yrj +Zij)—Cj '(Yij +25», {KF,KJ-;j=l ..... n} {Yij-,Zij;i=l,...,k,j=l ..... n} subjectto Yij—K,- :0, i=1 ..... k j=l ..... n, Zle—KFSO’ 1:1 ..... k, J Yij-20, i=1 ..... k, “l ..... n Z620, 1:1 ..... k J—1,... n K120, j—l ..... n KF20 The first two terms in the objective function are the technology acquisition costs for both flexible and dedicated technology. The term, pi, represents the probability associated with the state i of the world for market demand. The term, Rij(Yij+Zij)» is the revenue function of product family j in state i. In their sensitivity analysis, the authors assume the revenue functions to be quadratic functions (i.e., linear demand functions). The term, CJ-(Yij+Zij), represents the manufacturing variable cost. The terms, Yij and Zij, represent the amounts of product family j in state i produced by dedicated capacity j and flexible capacity, respectively. In numerical examples, due to the enormous number of possible states of the world and the condition for uniqueness of the optimal solution, Fine and Freund carry out sensitivity analysis in a two-product case. Moreover, by assuming that the variable costs for either technology are very low, they exclude the variable costs from the objective fiinction. In the first sensitivity analysis, they explore analytically the sensitivity of the optimal capacity levels to the acquisition costs. In the second sensitivity analysis, they explore the sensitivity of the optimal capacity levels to the changes in the distribution of demand. A major drawback of their model is that it separates two technologies only by their acquisition costs, and thus it misses many potential benefits of flexible technology in 49 the formulation. In fact, the elimination of the economies of scope from the model, i.e., the elimination of variable cost, may diminish the significance of their study. 2.3.4.2 The Model of Gaimon Gaimon conducted many studies related to automation (June 1985, September 1985, 1986, 1989, 1992). Her studies have addressed many issues such as diminishing returns (1985a), productivity enhancement of labor (1985b), application of game theory (1989), and lagged learning effect (1992). In most of her work, she views the acquisition of automation as the substitution of automated capacity for manual output capacity. Below is one of her models that is closely related to this dissertation. j: {[p(t) - BU) — b(1)]d(t)— c(t)a(t) + s(t)r(t) — m(t)q(t)}c'“d[ +gfme‘” + mime-rT subject to d(t) = d1(t) - d2(t) + d3f(t), q'(t) a(t) - r(t), (1(0) =qo, b'(t) = -0t(t)a(t)b(t), b(0) =bo, f(0 = 13(080), f(0)=fo, d(t) 2 0, t e (0,T), d(t) SQG), t6 (0,T), c1(0 - a(t) 2 r(t), a(t) e (0, A(t)), r(t) e (0,R(t)), p(t) 2 0. In this model, the term, f(t), represents the flexibility index summarizing a firm's capabilities with respect to product-mix, volume, quality of output, and customer service. Therefore, this model assumes, as described in the demand function, that a firm with a high index level of flexibility has the potential to increase its demand by manufacturing products whereas a firm with less flexible productive capacity may be unable to produce. One important aspect of this model is related to the state equation for the flexibility index. 50 The term, B(t), is defined as a measure of the effectiveness that acquiring technology has on increasing the level of flexibility index at time t. Gaimon defined this term as a function of time to capture the possibility of anticipated technological advancement. According to the constraint for f(t), the more flexible the system, the higher the value of B(t), and Vice versa. In addition, according to the third and fourth constraints, both the unit operating cost and the system's flexibility depend on the acquisition rate of automation. As discussed in Section 2.3.2.3, flexibility is the capability of the system which may be independent of the level of automation capacity. Therefore, it might be appropriate to have a control variable representing the rate of acquired flexibility independent from both the rate and the accumulated level of acquired automation capacity. 2.3.4.3 The Model ofRoth et al. The model of Roth et al. (1991) is an extension of Gaimon's model (1986). Roth et al. consider in their model the lagged learning effect, called as "technological progress," the market share increases due to the acquisition of new process technology, and the reduction of unit operating costs. In this model, the flexibility of the process system is proportionate to the rate of acquired flexible automation capacity. Thus the impact on market share and unit operating costs is proportionate to the rate of acquired flexible automation capacity. Below is the formulation of their problem. Max [o,s(r) +G,k(T) +G3a(T)]E(T) —joT{n[s-s12 +c1a2+c3r2 +[B+c3]s+c,[dk -s]2 -c,[dk -s]}Edt subject to s' = yl(a+ak)(N-s)+yzs. x' = a. k' = a + onk - 1'. OL' = -\llOL(1-¢8/X). 51 d3 = -BaC3. a e (0, A), r e (0, R). They assume that all demand is met through the available operating capacity or through the use of short-term measures that increase capacity. They also assume incremental timing strategy to upgrade a firm's capacity by asserting that radical diffusions of new technologies may cause disruptions and be more costly than are incremental diffusions. Hence, the cost functions for acquiring flexible automation and retiring conventional capacity are defined as quadratic. The presence of the capacity variable, k, makes the model capture the impact of technological progress due to the organizational learning in the total capacity and the market share. Hence, the change in total capacity and the market share over time is the function of both the change in flexible automation capacity and the change in the total capacity due to technological progress. Interestingly, this technological progress factor, or, is considered as a state variable that is decreasing over time. Roth et al. also assume the rate of decrease in technological progress over time is decelerated by the system's synergy effect (Meredith and Cam 1989). This concept is the same as the flattening of the learning curve due to the acquisition of automation as discussed before. However, empirical evidence suggests that the learning effect is a fiinction of cumulative experience, which is usually represented as a cumulative production level (see Yelle 1979). Hence, it may be appropriate to relate the value of reduction factor, w, to the cumulative demand level rather than to choose an arbitrary value for the factor. Overall, the contribution of this model is the extension of previous studies in this area including the detailed modeling aspects with the consideration of the interaction effects of both the organizational learning and the acquisition of flexible automation on both market share and unit operating costs. 52 2.3.5 Summary and Implications In the review of the past analytic studies, there are three findings: (1) Lack of consideration for the flexibility aspect of new process technology; (2) Failure to link technology acquisition decisions with the corporate strategy; (3) Single dimensional definition of process technology as production capacity in units of output. In fact, these drawbacks are found mainly due to the single dimensional definition of new process technology as production capacity in units of volume. It is expected to enhance the richness of analytic models in the future studies with continuous efforts in developing conceptual works to distinguish two aspects of new process technology, automation capacity and flexibility. Hence, this dissertation attempts to propose an analytic model overcoming the drawbacks of the past studies by considering the two-dimensional aspect of new process technology. 2.4 RESEARCH STATEMENT This dissertation addresses the linkage of the strategic orientation of SBUs regarding cost efficiency and flexibility with the acquisition decisions for new process technologies. Specifically, this study attempts to provide managers with a strategic framework for the acquisition decision process of new advanced technology as a practical tool for strategic decision making in technology acquisition. The focus of this study is on the development of an optimal control model for optimal decisions regarding the acquisition of new process technology to enhance the production volume and variety capability of the system and, using the scenario analysis, on the investigation of relationships between process technology choice and competitve priorities (cost efficiency vs. flexibility) of a firm's SBU. The contribution of this study is also due to the fact that the acquisition of new process technology is considered to affect the production system along two important 53 dimensions, its production capacity in units of output and its ability to produce a variety of products or part-types. The choice of the mix of these two dimensions of process technology is also considered to affect its market potential regarding volume and variety in addition to unit production costs. The objective function considers the trade-off between the market potential and SBU goals regarding volume and variety. Through systematic scenario analysis, the evolution of process technology choice in relation to the strategic competitive priorities of the SBU along the stages of the product life cycle is thoroughly investigated. CHAPTER HI STRUCTURAL FRAMEWORK FOR THE STRATEGIC ACQUISITION OF NEW PROCESS TECHNOLOGY: A CONCEPTUAL FRAMEWORK As discussed in the previous chapter, the acquisition of new technology has significant strategic implications for a firm's long term market competitiveness. The understanding of relationships among various contingent variables surrounding a firm's internal and external operation environment is a prerequisite for the analysis of the strategic implications of technology acquisition (Madu and Georgantzas 1991). The structural framework proposed in the first section explains how technology decisions are related to the corporate strategy through the business strategy. This relationship will be a conceptual basis for developing the objective function of the model presented in Chapter 4. The following two sections explain how changes in a firm's market demand and manufacturing cost structure are affected by other variables. Hence, these two sections provide conceptual bases for the development of constraints of the model in Chapter 4. The last section introduces a new concept called the "Product-Technology Box" by incorporating two dimensions of technology -'- automation and flexibility -- into the normative concept of product-process matrix. This new concept has been supported by many researchers (Boothroyd 1982; Hayes and Wheelwright 1984; Adler 1988). With the results from the scenario analyses at the end, this research attempts to explore how the best fit between technology choice and market requirements within the new "Product- Technology Matrix" evolves due to the optimal acquisition of new process technology along the stages of a product life cycle. 54 55 3.1 DECISION PROCESS FOR NEW TECHNOLOGY ACQUISITION The decision process for the acquisition of technology is comprised of three levels of decision as depicted in Figure 3.1. In this structural diagram decisions for the acquisition of new technology made at the functional level are linked to corporate strategy and business strategy. Decisions related to the corporate strategy are the relative importance weight given to a particular strategic priority, either flexibility focus or cost efliciency focus, depending upon whether the firm pursues a differentiation strategy or a cost leadership strategy (Porter 1980). Decisions related to business strategy include each SB U 's goals for the growth of market share in volume and the growth of market share in product @art) variety. The growth of market share in volume is related to the corporate strategy of pursuing cost efficiency whereas the growth of market share in product (part) variety is related to the corporate strategy of pursuing flexibility. Firms with the strategic priority of cost efficiency will try to achieve economies of scale by increasing sales in volume. On the contrary, firms with the strategic priority of flexibility will try to achieve economies of scope by increasing sales in variety. Hence, the goals of a firm's SBU are closely related to the strategic priority stipulated by the corporate strategy. In addition, the real values of SBU's goals are to be set according to the firm's projection for future market conditions along the market product life cycle. When the SBU's goals are set at the business level, decisions regarding the acquisition of new technology will be made at the functional level. Decisions for the acquisition of technology include the timing and size of investment in automation capacity of new process technology, flexibility of new process technology, and the reduction of conventional capacity. Since new process technology is defined in two dimensions - automation capacity and flexibility - in this dissertation, the acquisition of new process technology entails two separate decisions, one for the acquisition of 56 automation capacity and another for the acquisition of flexibility. Corporate Strategy Strategic Priority Business Projection of Strategy Future Market Condition SBU's Goal Funcflonal Actual Strategy Market Position Technology Decision Internal ______> Production Environment Performance - Flexibility - Cost Efficiency - Sales Growth in Volume - Sales Growth in Product (Part) Variety - Purchase of Automation Capacity - Purchase of Flexibility - Reduction of Conventional Capachy - Product Volume - Product Variety - Production Cost - Flexibility - Total Manufacturing Cost - Strategic “Fit” Figure 3.1 Structural Diagram of the Decision Process for the Acquisition of New Advanced Technology As discussed in Chapter H, many have asserted that optimal decisions for the acquisition of technology should be contingent upon the strategic priority decided at the corporate level and the following SBU's goals decided at the business level (Porter 1980; 57 Skinner 1967; Simmonds 1981; see also Swamidass and Newell 19878). Hence, the model in this dissertation includes variables related to the SBU's goals and the coefficients related to the strategic priority. In doing so, this dissertation explores how decisions for the acquisition of new technology are related to the desired business goals as well as the firm's strategic priority. According to Figure 3.1, decisions for the acquisition of new technology affect internal production environment, which consists of production volume, production variety, and manufacturing cost structure. Technology decisions also affect the firm's actual market position in terms of both sales volume and product variety (refer to section 2.3.2.3). Hence, the model in this dissertation will measure both product volume flexibility and product mix flexibility of the firm by comparing the internal production environment with the actual market position. Detailed relationships among variables related to technology decisions will be discussed in the next section. The overall performance to be measured in the model includes not only the total costs related to both technology investment and production as well as flexibility in terms of production volume and product variety, but also the costs related to the strategic "fit," which represents the consistency of decisions made at various levels in the hierarchy of strategy as emphasized explicitly or implicitly by many researchers (F rohman 1982, 1985; Garrett 1986; Hayes and Wheelwright 1984; Schroeder 1990; Skinner 1974; Wheelwright 1984). The relationship between conceptual linkage from corporate strategy to production environment and actual scheme for measuring the overall performance in the model is depicted in Figure 3.2, in which the corporate strategy dictates SBU's goal, the technology decisions affects SBU's actual demand, and the SBU's production environment dictates its production decisions. The model can measure the strategic "fit" by comparing 8 This empirical study reveals the positive relationship between the perceived environmental uncertainty and the manufacturing flexibility of a firm. In this study. the manufacturing flexibility is measured with respect to the product mix flexibility. 58 SBU's goals with its actual demand while it can measure flexibility by comparing actual demand with production decisions. Business Goal Actual Demand Production A _ _ Decisions Volume ‘ Volume Volume Variety Variety Variety 11 ii 1 Corporate Strategy Technology Acquisition Production A __ A Environemnt Cost Efficiency ‘ Automation Capacity T Volume Flexibility Flexibility Variety Figure 3.2 Linkage Among Various Decisions Table 3.1 Objectives of The Acquisition of New Technology Tacficaland Strategic Objective , , _ Operational Objectives Consistency with the Flexibility maximization firm’s competitive - volume strategy - variety Min [ Desired Business Total costs minimization Goals -Actual Market - investment cost Position ] - production cost As described in Table 3.1, the strategic objective for technology decision making, called strategic fit, is the minimization of the deviation between a firm's actual market position through the acquisition of new technology and the desired business goals to align the acquisition of new technology with the finn's overall competitive strategy. The tactical objective in this model is the minimization of costs related to the product volume flexibility 59 and product mix flexibility. Operationally, the acquisition of new technology should be decided to minimize the total costs combining fixed investment costs and the production variable costs through trade-offs with flexibility (Simmonds 1981). Particularly, by incorporating the strategic objective, the model can measure linkage between the strategic priority of corporate strategy and the acquisition decision for new advanced technology of functional strategy. Hence, the optimal control model presented in Chapter IV attempts to capture the linkage among strategies in the hierarchy of strategy and measure this linkage as the overall performance. 3.2 TECHNOLOGY AND EXTERNAL EN VIRONIVIENT There are two possible causes for changes in a firm's aggregate demand. One of these causes is the change in internal competitive force, and the other is the change in external competitive force. This type of view can be found in the well-known Vidale- Wolfe sales response model (1957). Internal change may include change in delivery performance, service quality, product quality, or information systems for data communication. External change may include change in the position of the product in its own product life cycle, elasticity of demand due to competitive forces, general economic conditions, or other environmental forces (Roth et al. 1991). As depicted in Figure 3.1, the actual market position of a manufacturing firm is assumed to be postulated in terms of both volume and variety. Anecdotal and empirical evidence provided in the literature suggest that the acquisition of new process technology improves a firm's market position (Goldhar and Jelinek 1984; Thietart and Vivas 1984). Unfortunately, these studies fail to address the distinction between improvement in market position in terms of volume and in terms of variety. Despite the lack of support from the literature, this study proposes that the improvement of market position in terms of volume is more closely related to the level of investment in the automation capacity of new 60 technology, whereas improvement of the market position in terms of variety is more closely related to the level of investment in the flexibility of new technology. The reasoning of this proposition is that volume change in market demand can be better met as more investment in new technology capacity improves the capacity utilization of a production system to meet the variation of sales in volume (Roth et al. 1991), whereas variety change in market demand can be better met as the acquisition of flexibility firrnishes a production system with more flexibility since the customer's brand loyalty may be enhanced due to the benefits of better product quality, reduced lead time, and fast response to market shifts (Groover 1985; Hayes and Jaikumar 1988; Kaplan 1986). As discussed in the previous section, minimization of the discrepancy between the SBU's goals and its actual market position plays a pivotal role in linking the strategic Priority of the corporate strategy and decisions for the acquisition of new technology of the functional strategy. Since the technology decision affects a firm's actual market Position, and the business goal is a replica of the corporate strategy with respect to sales gr OW‘th and introduction of new products, by matching these two in close proximity, managers can make the technology decision that is sound and consistent with the Objectives stipulated by corporate strategy. 3.3 TECHNOLOGY AND MANUFACTURING COST STRUCTURE In this study it is assumed that a firm pays a fixed investment cost for new techlIOlogy capacity and its flexibility separately. In fact, a stand-alone NC machine usually costs around $300,000 to $400,00 whereas an FMS, which combines these NC macllines with other automated systems to increase the flexibility of a production system, can CO St sometimes tens of millions of dollars (Bobrowski and Mabert 1988). Hence, the fleXibility embedded in new process technology will require more investment in addition to the . . . . . b21810 cost for mere automation capacrty as shown in Figure 3.3. ‘ 61 Fixed investment (1 Cost / Flexible Automation $100,000 / / Rigid / Automation $0 0 . 100,000 uints Capacrty Figure 3.3 Technology and Investment Cost As Stalk (1988) suggests, this study assumes there are two types of production variable costs — volume—related variable costs and variety-related variable costs. The average volume-related cost decreases as the degree of automation increases (see Roth et al. 1 99 1). This type of cost decline results from the elimination of manual activities in Production capacity due to the acquisition of new process technology. It is important to note that as the cumulative automation capacity increases, there will be less opportunity to gain the learning effect due to the low level of labor—intense activities and the high level of integration among machinery and tools (Wild and Port 1987). Hence, the rate of decline Of VOIUme-related variable costs will decrease as the degree of automation increases (see Figul‘e 3.4). In the model the initial value of volume-related variable costs of a production Syst em will be determined by its initial degree of automation. In this dissertation, the VOlurTle-related learning effect is not considered. The average variety-related costs reflect the costs of complexity in manufacturing, including setup, material handling, inventory, and many of the overhead costs attributed to the degree of product or part variety of a factory (Stalk 1988). The average variety- related costs increase as the production system processes more product (part) types. Si . . . . . . . . . flee the system's flexrbrlrty 1n thrs dissertatron rs defined as the number of unique product ‘ 62 (part) types to be produced, the average variety-related costs increase as the system's flexibility increases as depicted in Figure 3.5. It is also assumed that the rate of increase in average variety-related costs becomes smaller as the cumulative system's flexibility increases. Volume- Related Variable Costs $0 Low High Degree of Automation Figure 3.4. Automation and Average Volume-Related Variable Costs Variety— Related Variable Costs $0 Low High System's Flexibility Figure 3.5. Flexibility and Average Variety-Related Variable Costs 3.2.1 PRODUCT-TECHNOLOGY BOX In this dissertation new process technology is defined in two dimensions -- ¥ 63 automation capacity and flexibility. Definitions for these two dimensions were already discussed in Chapter 11 (see section 2.3.1). Automation capacity of new technology is defined as production capacity substituted for the conventional production capacity. Flexibility is defined as the capability of the production system in terms of the number of product or part types the system can produce. Hence, any process technology can be defined by the combination of the degree of automation and the degree of flexibility. This definition of process technology is quite different from the definition of process technology found in the product-process matrix prOposed by Hayes and Wheelwright ( 1 97 9a). In the product-process matrix the process technology is defined by the combination of a fixed degree of automation and a fixed degree of flexibility, such as high automation/low flexibility for transfer line and low automation/high flexibility for job shop. Unfortunately, as Hayes and Wheelwight (1979b) pointed out, their product-process matrix seems to be inappropriate as new process technology with a high degree of both automation and flexibility becomes available. New process technology allows firms to PTOduce a wider range of product variety and volume at the same or lower cost than does conventional technology (high-automation/low-flexibility or low-automation/high- flexibility) defined along the vertical axis of the matrix. Hence, new process technology might allow firms to pursue both cost efficiency and flexibility as the same time (Porter 1980) . Therefore, to be a viable paradigm of manufacturing strategy the product-process matrix may need to be revised to fill the gap between the existing normative concept behind the product-process matrix theory and the current trend of rapid diffusion of new proceSS technology. 64 Flexibility High Low High Automation Figure 3.6 Technology Matrix To fill such a gap, a new term, called a "Product-Technology Box," is proposed to integrate the PLC and a firm's technology choice as depicted in Figure 3.6. Unlike the product—process matrix suggested by Hayes and Wheelwright (1984), a new matrix, called a "Technology Matrix," combining automation in the horizontal axis and flexibility in the vertical axis is first devised to represent technology choices (see Figure 3.6). Technology choices on the diagonal in the new technology matrix have the same definitions as in the pr Oduct-process matrix. The upper left corner (Job Shop), labeled as point A in Figure 3.6, represents a technology choice with high flexibility and low automation. The lower right corner (Transfer line), labeled as point E in Figure 3.6 , represents a technology ChOiCe with low flexibility and high automation. The major difference of this new matrix is that any technology choice can be defined with the combination of the level of flexibility given by a point on the vertical axis and the level of automation given by a point on the horizOntal axis. Any off-diagonal points, such as B and D, have different definitions than in the Conventional product-process matrix. In the conventional product-process matrix, the firm at point B has the same process technology as the firm at point A. In this new matrix the firth at point B (Firm B) has higher level of automation than the firm at point A (Firm A) and the higher level of flexibility than the firm at point C (Firm C). The firm at point D ‘ 65 (Firm D) has a lower level of flexibility than Firm A and a lower level of automation than Firm C. Hence, it is certain that Firm B (Firm D) needs more (less) fixed capital investment for its process technology than does Firm A or Firm C. According to the normative concept of the product—process matrix, Firm A is suitable for the product structure of low volume and high variety whereas Firm C is suitable for the product structure of medium volume and medium variety. In this new matrix, Firm B has not only a high degree of flexibility, as does Firm A, to operate efficiently for up to high levels of product variety but also the medium degree of automation, as does Firm C, to operate economically for up to medium size of product volume. Therefore, Firm B might be able to achieve the economies of scale of Firm C as well as the econorrries of scope of Firm A, simultaneously. On the contrary, Firm D has a technology choice with the same level of flexibility as Firm C and the same level of automation as Firm A. According to the normative concept, Firm D would be suitable for the pro duct structure of low volume and medium variety. Unfortunately, this kind of produ ct structure cannot be found along the stages in the normal product life cycle. consequently, the technology choice of Firm D is not considered in the conventional Product-process matrix. However, since this choice has a lower level of fixed capital investrnent than others such as A, B, and C, and a firm's technology choice changes ConStantly over time due to the acquisition of new technology, this choice as initial process tecl'11‘1010gy might have an advantage over the other positions in the dynamic environment. Therefore, this study does not exclude this position in the scenario analysis. To this new technology matrix, a third dimension is attached to represent the Stages of product life cycle. This third dimension is the same as the horizontal axis of the Conventional product-process matrix. The product structures of the introductory stage through the declining stage of product life cycle, from low volume/high variety to high volume/low variety, are represented along this additional dimension. Hence, the "Product- e(“'hl'lology Box" combines all three drmensrons in one diagram as depicted in Figure 3.7. ‘ 66 A point in the box has a corresponding product structure of a stage of the PLC and a corresponding technology choice. For example, at point X in Figure 3.7 Firm F is operating with a technology choice of a high degree of both automation and flexibility and facing with a product structure of the mature stage of the PLC. -<—i"'r—“rn">rmr-n AUTOMATION Figure 3.7 Product—Technology Box One important feature ofthis box is that it is possible to identify the range of pr Oduct structures suitable for any technology choice defined by the new technology matrix) which would be difficult with a two—dimensional matrix such as the product- p r 0(3683 matrix. For example, Firm B might be suitable for the product structures ranging from the introductory stage to the growth stage of the product life cycle. A firm positiOned at the upper right corner ofthe technology matrix may be suitable for the Product structures ranging from the introductory stage to the mature stage ofthe product life c=ycre. With the above new concept of Product-Technology Box, this dissertation a . . ttempts to explore analytically how a firm's technology chorce evolves on the new ‘ 67 technology matrix according to the transition of PLC and the firm's choice of competitive priority for each stage of PLC. In particular, this study tries to identify how the evolution of technology choice on the technology matrix differs according to a firm's competitive priority -- either flexibility or cost minimization. This study also tries to explore how the technology choice evolves in two different market environments of the normal PLC and the reversal of PLC. Significantly, the analysis of the interaction effect between the types of PLC and the competitive priority will provide managers with an insightful guide for how to make optimal acquisition decisions for new process technology consistent with the competitive strategy under different projections for future market behavior. CHAPTER IV THE MODEL In this chapter, a dynamic optimal control model for the acquisition of new technology is developed according to the conceptual frameworks proposed in Chapter III. This chapter is separated into three parts. In the first part, the general description of the model is provided along with the assumptions and the variable definitions. The second part presents the objective function of the model. The state constraints and the control variable boundaries are depicted in the third part. The model for this study follows the standard approaches taken by many analytic studies in the literature on process technology. The notation for time, (t), is suppressed in the objective function and the subsequent constraints unless it is required. 4.1 INTRODUCTION This section starts with a brief description of the proposed model. Then the assumptions for the model are presented, followed by the description of the notation used in the model. 4°1‘ 1 General Description of The Model As mentioned in chapter 11, few past analytic studies regarding process technolog rWestrnent have consrderedflexzbzlity as one of the objectives in their models. Past a . . . . . . nalYtlc studies related to process technology generally vrew the acqursrtron of technology a . . . . . . S the capacrty expansron, frequently combined With the learning effect. The major 68' ‘ 69 drawback of their models is that they consider the cost efficiency as the only objective. The model presented here includes not only cost efficiency but also costs related to flexibility as the performance measures in the objective fiinction. Particularly, by incorporating implicitly relative importance weights placed on two strategic priorities, cost e ficiency and flexibility, according to the corporate strategy, the model attempts to embed the linkage between the corporate strategy and the functional strategy. Hence, it allows decision makers to render the optimal policy not only for its best positioning of technology choice in the product-technology matrix but also for the technology investment strategically viable according to the relative emphasis given to each of two competitive priorities -- cost efficiency and flexibility. The model in this dissertation considers both product volume flexibility and product mix flexibility. These two types of flexibility are included explicitly in the Objective function as the difference between actual production in volume and market demand in volume and the difference between actual production in variety and market demand in variety. Hence, any deviations in terms of volume and variety incur costs deteriorating the overall performance of a firm's SBU. According to the perspective of this study, the types of flexibility considered in the model should adequately portray the strategic implications of new technology acquisition conditioned by the external strategic variables, actual market demand in volume and variety. In this regard, the above two types of flexibility are considered as the most suitable ones in reflecting strategic implications with respect to the positioning of technology choice in the product— technology matrix proposed in Chapter III (see Figure 3.6)- 4'1 '2 Assumptions 1‘ The firm is manufacturing products of a single type. 2‘ The market behaves in perfect competition. Hence, there is no dominant competitor in the market. 70 3, There is no change in the technology choices of the competitors in the market. 4, The prices are fixed over the planning horizon. 5, New advanced technologies are readily available and there is no barrier for individual firms to access new advanced technologies. 6. Changes in the aggregate market demand follow a smooth transition as in the PLC curve. 7, There is no volume related learning effect in the average volume-related costs with new process technology. 8. The average variety-related costs are decided by the flexibility of the system, not by actual product (part) variety produced by the system. 8. There is no change in managerial control policy, other than specified by the variables in this model, over the planning horizon. 9. The timing policy for the process technology acquisition is smooth, evolutionary changeover from old to new technology within an existing plant. 4.1.3 Variables In this section, endogenous variables and exogenous variables are grouped separately. State and control variables are defined as endogenous variables. Cost CoetTicients and other parameters are defined as exogenous variables. 4.1.3. 1 Endogenous Variables X“) = accumulated level of automation capacity of new technology as production capacity in units of output at time t, x(0) = x0, (state variable), k“) = total production capacity in units of output at time t, k(0) = k0, (state variable), SQ) = overall system flexibility (manufacturing flexibility) defined as the number of products (parts) the production capacity handles at time t, 5(0) = so, (state variable), V“) = aggregate market demand of the firm in units of output at time t, v(0)=v0, (state y ._ 71 variable), m(t) = market demand of the firm in number of products (or parts) types at time t, m(0) = m0, (state variable), b1(t) = average volume-related variable cost at time t, b1(0) = blo, (state variable), b2(t) = average variety-related variable cost at time t, b2(0) = b20, (state variable), a(t) = acquisition rate of automation capacity of new technology in units of output at unit time t, a(t)e(0,A(t)), (control variable), f(t) = acquisition rate of flexibility of new technology measured in number of product (part) types newly acquired new technology can handle at unit time t, f(t)e (0,F(t)), (control variable), r(t) = rate of scrapping/reducing the level of conventional (existing) capacity in units of output at unit time t, r(t)e(0,R(t)), (control variable), p(t) = rate of actual production in units of output at unit time t, p(t)e(0,P(t)), (control variable), q(t) = rate of actual production in number of product (part) types at unit time t, q(t)e (0,Q(t)), O < Q(t) S k(t)/\i/, (control variable). 4.1.3.2 Exogenous Variables Yv(t) = total market size exogenously determined in terms of aggregate volume at time t, Ym(t) = total market size exogenously determined in terms of product (part) variety at time t, V(t) = business unit (SBU) goal level of aggregate demand in units of output at time t M(t) = business unit (SBU) goal level of demand in number of product (part) types at time t, c1 = acquisition cost per squared unit of automation, a(t), 02 = acquisition cost per squared unit of flexibility, f(t), c3 = scrapping/reducing cost per squared unit of conventional (existing) capacity, r(t), c4 = cost per unit squared deviation between the actual production in volume and the 0 UI e1 e2 7:1 7‘2 (hr (1)2 Br B 91 92 72 market demand in volume, = cost per unit squared deviation between the actual number of products (parts) types produced by the production capacity and the number of products (parts) types demanded by the market, = cost per unit squared deviation between actual aggregate market demand in volume and SBU's goal in volume, = cost per unit squared deviation between actual number of products (parts) types demanded by the market and SBU's goal in number of products (parts) types, = market response factor per unit of acquired new technology capacity for the change of SBU's aggregate demand in volume, 0 < n1 < 1/A(t), = exogenously predetermined coefficient representing natural growth/decay of the product life cycle in volume, -1 < n2 < DV, where DV is a predetermined upper bound (saturation level) on the growth factor in terms of volume, = market response factor per unit of acquired flexibility for the change of SBU's demand in variety, 0 < (1)1 < 1/F(t), = exogenously predetermined coefficient representing natural growth/decay of the product life cycle in variety, -1 < (1)2 < Dm, where Dm is a predetermined upper bound (saturation level) on the growth factor in terms of product (part) variety, = effectiveness factor of acquired automation capacity on reducing average volume-related cost, 0 < or < 1, = the lower bound of b1, the volume-related cost, = effectiveness factor of acquired flexibility on reducing average variety-related cost, 0 < [32 <1, = effectiveness (synergy) factor of acquired flexibility of new technology on improving overall system flexibility due to the system synergy, 0< p1 < 1, = effectiveness factor of reducing system's overall flexibility due to the attrition of conventional (existing) capacity, 0< p1 < 1, = a large number in the constraint for the relationship between a and f, = a large number in the constraint for the relationship between p and q, 73 w = a coefficient representing the proportionate loss of volume capacity due to the production of one product (part) type, r = a coefficient representing the proportionate loss of system's flexibility due to the scrapping/reducing of conventional (old) capacity. G1 = value per unit market demand of the firm in volume at terminal time, T, G2 = value per unit market demand of the firm in variety at terminal time, T, G3 = value per unit capacity at terminal time, T, G4 = value per unit of overall system flexibility at terminal time, T, E(t) = e'il, a discounting factor, where i is a discount rate, T = a terminal time of the planning horizon. 4.2 THE OBJECTIVE FUNCTION The overall objective of the model is to maximize a SBU's strategic value, including discounted values of its market share and technological capabilities at the terminal time, while subtracting the total tangible and intangible costs over the planning horizon. The dual-criterion objective fiJnction depicted in equation (4.1) addresses the trade-offs between two competitive priorities -- cost efficiency and flexibility. This function includes both tangible costs and intangible costs related to a system's flexibility. Tangible costs include both the capital investment cost for the acquisition of automation capacity and flexibility of new technology and the production variable costs. Intangible costs include the costs of over-flexibility and under-flexibility represented by the difference between achieved flexibility and required flexibility in terms of both volume and product (part) variety. Also included in the objective fiinction are the costs related to the discrepancies between actual demand level and business goal level in terms of both volume and product variety. Minimization of these costs guides the acquisition decision for new 74 advanced technology to maintain the consistency between the firm's strategic focus and its technology acquisition decision as a part of functional strategy. Max [le(T) + szm + 63km + Gis(T)1E(t) — like? + sz2 + <2er + (bi + (4. 1) bz)p+ c4(p— V)2 + c5(q — m)2 +<-31(v—V)2 +ez(m — M)2}Edt The salvage values, G1v(T), G2m(T), G3k(T), and G4s(T), correspond to the market valuation of the firm's market share in volume and variety and the valuation of the capability of the firm's production technology at the terminal time. These values also provide linkages for the continuation of the firm beyond the planning horizon (see also Roth et a1. 1991). The first term, claz, represents the acquisition cost for automation capacity of new technology at time t. The term, czfQ, is the acquisition cost for flexibility of new technology at time t. The term, c3r2, is the cost for scrapping or reducing conventional (existing) capacity. The model represents the costs for acquiring automation capacity of new technology and its flexibility and scrapping/reducing conventional capacity as quadratic9 since it is assumed that large acquisition of new technology is more difficult to assimilate into the existing system than the smaller acquisitions (see also Gaimon 1990; Farley et al. 1987; Roth et al. 1991). With the same analogy, it is assumed that the larger the conventional capacity scrapped/reduced, the more destructive it is for the production system to reorganize. The total variable costs are denoted as the multiplication of (b 1+b2), the sum of the average volume-related variable cost and the average variety- related variable cost, by p ,the actual production level in units. These first four terms in the integration are the tangible costs, which are directly related to cost efficiency as a strategic priority. 9 For the purpose of simplicity, the quadratic cost functions for the technology acquisition and attrition are assumed as found in the literature (see studies of Gaimon and Roth et a1). For the sensitivity analysis as the extension of this study, it may be possible to assume different cost functions. 75 The next two terms in the objective function are the costs related to the system s flexrbllity in terms of both lead time and product variety. The first term c4(p v)2 rs the cost for over flexibility and under-flexibility of product volume flexibility represented by the squared difference between actual production volume and market demand in volume The second term c3(q - m)2, is the cost for over-flexibility and under flexibility of product mix flexibility, described as the squared difference between the number of products (parts) types produced by the system as achieved product (part) rnrx flexibility and the number of products (parts) types demanded by the market as requrred product (part) mix flexibility. The last two terms, e1(v-V)2 and e2(m—M)2, are the costs related to the firm 3 strat eOic consistency depicted as the discrepancy between the firm 5 actual market posrtron and the desired market position of the strategic business unit (SBU). In thrs model both desired level of volume and desired level of variety as business goals V and M are aSSIJI‘ITLed to be detemrined exogenously according to the firm 5 competrtrve strateoy and the p Fojection of future market conditions as discussed in section 3.1 A srmrlar quadratic representation like these two terms can be found in the model developed by Roth et a1 “9 9 1 ) (see also Holt et al. 1960; Bergstorm and Smith 1970; Chano and Jones 1970) Roth et a1 (1991) explain the term in their model as the squared difference between actual market share and desired market share as the SBU's goal such that "when the actual market position exceeds the goal, the firm incurs a penalty cost because its oroanizational Sim QtLlre and resources are strained. On the other hand, an opportunity cost is incurred Wh e 1'1 the firm underachieves its goal level of demand. Any devratron from the desired bu ‘ 1 hess coal has a Significant impact on the firm's competitiveness and lone term survrval et al.1991) Unlike the model of Roth et al., this model considers not only the sales V6 1 1‘ll‘lle but also the product (part) variety as the elements of market position. The firm' 5 ab 1 1 lty to introduce more products to the market is directly related to the firrn' s "In Q Vativeness, which rs another key element for the firm' 5 competitiveness and long term ¥ 76 survival. Hence, the last two terms in the objective fiinction measure the firm's strategic competitiveness in two dimensions of both product volume and product variety. As described above, the objective fianction of this model consists of three parts, tangible costs combining fixed investment costs and production variable costs, costs related to the system's flexibility, and costs related to the firm's strategic competitiveness. The relative importance recognized by a firm can be accommodated by manipulating the values of cost coefficients for three parts. Particularly, the trade-offs between cost efficiency and flexibility can be explored by changing the values of cost coefficients for flexibility, c4 and c5, in the objective fimction. Technically, the small values of cost coefficients for flexibility, c4 and c5 , relative to those for tangible costs, cl, c2, c3, b1, and b2, means the firm's strategic priority is focused on cost efficiency. On the contrary, the large values of cost coefficients for flexibility relative to those for tangible costs means the firm's strategic priority is focused on flexibility. Hence, depending upon the firm's strategic focuses according to the values of these coefficients the firm will make different optimal acquisition decisions for new advanced technology. To reflect the top priority of the strategic "fit" emphasized by most manufacturing firms (see Fine and Hax 1985), the values for cost coefficients of strategic competitiveness, el and e2, might be given the highest values among the cost coefficients. 4.3 CONSTRAINTS In this section, the constraints for state variables will be presented followed by the upper bound conditions for control variables. The factors not included in the constraints ar e assumed to be constant. 4 - 3- 1 Change in Market Demand g A firm may increase total demand in two separate ways. First, a firm can increase 77 the total demand by capturing the portion of the market share of its competitors through the acquisition of new technology. Second, the total demand can be increased according to the natural growth or decay of the market. As discussed previously, in this study the market demand is viewed in two dimensions, volume and variety. Hence the state equations for the change in market demand are depicted by equation (4.2. 1) and equation (4.2.2) in terms of volume and variety respectively. As discussed in section 3.2, both demand response fiinctions are formulated like the well-known Vidale-Wolfe sales response model (Vidale and Wolfe 1957; see also Bass 1980). YV and Ym are total market size in terms of volume and variety respectively, whereas v and m are the firm's actual market demand in terms of volume and variety respectively. v' = 7t1a(YV - v) + TCZV, v(0) = v0 (4.2.1) The acquisition of automation capacity in equation (4.2. 1) serves as the market stimulus. Since 7:1 represents the market responsiveness to the acquisition of automation capacity, the term, rt1a(YV - v), denotes the fraction of the competitors' market share in volume captured at any instant time due to the acquisition of automation capacity(see also Kotabe and Murray 1990). The market growth/decay parameter, 7:2, in equation (4.2. 1) characterizes the shape of the curve of the exogenous total market demand in volume of the PLC. In this study, the firm's aggregate demand is assumed to follow the typical S- Shaped curve according to the conventional PLC theory. Hence, the term, 7t2V, in equation (4.2.1) denotes the instantaneous effect on the firm's demand in volume due to the changes of the external market forces at a particular stage of PLC. m' = ¢lf(Ym - m) + (bzm, m(O) = m0 (4.22) The state equation for demand variety as depicted by equation (4.2.2) is very k 78 similar to equation (4.2. 1) except for the market stimulus factor. As in equation (4.2.1), the parameter, 6m, in equation (4.2.2) denotes the market responsiveness to the acquisition of flexibility for increasing the firm’s product (part) variety of the market demand. It may represent the customers’ perception for the finn's innovativeness to introduce more new products (parts) in the market due to the improved flexibility (see Swamidass and Newell 1987). The term, (Ym - m), represents product (part) types that are available and unique in the market, but recognized by customers as product (part) types not produced by the firm. Accordingly, the first term, ¢1f(Ym - m), denotes the fraction of market product (part) variety captured by the firm at any instant time due to the acquisition of flexibility. The parameter for natural market growth/decay in variety, (b2, characterizes the changes in product (part) variety along the PLC curve (see also Butler 1988). Hence, the term, (132m, represents the total exogenous change in the firm's demand variety induced by other competitive factors than the acquisition of new technology. According to the formulas given in equation (4.2.1) and equation (4.2.2), both v and m are nonnegative for all te[O,T]. 4.3.2 Change in Variable Costs As discussed in section 3.3, the production variable costs are comprised of both volume-related costs and variety-related costs. In equation (4.3.1), the parameter, 0L, denotes the effectiveness in structural downward shift of the average volume-related cost curve due to the acquisition of automation capacity. According to a study by Buzzell and Wiersema (1981), the reduction in the per unit production and inventory cost is found Proportionate to the level of the per unit production and inventory cost at time t (see also Roth et al. 1991). The acquisition of automation capacity, a, causes this structural shift of the volume-related cost. Accordingly, the term, - eta, represents the fraction of the Volume-related cost, b1, to be reduced due to the acquisition of automation capacity at time t. As defined before, BI is the lower bound of b1. Hence, b1 approaches k 79 asymtotically B1 as t goes to infinity. Since new process technology has low level of manual activities involved, the volume-related learning effect becomes insignificant (see section 3.3). Hence, the volume-related learning effect is not considered in this model. b1'=-0La(b1-B1), b1(0):b10 (4.31) b2' 2 B(S'/S)b2, 132(0) 2 bzo (4.3.2) The next state equation (4.3.2) depicts the change in the average variety—related cost due to the acquisition of flexibility. The parameter, [3, denotes the proportionate increase in the average variety-related cost due to the increase in product (part) variety that the system processes. The term, B(s'/s), represents the fraction of the average variety-related cost that changes according to the change in the product (part) variety processed by the system, s, as illustrated in Figure 3.5. As discussed in section 3.3, the term, s'/s, shows that the increase in the average variety-related cost is decelerated as the cumulative level of overall system flexibility increases. According to the formulas given in equation (4.3.1) and equation (4.3.2), it follows that both b1 and b2 are nonnegative for all te[O,T]. 4.3.3 Change in Production Capacity The change in the production capacity in units of output, k', is depicted in equation (4.4) as the net difference between the acquisition of automation capacity and the scrapping/ reducing of conventional (existing) capacity. k' = a - r, k(0) = k0 (4.4) 4.3.4 Change in Cumulative Capacity of New Technology The change in total accumulated capacity of new technology is represented by the 80 acquisition rate of new capacity as expressed in equation (4.5). In the scenario analysis, the initial value of x, x(0), determines the initial values of v, the firm's market demand in volume, and b1, the average volume-related cost. According to equation (4.5), x is nonnegative for all te[O,T]. This variable is included to calculate the cumulative investment for automation capacity. x' = a, x(0) = x0 (4.5) 4.3.5 Change in Overall System Flexibility The overall system flexibility at time t is defined as the number of products (parts) processed by the system at time t. The state equation (4.6) expresses the change in the overall system flexibility due to the acquisition of flexibility. The parameter, p1, is the effectiveness factor in assimilating the flexibility embedded in the acquired new technology into the total production system to improve the overall system flexibility. This effectiveness factor is quite similar to the synergy factor introduced by Meredith and Cam (1989), which represents the synergy effect among different technologies to reduce in the average production cost due to the acquisition of new technology (see also Roth et al. 1991). Hence, p1 is a maximum system effectiveness representing the organization's ability to assimilate the flexibilities of different technologies. The high value of p1 shows the high effectiveness between the existing technology and the newly acquired automation that realizes the full potential of newly acquired flexibility in improving the overall system flexibility (see also Ettlie and Reifeis 1987; Schroeder 1990). However, the system's flexibility can be reduced due to the scrapping/reduction of conventional or existing capacity. It is assumed here that the system's overall flexibility decreases proportionately by the scrapping/reduction of conventional or existing capacity. In this study, both p1 and D; are assumed fixed. 81 S' = Plf' 92F , 8(0) 2 50 (4-6) 4.3.6 Inequality Constaints According to the definitions of automation and flexibility in this model, the acquisition of flexibility should be accompanied by the acquisition of automation. To ensure this, the inequality constraint (4.7.1) is required, where P is a large number. This constraint ensures the nonacquisition of flexibility required in the case of the nonacquisition of automation. It also permits the case of the positive acquisition of automation accompanied by the nonacquisition of flexibility, in which the acquired automation merely increase the capacity of existing automation without any change in the level of overall system flexibility. The constraint (4.7.2) depicts the relationship between p, actual production volume and q, actual production in variety. The explanation for constraint (4.7.2) is analogous to that for constraint (4.7.2) Ha-fZO (4.7.1) Wp - q 2 0 (4.7.2) p gk_wq (4.73) q S 5 (4.7.4) k 2 0 (4.7.5) The constraints (4.7.3) and (4.7.4) depict that acual production in volume and variety, p and q, are bounded by actual production capacity in volume and variety, repspectively. Note that actual production volume capacity will be reduced proportionately as the number of product (part) types increases as shown in the right hand side of the constraint (4.7.3). The last (4.7.5) is the non-negativity constraints for the production capacity at time t. 82 4.3.7 Variable Boundaries The upper and lower bounds for control variables are presented in equation (4.8). The maximum rate of new technology capacity acquisition at time t may be constrained by factors, such as budget (Gaimon 1985), the ability of the organizational infrastructure to assimilate the new technology, and the availability of the technology (Roth et a1. 1991). The factors affecting the maximum rate of flexibility acquisition may be similar to those for new technology capacity acquisition. The upper bound on reducing existing old capacity may depend on restrictions due to labor contracts, the ability of the organization to make production process changeovers, and the impact of such changes on the organization (Roth et al. 1991). The upper bounds for p and q, P and Q, are depicted by the constraints (4.7.3) and (4.7.4). a e (O, A), f e (O, F), r e (O, R), p e (O, P), and q e (O, Q). (4.8) 4.4 SUMMARY In summary, the optimal control model presented in this chapter contains a multi- criterion objective function to maximize the total value of a firm over the planning horizon. The model has seven state constraints, six inequality constraints, and the boundary constraints for the control variables. The whole model is depicted below. Max [G1v(T) + sz(T) + G3k(T) + Gis(T)]E(t) — Ema? + c2f2 + c3r2 + (b1 + b2)p+ c4(p- V)2 +c5(q - m)2 +el(v—V)2 + e2(m — M)2}Edt subject to v‘ = n1a(Yv- v) + thv, m' = ¢1f(Ym-m) + 2m , b1' = ' (13031-3), 83 132' = B(S'/S)b2 , S':Plf'92F, Ha-fZO, WP-qZO, psk-wq, qu, k20, V(O) = V0, m(O) = m0 b1(0) = b10> b2(0) = bZO, W» = k0, x(0) 2 X0, s(O) = so, and a E (0, A), f E (O, F), r E (O, R), p E (O, P), and q E (O, Q) CHAPTER V SOLUTION APPROACH This chapter presents both the analytical and the numerical solution approaches for the optimal control model presented in Chapter IV. Section 5.1 discusses the analytic optimal solutions for the acquisition of new technology and the scrapping/reducing of conventional capacity. Section 5.2 describes the discretized non-linear version of the model to solve with GRG2 in Microsoft Excel®. 5.1 OPTIIVIAL POLICIES This section first displays the necessary optimality conditions from the Hamiltonian equation (5.1) and the Lagrangian equation (5.2). Then it presents the optimal policies for the acquisition rate of new technology and the scrapping/reducing rate of conventional capacity. 5.1.1 Necessary Conditions The Hamiltonian equation corresponding to the model presented in Chapter IV is given in equation (5.1), where M, M, 7&3, I4, 15, X6, and k7 are the adjoint variables related to the state variables, v, m, b1, b2, k, x, and s, respectively, and E is the discounting factor as defined in section 4.1.3.210 g 10 variable. The interpretation for each of the adjoint variables is the marginal value of its respective state 84 85 H = AIV' + sz' + A3b1' + A4132. + Ask. + A6X' + A75. - {0132 + C2f2 + c3r2 + (b1 + b2)!3 + 04((P - CDZ + Cs(q - m)2 + 6M - V)2 + eztm - M)2}E, (5.1) Due to the inequality constraints in section 4.3.6, a Lagrangian equation is formed as L = H + m(Ha - f) + n2(Wp - q) + T13(k - wq - p) + m(s-q), (5.2) where n1, n2, T13, and 114 are Lagrange multipliers. The last nonnegativity constraint for the production capacity is not considered explicitly in the above equation. Later, during the numerical solution procedure if r(t) - a(t) > k(t) occurs for any time t E [0,T] , then the algorithm will modify the value of r(t) in order to make r(t) - a(t) = k(t) so that k(t) 2 O is not violated. The optimality conditions for the model are presented below. Equations (5.3) through (5.9) are the same constraints as presented in Chapter IV with the initial values for the state variables. Equations (5.10) through (5.21) are derived from the Lagrangian equation (5.2) by taking the first derivatives with respect to the state and adjoint variables. Detailed expressions for the conditions from (5.10) through (5.16) are shown in Appendix I. The complimentary slackness conditions of (5.22)-(5.25) are needed for the inequality constraints (4.7.1)- (4.7.4) to be satisfied. The necessary conditions and the complementary slackness conditions for the optimality of the model are; v' = 7t1a(YV - v) + nzv, V(O) = v0, (5.3) m' = ¢1f(Ym-m) + (bzm, m(O) = m0 , (5.4) b1’=-aa(b1-B), b1(0)=b10, (5-5) b2'= B(S'/S)b2, b2(0)=b20, (5-6) 86 k'=a-r, k(0)=k0, (5.7) x' = a, x(0) = x0 , (5.8) S' = Prf- 92E 5(0) = 50, (5.9) l1'= - ESL/5v, K1(T)= G1E(T), (5.10) 12' = - ESL/6m, M(T) = G2E(T) , (5.11) 7L3‘ = - 6L/5b1, 13(T) = O, (5.12) M' = - oL/bbz, 14(T)= O, (5.13) 15' = - ESL/5k, 15(T) = G3E(T) , (5.14) 7&6' = - oL/bx, K6(T) = O , (5.15) k7' = - ESL/55, M(T) = G4E(T) , (5.16) 6L/5a = O, for a = (O, A) , (5.17) ESL/5f: O, for f= (O, F) , (5.18) ESL/5r = O, for r = (O, R), (5.19) 6L/5p = O, for p = (O, P), (5.20) ESL/Sq =0, forq=(0, Q), (5.21) n120,Ha-f20,m(Ha-f)=0, (5.22) 712 20,Wp-t120m2(Wp-q)=0, (5.23) n3ZO,k-wq-p20,n3(k-wq-p)=0,and (5.24) 11420, s-q20,n4(s-q)=0. (5.25) 5.1.2 Optimal Policies for Acquiring New Technology In Theorem 1, the optimal policies for the control variables, a(t) and f(t), satisfying both the necessary conditions and the complementary slackness condition in section 5.1.1, are stated. Theorem 1: The optimal policies for the acquisition rates of automation capacity and flexibility are depicted below. According to the complementary slackness condition of 87 (5.22), there are two possible cases for the optimal policies for the acquisition rates of automation capacity and flexibility. Case 1. If Ha(t) - f(t) > 0, then n 1(t) = O and a(t) > O and f(t) 2 0. Hence, 9,0), if0<6,(t) 0, then m(t) = O, p(t) > 0, and q(t) 2 0 Hence, _ P(t), ife (t)ZP(t) (5.“2) P“) ‘ {9pm, if 0p<9p(t) 1-3 Flexibility ' l stands for introductory stage, (3 stands for growth stage, and M stands for maturity stage. In Table 6.1 the possible choices of the combination of strategic priorities during the evolution of the PLC are given: (1) choice l-l, in which a firm's strategic priority is changing from flexibility for the introductory stage, and flexibility/cost efficiency for the growth stage, to cost efficiency for the mature stage; (2) choice 1-2, in which the strategic priority is always cost efficiency; and (3) choice 1-3, in which the strategic priority is always flexibility. Choice 1—1 is the one generally supported by the literature (Hayes and Wheelwright 1984; Hill 1980). In Table 6.2 the PLC is assumed to evolve from the introductory stage (I), through the growth stage (G), and then back to the growth stage (RG) again. The possible strategic choices are (1) choice 2-1, which is the same as choice 1-1 in Table 6.1; (2) choices 2-2 and 2-3, which are the same as choices 1-2 and 1-3, respectively; and (3) choice 2-4, which is similar to choice 1-1 except that it includes flexibility for the RG stage. 97 Table 6.2 Strategic Choices and the Reversed PLC Strategic Choice Reversal of PLC G RG' 2-1 2-2 2—3 2—4 Flexibility Cost Efficiency Flexibility Flexibility F lexibility/ Cost Efficiency Flexibility p > Flexibility! Flexibility Cost Efficiency " RG stands for reversed growth stage. 6.1.2 Positioning of Initial Process Technology There are two structural variables related to the positioning of initial process technology: (1) so, the initial level of flexibility and (2) Xo, the initial level of automation. As shown in Figure 6.1, there are four types of initial process technology with different combinations of so and Xo; (1) Firm A with high so and low Xo; (2) Firm B with high so and high Xo; (3) Firm C with low so and low Xo; and (4) Firm D with low so and high Xo. In Figure 6.2 the combination of levels of these two variables determines a firm's initial process technology in the technology matrix. 98 F L A B E “'9“ High 30 High 50 I Low xo High x0 8 | L C D | Low T Low 80 Low 80 Y Low X0 High X0 Low High AUTOMATION Figure 6.1 Four Types of Initial Technology Choice F ng’l L E X I B l L l T . 0 Low Low Autanation Higi Allorration AUTOMATION Figure 6.2 Positioning of Initial Technology Choice According to the normative concept of the Product-Process Matrix, Firm A, with high flexibility and low economies of scale, is suitable for the production of products of high variety and low volume. On the contrary, Firm D, with low flexibility and high economies of scale, is suitable for the production of products of low variety and high volume. Firm B, with high flexibility and high economies of scale, can operate efficiently in a wide span of operating environments from the high variety/low volume stage to the low variety/high volume stage. Firm C is also included in the analysis, although this has 99 not been considered as a practical setting in a static environment. However, there is yet no conclusive theory for the best initial technology choice in a dynamic environment such as the one in this study. 6.1.3 Settings of Variables The values of some exogenous variables as defined in section 4.1.3.2 change according to the different combination of levels of the factors discussed in section 6.3.1. In Table 6.3 the variables whose settings are changing according to the evolution of the PLC include (1) Yv(t), the total market size in volume; (2) Ym(t), the total market size in variety; (3) 7:2, the growth/decay factor of a firm's market share in volume; and (4) (1)2, the growth/decay factor of a firm's market share in variety. The settings for YV(t) and n2 are the same for all stages of the PLC. The settings for Ym(t) and (1)2 are also the same for all the stages of PLC. This is due to the fact that both 7:2 and ()2 follow the behavior of the total market size according to the evolution of the PLC. Table 6.3 Variable Settings for the Stages of the PLC Stages of PLC Variables I G M RG Yv(t) MG FG so MG let) FG FD so so 1!: MG FG SG MG 4), FG FD so so ' F stands for fast. M for medium. S for slow, G for growth, and D for decline 100 The settings of Yv(t) —- the total market size in volume -- are (1) medium growth at the introductory stage; (2) fast growth at the growth stage; (3) slow growth at the mature stage; and (4) medium growth at the reversed growth stage. The settings for Ym(t) -- the total market size in variety -- are (1) fast growth at the introductory stage; (2) fast decline at the growth stage; (3) slow decline at the mature stage; and (4) slow growth at the reversed growth stage. Table 6.4 Variable Settings for Strategic Priorities Strategic Priority Variables ' ' ' Flexibility C°5t Efficiency /Co:tleExflfti)cl:lieyncy C4 H' L M Cs H L M W) L H M M(t) H L M As depicted in Table 6.4, variables related to the strategic priority are (1) C4, costs for inflexibility of lead time; (2) c5, costs for inflexibility of variety; (3) V(t), the business goal of the market share in terms of volume; and (4) M(t), the business goal of the market share in terms of variety. The settings of C4 and c5 are (1) high values for C4 and 05 when a firm's strategic priority is flexibility; (2) low values for C4 and C5 when a firm's strategic priority is cost efficiency; and (3) medium values for C4 and c5 when a firm's strategic priority is flexibility/cost efficiency. The settings of V(t) are (1) low for the priority of flexibility; (2) high for the priority of cost efficiency; and (3) medium for the priority of flexibility/cost efficiency. The settings of M(t) are (1) high for the priority of flexibility; 101 (2) low for the priority of cost efficiency; and (3) medium for the priority of flexibility/cost efficiency. Table 6.5 Value Settings of Research Variables Related to the Strategic Choice High Medium Low C4 100 50 25 C 5 200 100 50 “v .002 .001 .0005 um .01 .005 .0025 The values corresponding to the levels of research variables related to the strategic choice are depicted in Table 6.5. The values of C5 are twice that of C4 at each level to reflect the difference in the acquisition cost between automation capacity and flexibility. The value settings for V(t) and M(t), the SBU's goals of the market share in terms of volume and variety, are represented by the growth factors -- Liv and Hm -- in the functions defined below. It is assumed that V(t) = Yv(t) (c + ttv(t-i)) and Mt) = Ym(t) (d + umrt-i». where both C and d are .l at time 0 and represent the firm‘s market share at the starting time of each stage of the PLC, and i is the index of the initial period of each stage of the PLC. 102 As depicted in Table 6.6, the variables related to structural factors -- so and x0 -- are (1) blo, the initial average volume—related cost and (2) b20) the initial average variety- related cost. The settings of these two variables are (1) high b10 and low bzo for Firm A; (2) low b10 and low bzo for Firm B; (3) high blo and high b20 for Firm C; and (4) low blo and high 1320 for Firm D. Table 6.6 Variable Settings for the Positioning of Initial Technology Choice Initial Production Cost Initial Technology Choice V olume- V ari ety- x. .. “flit?" Riéiii“ A L H H H B H H L H C L L H L D H L L L " H stands for high and L for low Table 6.7 Value Settings of Research Variables Related to Initial Process Technology High Low X0 (bio) 150 (25) 50 (30) SO (b20) 100 (15) 10 (10) 103 The value setting for these two variables are presented in Table 6.7. The values in parentheses are the initial values for average volume- and variety- related variable costs -- blo and b20 -- corresponding to each level of xo and so. These values are determined after the preliminary runs. The high value of bio and b20 are set arbitrarily at 30 and 10 with xo at 50 and so at 10. After running several problems with different settings of the other variables, the values of b10 and b20 corresponding to the values of x and s, 150 and 100 are interpolated from the solutions. As a result, the value for b10 is 25 when xo is 150 and the value of b20 is 15 when so is 100. 6.1.4 Values of Exogenous Variables The values of the base scenario is presented in Table 6.8. Most of the values in the base scenario are similar to those in the study by Roth and Gaimon (1991). However, there are some exogenous variables unique to this model. The values of these exogenous variables are set after some experimentation with various levels for each one. The values of G1, G2, G3, and G4 -- the marginal values at the terminal time of the firm's market share in volume, V(T), the firm's market share in variety, m(T), the capacity in volume, k(T), and the system's flexibility, s(T) -- are set at 300, 600, 50, and 100, respectively. The values of C 1, C2, and C3 -- the acquisition cost of automation and flexibility, and the scrapping cost of existing or conventional capacity -- are set at 50, 100, and 20, respectively. The cost of flexibility, C2, is set as twice the cost of automation, C1, in order to reflect the fact that the flexible system costs an enormous amount compared to a mere automated system such as the stand-alone NC machine. The values of E1 and E2 -- the costs related to the strategic fit between business strategy and corporate strategy -- are set at 1000 and 2000, respectively. These values are larger than others because of the high priority given to the fit among strategies in different levels. During the preliminary runs, these two values have been found as major factors 104 influencing the effectiveness of initial process technology on the optimal policy for technology acquisition. This will be further discussed in detail later in Section 6.2. Table 6.8 Value Settings for Exogenous Variables Parameters Value G] 300 G2 600 (33 50 G4 100 C] 50 C2 100 (:3 20 E] 1000 E2 2000 R] .00002 9] .0005 or .002 [3 1 B 1 Di 8 pz .1 MA 200 ME 40 MR 20 \l/ .5 H l .0E+10 W 1.0E+10 The values of the market responsiveness factor, 7:1 and (b1, are set at .00002 and .0005, respectively. It has been found that these two values have a significant impact on the overall cost performance of each problem during preliminary runs (see also Roth and Gaimon 1991). Since the purpose of this dissertation is not to study the sensitivity analysis of research parameters with the model here, the values of these factors are 105 determined at reasonable levels after Close scrutiny of the results of the preliminary runs. The values of effectiveness factors for the variables costs, or and B, are set at .002 and .1, respectively. It has been found that these two factors have a significant impact on the size of investment of both automation capacity and flexibility. However, they are not affecting the general pattern (policy) for the acquisition of automation capacity and flexibility. The minimum value of the volume-related cost is assigned at l. The effectiveness factors for the system's flexibility, p1 and p2, are set at .8 and .1, respectively. Within the reasonable range of the values for these two factors, there has been no significant impact on the Optimal solution with the base scenario. The upper bounds for the control variables —- a(t), f(t), and r(t) —- are set at 200, 40, and 20, respectively. The value of w, the effectiveness factor of scrapping existing capacity(r(t)) on the total capacity on volume(k(t)), is set at .8. This factor is also found to be an insignificant factor affecting the optimal solutions. The values of H and W are set at both l.0E+lO as a large number. 6.1.5 Hypothetical PLC Curves The difference between the normal PLC curve and the reversed PLC curve has already been discussed in Chapter II. Tables 6.9 and 6.10 depict the discretized values for 1:2 and (192 for the normal PLC curve and the reversed PLC curve. n2 defines the shape of the total market size of volume, and (b2 the shape of the total market size of variety. The shapes of the PLC curves used for the scenario analysis are illustrated in Figures 6.3 through Figure 6.6. The curves in Figure 6.3 and Figure 6.5 depict the aggregate demand in units of output by combining the market demand of all products in a family. The total market demand of variety is also the aggregate demand of all products in the same family. It is assumed that Yv(t) = Yv(0)(a+32(t-i>>, Yv(0)=1000, and Ym(t)= Ym(0)(b+¢2(t'i)), Ym(0)=100, 106 where both a and b are the exponents for the values of YV and Ym at the starting time of each stage of the PLC, both 7:2 and (1)2 are the growth factor of YV and Ym determined exogenously, and i is the index of the initial period of each stage of the PLC. Table 6.9 The Values of Variables of the Normal PLC n2 (1)2 YV Ym Introductory 0 .02 .20 1000.0 100.0 stage 1 .04 .50 1020.0 120.0 2 .10 .70 1060.8 180.0 3 .20 .30 1166.9 306.0 Growth 4 .30 .10 1400.3 397.8 Stage 5 .30 .01 1820.3 437.6 6 .30 -.01 2366.4 441.9 7 .30 -.10 3076.4 437.5 8 .30 -.40 3999.3 393.8 9 .20 -.30 5199.1 236.3 10 .10 -.20 6238.9 165.4 11 .05 -. 10 6862.7 132.3 Mature 12 .03 -.05 7205.9 119.1 Stage 13 .02 -.01 7422.1 113.1 14 .01 -.01 7570.5 112.0 15 .01 -.01 7646.2 110.9 16 .01 -.01 7722.7 109.8 17 .01 -.01 7799.9 108.7 18 .01 -.01 7877.9 107.6 19 .01 -.01 7956.7 106.5 20 .01 -.01 8036.2 105.4 The shape of the normal PLC curves can be found ubiquitously in the literature. The shape of the reversed PLC curves as in Figure 6.4 and Figure 6.6 are yet unpopular. 107 Generally it is argued that the demand of volume after the growth stage of the reversed PLC continue to grow without entering into the maturity stage, and also the demand of variety continues to rise after the growth stage (Albernathy et al. 1983; Ayres and Steger 1983) Table 6.10 The Values of Variables of the Reversed PLC 7:2 2 Yv Ym Introductory 0 .02 .20 1000.0 100.0 Stage 1 .04 .50 1020.0 120.0 2 .10 .70 1060.8 180.0 3 .20 .30 1166.9 306.0 Growth 4 .30 .10 1400.3 397.8 Stage 5 .30 .01 1820.3 437.6 6 .30 - 01 2366.4 442.0 7 .30 - 10 3076.4 437.5 8 .30 -.20 3999.3 393.8 9 .20 -.10 5199.1 315.0 10 .10 -.05 6238.9 283.5 11 .05 -.01 6862.7 269.3 Reversed 12 .01 .01 7205.9 266.7 Growth 13 .05 .05 7277.9 269.3 Stage 14 .05 .10 7641.8 282.8 15 .05 .10 8023.9 311.1 16 .05 .10 8425.1 342.2 17 .03 .05 8846.4 376.4 18 .02 .03 9111.8 395.2 19 .01 .01 9294.0 407.1 20 .01 .01 9387.0 411.1 108 9000 -- 80m .. 7000 - Total Market Size in Volume Time Figure 6.3 Changes in Total Market Volume of the Normal PLC 8 o 8 o 8 o 200 -» 100 -- Total Market Size in Variety Ott‘fttittrkitttiitf. 1 4 7 101316 19 Time Figure 6.4 Changes in Total Market Variety of the Normal PLC 109 10000 ~- 9000 - 8000 ~ 7000 - 6000 ' 5000 - 4000 ' 3000 - 2000 - 1000 - Total Market Size in Volume Figure 6.5 Changes in Total Market Volume of the Reversed PLC 8 c> ‘8 300 -- 200 ‘- 100 -- Total Market Size in Variety Figure 6.6 Changes in Total Market Variety of the Reversed PLC 6.1.6 Frameworks of Design There are two frameworks of design for the scenario analysis according to the type of PLC. Table 6.11 presents a design framework with the normal (conventional) PLC, whereas Table 6.12 presents a design framework with the reversed PLC. There are two factors for the design (1) strategic Choice and (2) initial technology choice. In Table 6.11 there are three levels of the first factor: (1) choice 1-1; (2) choice 1-2; and (3) choice 1-3. 110 In Table 6.12 there are four levels of the first factor: (1) Choice 2-1; (2) Choice 2-2; (3) Choice 2-3; and (4) Choice 2-4. There are the same four levels of the second factor in both Table 6.11 and Table 6.12: (1) Firm A; (2) Firm B; (3) Firm C; and (4) Firm D. The number of replications of the first design, as depicted in Table 6.11, is twelve, whereas for the second design there are sixteen replications. Therefore, the total number of replications for the scenario analysis is twenty-eight. This number will increase exponentially as more variables are included for other sensitivity analyses. The actual settings for the research variables for all cases are depicted in Table 6.13 and 6.14. Table 6.11 Design Framework with the Normal PLC Factors Levels Strategic 1_1 1_2 1_3 Choice Initial Technology A B C D Choice Table 6.12 Design Framework with the Reversed PLC Factors Levels Strategic 2-1 2-2 2—3 24 Choice Initial Technology A B C D Choice 111 Table 6.6 Settings of the Research Variables 11 v “m EEEEEEQEA 4...} 4 mEmmzmgzg :4 cx'i m mzmmzmAZJ 4E < EEEEEEAEJ :m D EEEEEEEJAJ ~44 2 U mmmmmmuduz :4 66 3 m mmmmmmuuqz 43 < EEIEEEAAJE III 0 555555.412: AA N <4: 0 Aqqqqqmm: 421-1 08 2 m qqqqqqm m 53—15: < qngqqmma. 2:: Q EEACEEAA :2 4.1.4 Z EEAEEAA 3‘ :4 9'8 '7 32432.44 : :25: < 3243.12.41.12“ :2: a) CD C3 C3 .2 $2 94 iii 2’. 2 8% ~08~08~oe L) O— .2 5:0 2 2 2 GOES! g0 73% xmnn % H 112 o_ o_ 2 E o_ o_ 2 m. o_ o. 2 2 o_ o_ 2 2 in mm on mm om mm om mm om mm om mm om mm om mm om So 2 2 oom o2 2 2 oo_ o2 2 oz o2 ooz o. oo oo_ oo_ om owe om omfi om omo om omo om om_ om omo om om_ om oom om oz S. S. S. S. S. S. S. S. mmoo. mmoo. mmoo. mmoo. mmoo. mmoo. mmoo. mmoo. oz 5 2 woo. woo. woo. woo. S. S. S. S. mmoo. mmoo. mmoo. mmoo. moo. moo. moo. So. 0 ...: go. _o. _o. _o. _o. _o. _o. _o. mmoo. mmoo. mmoo. mmoo. _o. _o. _o. _o. _ mooo. mooo. mooo. mooo. mooo. mooo. mooo. mooo. moo. moo. moo. moo. moo. moo. moo. moo. oz 8 2 So. So. So. So. 88. mooo. mooo. mooo. moo. moo. moo. moo. So. So. So. So. 0 >1 mooo. mooo. mooo. mooo. mooo. mooo. mooo. mooo. moo. moo. moo. moo. mooo. mooo. mooo. mooo. H oom oom oom oom oom oom oom oom om om om om on om om om mm a 2 o2 o2 o2 oom oom oom oom oom om. om. om om oom o2 o2 o2 0 mo oom oom oom oom oom oom oom oom om om. om om oom oom oom oom o ooo o2 o2 o2 oom ooo o2 o2 mm mm mm mm mm mm mm mm om a 2 om om om om oo. oo_ 2: oo_ mm mm mm mm om om om om o 5 o2 o2 o2 o2 o2 o2 o2 o2 mm mm mm mm o2 o2 oom o2 E 32858. a o m < a o m < a o m < o o m < 3865 3:5 mm mm as 2 mm as m-_ Tm mo Z 866 BmeEm SEES> :888m of mo mmcfiom o:_o> he 053. 113 6.1.7 Performance Measures The optimal solution in the spreadsheet of Microsoft Excel® generates the total objective function value, the dichotomized values of the terms in the objective fiinction, and the values of control variables and state variables. The possible performance measures comparing different scenarios include (1) optimal policy for the acquisition of automation and flexibility, (2) optimal policy for the scrapping/reducing of conventional capacity, (3) total automation capacity and the system's flexibility at time T, (4) total cost performance, and (5) total flexibility performance. The first and second performance measures might provide managers with insight for the optimal investment policy in the presence of a specific corporate strategy. The fourth and fifth performance measures might help corporate executives to understand how they should plan their corporate strategies by bearing market demand behavior in their minds. The third performance measure indicates how a firm's process technology evolves in the product-technology box differently according to the choice of initial process technology and corporate strategy. 114 6.2 RESULTS OF ANALYSIS One interesting observation from the results of all the cases is that there is no scrapping of existing or conventional capacity. Hence, the scrapping of conventional or existing capacity will not be discussed. 6.2.1 Analysis with the Normal PLC As described in Chapter V, there are two research factors: strategic choice and initial process technology. For analysis with the Normal PLC, there are three levels of strategic Choice —-1-1( mixed I ), 1-2 (cost efficiency), and 1-3 (flexibility) -- and four levels of initial process technology -- A (high automation/low flexibility), B (high automation/high flexibility), C (low automation/low flexibility), and D (low automation/ high flexibility). Hence, there are twelve cases in analysis with the Normal PLC. 6.2.1.1 Effect of Initial Process Technology In this study a firm is assumed to Choose one of four different initial process technologies, such as A (low automation/high flexibility), B (high automation/high flBXibility), C (low automation/low flexibility), and D (high automation/low flexibility). Table 6.15 depicts the optimal acquisition of automation capacity for twelve cases with the Normal PLC, categorized according to the type of initial process technology. Figure 6.7 illustrates the results in Table 6.15 with graphs. Each graph depicts optimal policies (patterns) for automation capacity acquisition corresponding to three different strategic priorities for a given initial process technology. Since all four graphs look almost exactly the same, it can be readily concluded that there is no effect of initial process technology on the optimal acquisition policy for automation capacity. 115 Table 6.16 and Figure 6.8 illustrate the optimal acquisition of flexibility, categorized according to the type of initial process technology. Again, there is no effect of initial process technology on the optimal acquisition policy for flexibility. This is because the objective of the strategic fit in the objective fiinction of the model has a greater weight than the others. Table 6.17 and 6.18 display the cumulative automation capacity and the cumulative system's flexibility. Figures 6.9 and 6.10 present the results in Tables 6.17 and 6.18 with graphs. Since optimal acquisition policies for both automation capacity and for flexibility with four different initial process technologies are the same under a given strategic choice as depicted by Figures 6.7 and 6.8, the cumulative levels of production capacity as depicted by four graphs in each figure are also identical. The upper half of Table 6.31 displays the cost performance of cases with the Normal PLC. It is also found that there is no effect of initial process technology on the overall cost performance. No effect of initial process technology is caused by the highest weight being placed on el and e2, the penalty cost for the strategic fit between an SBU's actual market position and its goal for the market position in terms of volume and variety. For example, suppose an SBU starts its operation with an initial process technology that has a high level of automation capacity, and its corporate strategy is cost efficiency. The SBU's goal for the volume demand growth, therefore, will be set at a high level and, in addition, the initial market share is the same for all cases by the assumption. Hence, there should be the acquisition of automation capacity to boost the market demand in volume in order to minimize the excessive penalty of the strategic fit between actual market share and the business goal in volume, even though there are other penalty costs incurred by the acquisition of automation capacity. Additional experiments with low values of el and e2 (10 and 20) show an effect of initial process technology on the acquisition of flexibility (refer to Tables 6.32 - 6.35 and Figures 6.25 - 6.29). According to Figure 6.31, Firm C and Firm D, which have a low level of initial flexibility, invest in flexibility more than Firm A and Firm B, which have a 116 high level of initial flexibility. However, there is no difference in the acquisition of automation. This is because the same initial production capacity, k(O), for all cases. 6.2.1.2 Effect of Strategic Choice To analyze the effect of strategic choice, the results of cases with the Normal PLC are categorized in terms of strategic Choice as depicted by Tables 19 through 22. The results of these tables are graphically illustrated by Figures 6.11 through 6.14. Each graph depicts optimal policies for flexibility acquisition according to the four different initial process technologies within a strategic Choice. Since there is no effect of initial process technology on the optimal acquisition policy, each graph in Figure 6.11 and 6.12 shows almost a single line. The acquisition policy of automation capacity with strategic choice 1- l in Figure 6.11 depicts a skewed investment during the mature stage because of the strategic priority of cost efficiency pursued during that stage. The acquisition of automation capacity with strategic choice 1-2 is substantial over the three stages, tailing down toward the end of the mature stage. Since strategic choice 1-2 has the strategic priority of cost efficiency throughout all three stages, a firm would try to purchase automation capacity. This is partly because of the savings made from the economies of scale, and partly because of the need to boost its SBU's market share in volume to meet the SBU's goal for high growth in volume demand despite the relatively expensive acquisition cost of automation capacity. This is why the acquisition of automation capacity in the cases with strategic choice 1-2 is greater than the cases with the other two strategic choices. The last graph illustrates the acquisition of automation capacity for the cases with strategic Choice 1-3. Since the strategic priority for these cases is flexibility all the time, there is no significant investment in automation capacity partly due to the low growth goal for the market share in volume and partly due to a relatively high value for the unit acquisition cost of automation capacity compared to the other cost coefficients in the objective fiinction. 117 In Figure 6.12 the investment in flexibility is low for those cases with strategic choice 1-2 as compared to the other cases. Since the strategic priority is cost efficiency for all periods with strategic choice 1-2, a firm would not invest in flexibility due to the relatively expensive unit acquisition cost of flexibility and a low priority given to the objective of flexibility. The comparison between strategic choice 1-1 and strategic Choice 1-3 in Figure 6.12 indicates that the optimal acquisition policy for the flexibility of the two cases is almost identical during the introductory stage where the total market demand in variety increases sharply. Then, the acquisition of flexibility for the cases with strategic choice 1-3 becomes greater than those with strategic choice 1—1 until the end of the time horizon. The initial high investment in flexibility with these two strategic Choices is caused by the high priority given to flexibility during the introductory stage and, consequently, the relatively inexpensive unit acquisition cost of flexibility. The second reason is related to the objective of minimizing the penalty cost of strategic fit by boosting the actual market demand in variety, which meets the goal of high market share growth in variety pursued during the introductory stage with both strategic Choices. The greater investment in flexibility that strategic choice 1-3 shows compared to l-l after the introductory stage is due to the difference in strategic priorities. Figure 6.13 and Figure 6.14 depict the cumulative levels of automation capacity and the system’s flexibility, respectively. In Figure 6.13 cumulative levels of automation capacity for cases with strategic Choice 1-2 increase at a faster rate than the cases with the other two strategic Choices reflecting the high investment with the choice 1-2 as depicted by Figure 6.11. The cases with strategic choice 1-3 show the slowest growth of the cumulative levels of automation capacity. However, in Figure 6.14 the cases with strategic Choice 1-3 show a faster growth of the cumulative levels of the system's flexibility than the cases with the other two strategic choices. Here, the cases with strategic choice 1-2 show the slowest growth of the cumulative level of the system's flexibility. Figure 6.13 and 6.14 show that the cases with strategic choice 1-1 have a 118 medium level of both automation capacity and system flexibility at the end of the mature stage, compared to those cases with the other two Choices. The cases with strategic Choice 1-2 have the highest level of automation capacity, but the lowest level of system's flexibility at the end. The cases with strategic Choice 1-3 have the lowest level of automation capacity, but the highest level of system's flexibility at the end. As depicted by Table 6.31 there is again no effect of initial process technology on the total cost performance. Among the four cases within each strategic choice, there is no significant difference in the total cost performance. Among the three strategic choices, strategic Choice 1-3 show the lowest total cost performance. Strategic choice 1-2 is the worst in terms of total cost performance. 6.2.1.3 Conclusion The results of the cases with the Normal PLC suggest that when the market demand follows the pattern of the Normal PLC, the manufacturing firm pursuing a strategy of cost efficiency should focus on investment in a rigid process technology with high automation and low flexibility, whereas a firm pursuing a strategy of flexibility or differentiation should focus on the investment for flexibility rather than the investment for automation capacity. These results are managerially intuitive since a firm with the strategic priority of cost efficiency wants to operate in the market with high volume and low variety, and a firm with the strategic priority of flexibility wants to operate in the market with low volume and high variety. The first firm should have a process technology which allows it to achieve the economies of scale for price competition, and the latter one should have a process technology which allows it to achieve the economies of scope for a product-variety competition. The low total cost performance for the cases with strategic choice 1-3 may be due to their low investment in automation capacity. However, these cases have greater total penalty costs for flexibility and strategic fit compared to the other cases with the Normal 119 PLC. The most cost effective performances for flexibility and strategic fit are given by the cases with strategic choice 1-1, although they do have the highest total investment cost for technology acquisition compared to the cases with the Normal PLC. Considering the importance of the objectives of flexibility and strategic fit, the strategic Choice with the Normal PLC should be 1-1 in which the strategic priority of a firm changes according to external market conditions. 6.2.2 Analysis with the Reversed PLC In the analysis with the Reversed PLC, the first three levels of strategic choice -- 2- 1 (mixed one), 2—2 (cost efficiency), and 2-3 (flexibility) -- are the same as the three strategic Choices with the Normal PLC. The last choice, 2-4, is another mixed strategy in which the strategic priority is changing from flexibility for the introductory stage, through cost efficiency/flexibility for the growth stage, to flexibility for the mature stage. The levels of initial process technology are A, B, C, and D, the same as for the analysis with the Normal PLC. Therefore, the total number of cases with the Reversed PLC is sixteen. 6.2.2.1 Effect of Initial Process Technology The results of the analysis with the Reversed PLC also indicate no effect of initial process technology. Tables 6.23 and 6.24 present the acquisition of automation capacity and flexibility with the Reversed PLC. The results are categorized according to the type of initial process technology. These results are illustrated graphically in Figures 6.15 and 6.16. Tables 6.25 and 6.26 display the cumulative automation capacity and the system's flexibility with the Reversed PLC. Figures 6.17 and 6.18 illustrate the results in the two tables graphically. Again, all the graphs in each figure are almost identical. This is again due to the high weight on the penalty cost (el and e2) for the strategic fit. 120 6.2.2.2 Effect of Strategic Choice Tables 6.27 and 6.28 display the results for the optimal acquisition of automation capacity and flexibility with the Reversed PLC, categorized in terms of strategic choice. Figures 6.19 and 6.20 illustrate graphically the results in Tables 6.27 and 6.28, respectively. In Figure 6.19 each of the results with strategic Choice 2-1 (Mixed Strategy) and 2-2 (Cost Efficiency) shows almost the same pattern of optimal acquisition policy for automation capacity as does the counterpart with the Normal PLC. With strategic choice 2-3 (Flexibility), there is some investment in automation capacity and significant investment in flexibility during the reversed growth stage which is quite different from the result for strategic Choice 1-3 with the Normal PLC as depicted by Figures 6.12 and 6.13. This is mainly because of the continuing increase in market demand in terms of both volume and variety during the reversed growth stage. The optimal acquisition policy for automation capacity with strategic Choice 2-4 in Figure 6.19 shows a steady investment over the three stages. However, the optimal acquisition for flexibility with strategic choice 2-4 in Figure 6.20 shows a significant amount of investment during the reversed growth stage compared to the cases with strategic choices 2-1 and 2-2. One interesting observation from both Figure 6.19 and Table 6.31 is that the cases with strategic choice 2-4 have the lower total investment amount in automation capacity, and yet they allow firms to achieve a slightly higher market share in volume compared to those with strategic choice 2-3. In addition, with the strategic choice 2-4 firms can even have a higher level of cumulative automation capacity at the end of the mature stage than do those firms with strategic Choice 2-3. This is due to the difference in optimal acquisition policies for automation capacity between the two strategic choices, depicted in Figure 6.19. In Figure 6.19 strategic Choice 2-3 shows bulky investment during the growth stage and the reversed grth stage, but strategic Choice 2—4 shows a steady and incremental pattern of investment over the three stages. Because of the quadratic cost 121 function for the acquisition of automation capacity in the objective function, a bulky investment costs more than a steady and incremental investment. Tables 6.29 and 6.30 display cumulative automation capacity and flexibility, respectively, and are categorized in terms of strategic choice. Figures 6.21 and 6.22 depict the results of the two tables. These graphs show that the cases with strategic choice 2-1 generate process technology with medium levels of cumulative automation capacity, and flexibility at the end. The cases with strategic choice 2-2 generate process technology with the highest level of automation capacity and the lowest level of flexibility at the end. The cases with strategic Choice 2-3 have the process technology with the lowest level of automation capacity and the highest level of flexibility at the end. The cases with strategic choice 2-4 have the process technology with a low but slightly higher level of automation capacity than strategic choice 2-3, and a high but slightly lower level of flexibility than strategic choice 2—3. This reflects the presence of cost efficiency/flexibility priority with strategic choice 2-4 during the growth stage as compared to flexibility priority with strategic choice 2-3 during that stage. During the growth stage the cost efficiency/flexibility priority of strategic choice 2-4 encourages investment in automation capacity, but deters the investment in flexibility a little more than the flexibility priority of strategic choice 2-3. In Table 6.29 the cumulative automation capacities for the cases of 2-3 are 888.5 and 988.5; those for the cases of 2-4 are a little higher, 946.0 and 1045.9. In Table 6.20 the cumulative system's flexibility for the cases of 2-3 are 470.0 and 381.0, and those for the cases of 2-4 are a little lower, 395.0 and 305.0 The bottom half of Table 6.31 depicts the cost performance of all sixteen cases with the Reversed PLC. For total cost performance, the best strategic choice is 2-4, which has a changing strategic priority congruent to the external market conditions. The worst strategic Choice is 2-2 in which the strategic priority is cost efficiency for all three stages. 122 6.2.2.3 Conclusion Optimal investment policy for process technology for a firm operating in the market with the Reversed PLC seems to be the same as when operating in the market with the Normal PLC. Again, the results of analysis with Reversed PLC indicate that the investment policy is high on the acquisition of automation capacity and low on the acquisition of flexibility when the firin's strategic priority is cost efficiency. The results also indicate that when the firm's strategic priority is flexibility a firm should invest more in flexibility and less in automation capacity. The best strategy with the Reversed PLC is 2-4. The reason why the choice 2-3, flexibility, is inferior to 2-4 is that an SBU's goal of a relatively low growth in volume demand discourages the investment in automation capacity during the growth stage, which incurs an enormous penalty cost of the strategic fit later during the reversed growth stage. Compared to strategic Choice 2-1, strategic choice 2-4 has relatively low total penalty costs of strategic fit. This indicates that the firm's strategic priority has to be congruous with the conditions of the external market environment. However, the cases with strategic choice 2-4 have a higher total flexibility cost than those with strategic choice 2-1. Considering the high unit penalty cost of flexibility for the cases with strategic choice 2-4, one can see the difference is insignificant between the two choices. The structure of process technology at the terminal time has the most important strategic implication . Since the market following the pattern of the Reversed PLC will have the continued growth of volume and variety in the future, the process technology with high flexibility is key for survival in competition. Considering the importance of the system's flexibility in the future, the best Choice would be 2-3. However, strategic choice 2-3 is considerably inferior in terms of total cost performance to strategic Choices 2-4. In conclusion, the best strategic Choice with the Reversed PLC is strategic choice 2-4, which is a combination of strategic priorities: flexibility during the introductory stage, cost efficiency/flexibility during the growth stage, and flexibility during the reversed 123 growth stage. Again, with the Reversed PLC the structure of initial process technology has no effect on the overall performance and the optimal acquisition policy for process technology when the strategic fit in the hierarchy of strategies is considered the most important objective. 6.2.3 Comparison Between the Normal PLC and the Reversed PLC This section is devoted to the discussion of the interaction effects of three research factors -- the PLC, strategic choice, and initial process technology -- on the optimal acquisition of new process technology, on the evolution of process technology, and on overall cost performance. 6.2.3.1 Optimal Acquisition of New Process Technology In Table 6.31 the investment for new process technology for each case with the Reversed PLC is generally greater than the counterpart with the Normal PLC. This is due to the continuing growth of both volume and variety demand in the Reversed PLC. For example, the total investment cost with strategic Choice 2-3 is almost fifteen percent higher than that with strategic choice 1-3. However, there is no significant difference in the investment amount in both automation capacity and flexibility between strategic choices 1-1 and 2-1, and between strategic choices 1—2 and 2-2. There are no substitutions of new capacity for conventional capacity in any the cases of with both the Normal PLC and the Reversed PLC. This is because the scrapping cost outweighs the savings from the substitution with new process technology, which are mainly the decrease in volume- or variety- related costs. Finally, strategic choice 2-4 with the Reversed PLC has a slightly higher investment in automation capacity, but somewhat lower investment in flexibility than strategic choice 1-3 with the Normal PLC. 124 6.2.3.2 Evolution of Process Technology Figures 6.23 and 6.24 illustrate the evolution of process technology for the Normal PLC and the Reversed PLC, respectively. The first point of each line is the initial position of process technology. Each of the other points in each line represents the position of process technology at the end of each stage of the PLC. There is no difference in the pattern of evolution of process technology between the Normal PLC and the Reversed PLC for each case. For cases with the first (1-1 and 2-1) and the second (1-2 and 2-2) strategic Choices , there is little difference in the position between the Normal PLC and the Reversed PLC. However, with the strategic choice of flexibility, such as 1-3 or 2-3, the structure of process technology at the terminal time has higher levels of both automation capacity and flexibility with the Reversed PLC than with the Normal PLC. With strategic choice 2-4 the pattern of evolution of process technology shows a slightly higher level of automation capacity, but a slightly lower level of flexibility for all three stages than with strategic Choice 2-3. Table 6.31 provides detailed information regarding the level of automation and flexibility at the end of the mature stage for each case. In Table 6.31 the general salvage value of process technology for each case with the Reversed PLC is higher than the corresponding case with the Normal PLC. This reflects the higher amount of investment in process technology for the cases with the Reversed PLC than for those with the Normal PLC. One interesting observation is the higher level of the system's flexibility with strategic Choice 2-4 compared to that with strategic Choice l-3, although the total investment amount in flexibility with strategic choice 2-4 is lower than is with strategic choice 1-3. This is again caused by the bulky investment in flexibility during the introductory stage with strategic choice 1-3 shown in Figure 6.12, compared to the steady investment with strategic choice 2-4 shown in Figure 6.20. Table 6.31 shows that there is little difference in the cumulative automation capacity and a system's flexibility at the end between strategic choices 1-1 and 2-1, and 125 between strategic choices 1-2 and 2-2. However, strategic Choice 2-3 has a terminal level of cumulative automation capacity and system flexibility that is almost fifteen percent higher and more than twenty percent higher, respectively, than strategic Choice 1-3. Strategic Choice 2-4 has almost the same level of cumulative system flexibility as strategic choice 1-3, but an almost twenty percent higher level of cumulative automation capacity than strategic Choice 1—3 at the end. This high level of cumulative automation capacity at the end is due to the combined effect of continuing growth of market demand in both volume and variety with the Reversed PLC and the strategic priority of cost efficiency/flexibility during the growth stage with strategic choice 2-4. 6.2.3.3 Cost Performance Generally, the total cost performance of each of the first three strategic choices with the Reversed PLC is inferior to the counterpart with the Normal PLC. The first two strategic Choices, mixed I and cost efficiency, do not vary much in terms of the total cost performance between the two types of PLC. However, the third strategic Choice of flexibility with the Reversed PLC has a total cost much higher than does the one with the Normal PLC. This higher cost is due to the considerable difference in the last four penalty costs between the two types of PLC as depicted in Table 6.31. 6.3 SUMMARY OF FINDINGS In this section the findings from the scenario analysis are summarized. The findings related to the acquisition of new process technology are summarized in Section 6.3. 1, followed by those related to the evolution of process technology in Section 6.3.2, and finally those related to the best strategic choice are summarized in Section 6.3.3 126 6.3.1 Acquisition of New Process Technology 1. According to the results of this study, a firm pursuing the strategic priority of cost efficiency invests in new process technology more than does a firm pursuing flexibility does This may be due to the relatively small value for the unit acquisition cost for flexibility compared to that for automation capacity. The results would be different if the value for the unit acquisition of flexibility were significantly higher than the value in this study. The cost of automation capacity is generally for hardware, such as tools, control devices, and other automated equipment whereas the cost of flexibility is mostly for software, such as control programs for a CNC machine or an FMS. Since the cost for hardware is much higher than the cost for software, the values for the acquisition costs of both automation capacity and flexibility used in this study seem to be fairly practical. Therefore, it may be safe to conclude that the cost efficiency priority generally leads a firm to a greater capital investment in new process technology than does the flexibility priority. 2. It is found that a firm whose strategic priority is flexibility invests more in the system's flexibility than in automation capacity, as compared to low investment in automation capacity and high investment in flexibility for a firm whose strategic priority is cost efficiency. 6.3.2 Evolution of Process Technology 1. According to the results of this study, the size of investment in process technology is larger with the Reversed PLC than with the Normal PLC. 2. There is little difference between the Normal PLC and the Reversed PLC in the pattern of evolution of process technology when a firm's strategic Choice is changing priorities (flexibility-cost efficiency/flexibility-cost efficiency) or cost efficiency. 127 3. With its strategic priority of flexibility a firm should invest more in both automation capacity and flexibility when it Operates in a market with the Reversed PLC than when operating with the Normal PLC. 6.3.3 Best Strategic Choice 1. According to the results of this study, for overall cost performance a firm with cost efficiency priority is greatly inferior to the one with the flexibility priority in both types of the PLC. This is mainly due to the huge investment cost for automation capacity for the case with the strategic priority of cost efficiency, although the total production cost with the strategic priority of cost efficiency is lower than that with the priority of flexibility. 2. Based on the results of total cost performance, it can be concluded that a firm operating in a market with the Normal PLC should choose a strategic priority of flexibility, and that a firm operating in a market with the Reversed PLC should select its strategic priority congruent to the conditions of the external market environment. 3. A firm can achieve the best performance for both volume and variety flexibilities when it selects its strategic priority congruent to the external market environment. The firm can achieve both high volume and high variety flexibilities with the strategic choice of 1-1 when operating with the Normal PLC and with the strategic choice of 2-4 when operating with the Reversed PLC. 4. It is found that the acquisition policy of new process technology is really a strategic decision. A decision making process for technology acquisition should consider the strategic fit in the hierarchy of strategies. 6.3.4 Conclusion Due to the enormous computational time for each case, the parameters and variables other than the experimental factors have fixed values for the scenario analysis in 128 this study. It may be difficult to generalize the results of scenario analysis unless there is sufficient number of experiments with different settings of those parameters and variables. However, based on the findings from the scenario analysis in this study, it is found that a firm's corporate and/or business strategy is a key factor affecting an optimal acquisition policy for process technology. The initial positioning of process technology by a firm is found to affect the optimal acquisition policy only when a firm considers the strategic priority as an insignificant factor for the technology acquisition decision. The external market environment also affects the optimal acquisition policy. Particularly when a firm's strategic priority is flexibility, the eflect of external market conditions on both total cost performance and on an Optimal acquisition policy for new process technology becomes significant. In conclusion, the results of this scenario analysis demonstrate the importance of the technology acquisition decision in fitting the corporate or business strategy to external market conditions. For the optimal acquisition of new process technology, a firm has to select its strategic priority congruent to external market conditions. Finally, the findings from the scenario analysis in this dissertation support the argument by Wickham Skinner that the technology decision has to made strategically. 129 0.0 0.0 0.0 9 0.0 0.0 0.0 mp 0.0 0.0 0.0 m? 0.0 0.0 0.0 @— m.NF 0.0m _..mm 3 mdp 0.0m Tmm 2 mm..— m.0m Cam 0’ mdr 0.0m Sam 0— edp m.0m 0.00 t ed? m.0m 0.00 : e.mp mdm 0.00 5 e0? m.0m 0.00 t 0.0— méor 0.00 S 0.0. méor 0.0m 0P 0.2 méop 0.00 0? 0.2 méow 0.00 0— 0.: m0: 0N0, my 0.9 0.0: mdoo m— 0.2 m0: 0N9 m— 0.9 md: 0&2 mp oer 0K: e.e0P 3 we? 0.5: e.e0F e_. 0.3 OS: e.e0— er mew 0K: e.e0F er 0.0m .N.NN— néop mp 0.0m «.mwp 50— mp 0.0m ~.NN— héop m— 0.0N Ndm— ESP 0. N.0N m.0NP mew NP ~0N 0.02 mew NP Now 0.09 mem NF N.0N 0.0“: mew NP e00 0.9.: mém : e00 9mm— mém 2 e00 0.mm— mém 2 e00 0.mmp mém : 0.0m 5.0? Was 0F 0.0m 500— 5.05 or 0.0m fimmp fimn 0— 0.0m 500— 52. 0F e00 550— m.00 m 0.00 Comp e00 0 e00 Hump m.00 m 0.00 523 e00 m 0.3. eémr 0.0m 0 men e40— m.mm 0 men e._.mF mdm 0 men eémv 0.0m m 0S0 To? 0.0m m. 0.50 F02 0.0m n 0.50 v.00? 0.0m N. 0K0 —.om— 0.0m m. 0.e0 0.03 m.mm 0 0.e0 0.0m: m.mm 0 0.e0 0.09 mdm 0 0.e0 0.0“: m.mm 0 ~._.0 NéNF N.0m m ~15 «.2: Ndm m N40 N42 Ndm m NS NS: Ndm m mumm 9m: The e 5mm 9m: fine e mumm 00: fine e 5.0m 06: The e oée 0.0: 0Ke 0 fire 0.0: wfie m 0.; 0.0: mfie m hée 0.0: ofie m 0.0m 0.02 0.0e N 5.0.0. 0.02 0.0e N 0.0m 0.03 0.0e N 0.0m 0.02 0.0e N 0.2 QMNP m.me F TNF was m.me F 0d— m.mNP mde P 2: mam: mde P 0.0 mm: 0.0e 0 0.0 m0: 0.0e 0 0.0 m0: 0.0e 0 0.0 m0 5 0.0e 0 m; N; T— m; N; TF 0; N; T— m; N; —.F 83:0 20325 0 U m < $805 RES 015 38.82 05 ~23 guano 5:38:34 mo 50683504 m M .0 2an 130 0.0 0.0 0.0 0. : 0.0? 0.0.. 0.0_. 0.: 0.3. 0.0w 0.0V 0.0P 0.0.. e00 0.00 0.00 0.00 0.00 0.00 v.00 0-? 0.0 0.0 0.0 Oe we .3 .2. we 0.e we we 0.e _..e 0.e 0.0 0.0 0.0 0.0 0.0 0.0 0-? 0.0 0.0 0.e 0e 0.e 0.0 e0 0.0 0.0 OK 0.0 0.0 for 00—. e.0_. 0.0 0.00 0.00 o.e0 0.00 To mOFNmVl-OCONQC) f—FFV—‘I—PPFFF OFNWVLOCDNCO 0.0 0.0 0.0 0.: n.0_. 0.0V 0.0V 0.1. 0.: 0.0r 0.0r 00—. 00* v.00 0.00 0.00 0.00 0.00 0.00 _..00 04 0.0 0.0 0.0 0.e we _..e re re o.e re re 0.e we 0e 0.0 0.0 0.0 0.0 0.0 0.\. 0-? O: 3582 2o as, $583 0o 553% 0.0 0.0 0.e 0.e 0.e 0.0 e0 0.0 0.0 0.\. 0.0 0.0 for 0.0.. v.0? 0.0 0.00 0.00 o.e0 0.00 To GOV—NWVLOCONCDO) g—v-v-t-Y-FY‘V'F“ or00e00rxao 0.0 0.0 0.0 0.2. 0.0? 0.0V 0.0V 0.: 0.3 0.0V 0.0? 0.0V 0.0? v.00 0.00 0.00 0.00 0.00 0.00 v.00 0.0 0.0 0.0 Te we 0.e _..e we we 0.e 0e 0.e he 0.0 0.0 0.0 v.0 0.0 v.0 0.x. 0-_. 0 H .0 @351 0.0 0.0 0.e 0.e 0.0 0.0 0.0 0.0 0.0 0.0 0.\. 0.0 for 0.0.. 0.0_. 0.0 0.00 060 0.00 v.00 mOv-NWVLDCDINQO) V—T-V-Y—FT-FFFF QV—NmeON‘D 0.0 0.0 0.0 0.: 50—. 0.0? 0.0? 0.3. 0.3 0.0r 0.0V 0.0? 00—. v.00 0.00 0.00 0.00 0.00 0.00 0.00 0% 0.0 0.0 0.0 we re 0e we we we. 0.e 0e 0.e Ne 0.0 0.0 0.0 v.0 0.0 e0 0.x. 04 0.0 0.0 0.e 0.e 0.0 0.0 0.0 0.0 0.0 0.\. 0.0 0.0 for 0.0? 00—. 0.0 0.00 0.0 0.00 v.00 To QOFNC‘OVLDCONQO) V—T—T‘Y—rrv—FVF Ov-Nmfl’mwh‘m 029.0 20290 $890 32:. 0.00 .9502 on. 5.3 3.80.0 20.86930 00 2006803.. 0.0 820.0 0.2:. 0FOFn—0POFeF0—0—Zow 0 0 N. 0 0 e 0 0 F 0 THE... I T-IT..-il-..!T|IT|IT.. -. it .-I -..l. .. al.»-.. Til. I . Ta- .011.|T| 0.0 zo.so--oe:ol-¢//¢/x \\b\\ ./ . n 000 W In. 0/ S O Onaslxlxl odo. a - /U/.U A 7.001-...0/00- . \. _ m Imp/.4 \C- C..:. . IC 0 ......i. . n 0.00. I. 0.. IIIIOII‘ 0.. |-,IIC!-III 7.. 1.111% .. 0.000 4.1. 0 >00402100h 0000000 .7152. A. 1 0.2g 0FOFh—0—0Fe—0—0PFFO— 0 0 h 0 0 e 0 0 — 0 o 0 x. «r . w .r a L. L. . . L. . 0 0 I- 0.0 1120-9}:O'O/ . n 000 m I. S O II\I odo. .3 $1.. LTD/DI??? m fe/fif- \xa-‘5\.a - & C‘C n 0.00— I. 9. lol ~2— l'Uul ... Ill-III 0 0.000 0 >00402I00... 0000000 ..<_..._2_ 02F 0FOFn—0F0—eF0—0PZOP 0 0 n 0 0 e 0 0 . 0 T-l..|¢|... . 0 » o L. . ..lll.-T|-T.l-.-I|.T-«Tilll 0.0 l \ . 91030!o!9/6/, \b ’,/ . n 000 W I. \.\. m \l 0.09 m D . - \D/C/ -- \: n luv/A 0.14% . \AQIA..\D\~Q D AU uld. . -..-.. - n 0.00. 1. 0—lll‘ nélaiTl .—'|-||_ 0 . 0.000 0 >00002100» 0000000 :_<_._..Z_ 05:... 0.0—0—0—0Pe90—0— :0. 0 0 N. 0 0 e 0 0 — 0 v T .. 0 u 0 L. 0 0 . . o 0 ¢ 0 o a r DO.O IOI§-'Ol‘§/O/ / . n /n 08 m I. S 0 .0.00— .3 J m I. .d n .. 0.0.: I. 9.111% ~.—l.0ll Ell-Ill .. 0.000 < 50.62.83 3000.... ._<_.:z_ 2 3.. 3552 2: .23 E30250 850:3... 3 2%.... 0.2; 0.2; 0—0Ph—0—0—eP0—0P—F0—0 0 n 0 0 e 0 0 _. 0 0—0PNFOPOFeF0—0—Z0—0 0 N 0 0 e 0 0 F 0 fl .. 0 .p +1| 0 w 0 0 0 0 0 . L. o . p 0.0 . 0 . . 0 0 0 w 0 o . ¢ . . o . . . Tl 0.0 @flo n HflflfluflflflflmflmlmeTflwflzU/ : sflwTUHUHUHUHUHEHmHmLmI 3 ll/ 5\ / I. .\-Q\ ll . . n o /l/ 0 . n /o/o/? I/ co. m /9/0/ I/ co. m /O|IO O /O - O O / I 0.8 a / o.o~ a o / um \ f. M / < / / m / /u/ w o/ol /,l a /o I a I 0.8 M 0.8 u n..||:0-| SluCl ... IIII Slot! Elllflnll ... Ill! 0.0e 0.0e 0 >00402100h 0000000 .0352. 0 >OOAOZIUOP 0000000 ..(.....Z. 05.? 0.2: 0—0—0—0—0—ePOF0—ZOP0 0 N. 0 0 e 0 0 F 0 0—0th0—0Pe—0F0—PPOF0 0 h 0 0 e 0 0 F 0 v 0 0 o 0 ¢ 0 o b. o . . Tl-» « . T . T- 0.0 V 9 .. 0 o . . a 0 L. a o 0 ¢ 0 T 0 o L. . 0.0 A/C U “CHUNG! III-z- G DIC/D/DIWTDIUJ: M . nTu Dflfllmll IIGIGID/UIDIQDIO/nr 9/ /l/I/ 0 0.0. m ¢/¢ /I . 0.0. m /./ all! I ,/.l/ cl 0 0101;. a TTQII 0 /¢! 0 // o 6 0.00 M --- 0.00 M / / /./JK/f m \a// m 3 .4.— odm M l odm M. 2 0|: a; Ul:| .-. II 9. Cl 3 IIIDI ... III! . 0.0? 0. 0.0V 0 >00402100h 00000.00 ..<_0._Z_ < >00..02100... 0000000 05:2— 1/ 133 e000 0.090 0.003 00 e005 0.0000 0.000« 00 0.000 0.090 0.003 00 e005 0.0000 0.002 00 0.000 «.5000 «.053 0« 0.005 «.5000 0.050« 0« 0.000 «.5000 «.053 0« 0.005 «.5000 0.050« m« 0.30 0.5000 0.0«e« 0« 0.e«5 0.5000 0.0«0« 0« 0.30 0.5000 0.0:; 0« o.e«5 0.5000 0.0«0« 0« 0.000 0.300 0.«00« 5« 0.005 0.«0«0 0.«00« 5« 0.000 0.500 0.«00« 5« 0.005 0.«0«0 0.«00« 5« 0.505 0.030 e.e00« 0« 0.500 0.300 e.e0«« 0« 0.505 0.030 0.e00« 0« 0.500 0.0.50 e.e0«« 0« «.055 0.0000 m.«0«« m« «.050 0.000« 0.«00« 0« «.055 0.0000 0.«0«« 0« «.050 0.000« 0.«00« 0« 0.005 0.000« «.500« 3 0.000 0.000« 0.500 3 0.005 0.000« «.500« e« 0.000 0.000« 0.500 e« 0.005 0.000« 0.000 0« 0.000 0.005« 0.000 0« 0.005 «.000« 0.000 0« 0.000 «.005« 0.000 0« «.0«5 0.«50« 5.000 0« «.0«0 0.«50« 0.005 0« «.0«5 0.«50« 5.000 0« «.0«0 0.«5m« 0.005 0« 5.050 0.000« 0.0e5 «« 5.050 0.00e« 0.0e0 «« 5.050 0.003 0.0e5 «« 5.050 0.00.; 0.0e0 «« 0.«00 «.500« 0.050 0« 0.«00 «.500« 5.050 0« 0. «00 «.500« 0.050 0« 0.00 «.500« 5.050 0« e000 e.000« «.000 0 0.00... e.00«« 0.000 0 e000 e.000« «.000 0 0.00». e.00«« 0.000 0 «.05.. «.00 «« 0.0em , 0 0.050 «.000« e.0ee 0 «.05e «.00 «« 0.0.0 0 0.050 «.000« e.0ee 0 0.««e 0.000 e.00e 5 0.««0 0.000 0.000 5 0.««e 0.500 e.00e 5 0.««0 0.500 0.000 5 «.20 5.«50 0.09. 0 0.5.0 5. «55 0.000 0 «.20 5.«50 0.09. 0 0.5e0 5. «55 0.000 0 0.000 0.005 0.000 0 «.00« 0.000 0.000 m 0.000 0.005 0.000 0 «.00« 0.000 0.000 m 0.500 0.30 0.«e0 e e.50« 0600 0. «e0 e 0.500 0.e00 0.20 e e.50« 04.00 0.30 e 5.00« 0.0«0 0.e00 0 5.00 0.0«e «.em« 0 5.00« 0.0«0 0.3.0 0 5.00 0.03 «.e0« 0 0.00« 0.000 e.me0 0 «.00 0.000 e.me« 0 0.00« 0.000 e.0e0 0 «.00 0.000 e.me« 0 0.00« 0.000 0.00« « 0.00 0.00« 0.00 « 0.00« 0.000 0.00« « 0.00 0.00« 0.00 « 0.00« 0.00« 0.00« 0 00.00 00.00 00.00 0 0.00« 0.00« 0.00« 0 0.00 0.00 0.00 0 0-« 0-« «-« 0-« 0-« «-« 0-« 0-« «-« 0-« 0-« «-« 8.9.0 2082.0 0 0 0 < $820 EEE 0.5 38.52 2.. 5.3 0.896 5.8883... 3.32:5 5.0 0.0m... 134 0.000 5.000 #.000 0.050 0.000 0.000 0.0#0 0.000 0.000 0.0«0 5. «00 0.00« «.05« 0.50« 0.00« 0.0«« 0.00 5.05 5.00 0.00 0.0« 0-« 0.00 0.00 «.00 0.00 0.05 0.05 0.05 0.00 5.00 #.00 0.00 0.00 #.00 0.0# 0.0# #.«# 0.00 0.00 #.00 0.0« 0.0« 0-« 0.00« 0.00« «.00« 5.00« 0.05« 0.05« 0.05« 0.00« 0.«0« 0.00« 0.«0« 0.0#« 0.50« 0.00« #.0«« 0.00« 0.00 0.00 0.«0 0.«0 0.0« «-« ®OPNMVLOCDNQO50 r-v—v-r-w—r-v—v—v-r-N OFNC‘OVLOCONG) 0.000 5.000 #000 0.050 0.000 0.000 0.0#0 0.000 0.000 0.0«0 5.«00 0.00« «.05« 0.50« 0.00« 0.0«« 0.00 5.05 5.00 0.00 0.0« 0-« 0.00 0.00 «.00 0.00 0.05 005 0.05 0.00 5.00 #.00 0.00 0.00 #.00 0.0# 00# #.«# 0.00 0.00 #.00 0.0« 0.0« 0-« U 0.«0« 0.00« «.00« 5.00« 0.05« 0.05« 0.05« 0.00« 0.«0« 0.00« 0.«0« 0.0#« 0.50« 0.00« #.0«« 0.00« 0.00 0.00 0.«0 0.00 0.0« «-« OOPNWVLOCOYNQCDO P'v-r-v-V—v—u—v—v—f-N o—00#0050 #.000 5.000 #050 0.000 0.000 0.0#0 0.000 0.000 0.0«0 0.000 0. «00 0.050 0.000 0.5#0 0.000 «.000 «.00« 0.00« o.##« 0.00« 0.00« 0-« «.05« #.05« 0.05« #.05« 0.50« 0.00« 0.00« 0.50« 0.00« 0.00« 0.5#« 0.0#« 0.0#« 0.00« «.00« «.50« 0.«0« 0.0«« 0.««« 0.00« 0.00« 0-« 0. «00 0.050 0.050 0.050 0.000 «.000 0.000 0.000 0. «00 0.0#0 0.«#0 0.000 0.500 0.0«0 #.000 #.00« #.05« «.00« 0. «#« 0.00« 0.00« «-« 00«00#005000 q—v—v—r-v-I—F-v—I-v-N o—00#0050 0.000 5.000 #050 0.000 0.000 0.0#0 0.000 0.000 0.0«0 0.000 5.«00 0.050 «.000 0.5#0 0.000 «.000 «.00« 0.00« «.##« 5.00« 0.00« 0-« 0.5 .9:qu 2: «EB 3:5.on 982.000 038.255 0 _ .0 0.00.0 «.05« #.05« 0.05« #.05« 0.50« 0.00« 0.00« 0.50« 0.00« 0.00« 0.5#« 0.0#« 0.0#« 0.00« «.00« «.50« 0.«0« 0.0«« 0.««« 0.00« 0.00« 0-« 0.«00 0.050 0.050 0.050 0.000 «.000 0.000 0.000 0.«00 0.0#0 0.«#0 0.000 0.500 0.0«0 #.000 #.00« #.05« «.00« 0.«#« 0.00« 0.00« «-« GOPNMVDOBGDGO Y—I—‘f-T-f-PF'V—Q—I—N o~00#0050 8.9.0 0.03.50 3820 .m.::. 135 our. .mELoZ 2.. «...3 0:2.me :o..wE8:< 03.22::0 0.0 230.0 05:0 b P 000«0«5«0«0«#«0«0«««o«0 0 5 0 0 v m 0 « o 0 5004021000 0000000 450.2. lfldiflO :iO SilNfl 10:!an :10 Slan . 4 . l . . 4 0 . + . T L. o . u% 0.0 0.80 o 910.05010i019H - I... 0.89 \I .-l ..X. I I n. -o 0.08. -..x... mTD\r\C\r N. |Dl . o.ooo~ - ... Ilil III CIUFT . . ocean 0 5902280 mmmooE 2E2. Fill. I: 1" ll I||.|.I -l Incli¢1 0.2; 830.290.390.29 m m 0 0 m e m N — o T|.II$II+|I..I.|$|LI «T 4. 9 0|» 0 0 o 0 o J. L. .r » 0.0 033?. Hon-x \ . 0.80 \1\\ H. o|0|9l9I9i6|6 \I\ I\I\l .-. odoo. l\l\. \C\ I-.l\ _ -.-lviliél .1. . S oi: 082 S .liul . odoom \ I Ill-'1 0100. . 0.800 < >OOAOZIQE. 000000.... ._<.h_2. 0.2; ONBSCSESQSZSmm00¢,va.o T'... .r r 1 JL . b . 00.0 HDHDB . . .-. 40W.- oo 08 o. o. o- o- 0-6-6-60”... .3 m. I\I \« 0. . M .I.\ : oo 08. -.l. .. _ o I I .... .a I .l . ... - I l . 0 ._ 8 com.— n .\.; 3- ....--r . I. \C d U\8 N. m Ex Bl! 8.89.. 0000+Wh0 _; III! 8.8%.. 0 530229.: 0809: 2E2. 0.2: 009322339330me0mmvmu.o v'qT.-l.. ? 0 o .1 J. o .0 A o . a 0 4F 0 . .Ir.. .. 0.0 \0\ \0\ M9 \.\fiflfl\l\l oi \0“ u..- 0.80 n ¢|o\o\0\¢l6l. FHI\ w \ L \l\. Q.S\ 0.08. s \I .\ 0 I - x. ... , .....- i ll Eliot coon: a n... ' I. \a S D 0.88 D¥U\G\D Elli-ll .. edema 6 o: 2552 2: 5E 35%: yeoim 23328 OS 2%; 0.2: on—mphwormerm—NFZOF m 0 N. 0 0 v m N ul rlfll'f ,TllaT I. 1O1G1$1Q1D1D1D1D1 D1010 \,-\ I\I1I\I Q\9 I I '1-11l1l11.‘ \Q\ I I 1 -1 \O 1..- .\.\..\. 01-. -0201 o 9101 MBIHVA :10 8.1.an 0.2:. ONmpmpmpwpm—vwm—N—IOF m 0 n 0 0 v m N . 1 T-|. 1..-- 1.- I-.l-1..|...-l-..1 1. ? .lY'o- 1. , 11.1! ..rPlTIuTIlTI: ...... 0190151019313- Dial . \I\ .1- 11-\.11.\- Q\\ ..1- '11-11. .1\ I- I 1I- I- 91.101 o\¢1¢ \o\01 9191010 n; 9.1:--- 311113.11 E1111I ..... 1 _ MBIUVA :10 SUN” 0 >00402100h 00000mn. ..Stz. 05.; onpmwhporm—VPMFN—PPOF m m m 0 0 v m N F o r 1P L . 1 4 4 4 4 4 . . . 4 4 < 2111191111- :IiDl Z m >00402100h 0000090 03.22. MBIUVA :IO Sian 05.; on—m—hpopmpvwmwNP—POF m 0 h 0 m w m N F o r! p p P p > b p b 1? . 1 14 4 T 1 4 1 1 1< 1- . 4 J $.lllblll Wplllfilli 7. < >00402100H 0000001 .2752. MSIHVA :10 SUN“ 137 «.« «.« «.« «.« «F «.« «.« «.« «.« «F «.« «.« «.« «.« «F m.«. «.«F «.3 m.«_ «P «.«m «.«m «.«m «.«m «P ..«m Sm fimm gm «F 3.9 3.9 3.9 3.9 t «.3 «.3 «.3 «.3 C «.«« «.«« «.8 «.8 C «.«F «.«F «.2 «.9 3 «.SP «gov «.5? «.2: m: «.«« «.«« «.3 «.3 «F «.9 «.9 «.«_ «.«F «P «.o: «.o: «.o: «.o: «F «.«oF «.«E «.«3 «.«S «F «.3 «.3 «.3 «.3 3 cs: es: ox: as: 3 «do, :9 3.9 3.2 3 0.0« 0.0« 0.0« 0.0« «F «.««_ «.««P «.««. «.««F «F :9 :9 «.2: «.8— «— «.«« «.«« «.«« «.«« «F «.«9 «.««_ «.«S «.«S «_ «.¢« «.¢« «.¢« «.¢« «— 3.3 3.3 3.3 3.3 S oh? oh? om? 0.«9 Z «.5 «.5 «.5 «.5 2 «.«m «.«m «.«m «.«m o. «.«2 «.«9 «.«2 «.«3 op «4K «.3 «.«n 5K or 3.8 «.«o 3.3 «.«o « «$9 «.2: «««F «««F « «.8 3.8 «.3 3.8 « «.E «.E «.3. «.E « 3.x: 3.x: 3.2: 3.2: m «.«m «.«m «.«m «.«m « «.3 «.3 «.8 «.8 n 39 _.o«_ 3? <02 n «.3 «.mm «.«m «.«m « 04% 045 04$ «45 o «.«3 «.«S «.«S «.«E « «.mm «.mm «.«m «.«m o «.5 «.8 «.5 «.5 « «._«_ «.—«F «A? «.8. m «.o« «.o« «.8 «.8 m «.«m «.«m «.«m «.«m 3 «.«Z «.«5 «.m: «.«2 3 :3 :3 :3 :3 3 «.3 «.3 «.3 «.3 « «.«Z «.«2 «.«E «.«2 m «.3 «.3 «.3 «.3 « «.m« «.m« o.«« «.«« « 0.««F o.««F 0.««_ o.««F « «.3 «.3 «.3 «.3 « o.«P P.«— o.«F «F P «.«9 «.«S «.«2 «.««_ P «.3 «.3 «.3 «.3 _ 0.0 0.0 0.0 0.0 o m.«: m.«: «.«2 «.«2 o «.3 «.3 «.3 «.3 o o o m < o o m < o o m < 329.63 «809m 62:. m; N; F; «20:0 062.30 0.5 3882 «5 £3» 30830 cocmEoS< b8 20:60:3qu 9 .0 2%? 138 0.0 0.0 0.0 0.: «NV 0.0« 0.0« 0.: 0.: Now 0.0« N0? N0? vNN 0.0N QQN o.mN NMN 0.0N ...mN 0.0 0.0 0.0 0.: «N? 0.0« 0.0? 0.: 0.: N0? 0.0« Now Ndw VNN 0.0N 0.0N o.mN «.mm 0.0N «.mN 0 m-w 0.0 0.0 0.0 0.: EN? 0.? 0.0« 0.3. 0.: N.0.. 0.0« Now Ndr VNN 0.0N 0.0N 0.0N 0.mN 0.0N de m 0.0 0.0 0.0 m. : «NP 0.0« 0.0« 0.3. 0.: N0? 0.0« Ndr N.m_. vNN 0.0N 0.0N 0.0N 0.0N 0.0N 0.0m < or 0? t OOFNmVLDCD V—v-v—V—r-v—V— Ov-NCOVLDCDNCD 0.0 0.0 0.0 0.». iv we. 7v 7v o.v ...v 7v m.v ...v 0..» 0.0 0.0 Na 0.0 0.0 m.\- 000 .2502 2: 0.3 0:55.... 00 8005.090 0.0 0.0 0.0 o.v 0v fv fv ...v o.v 7v 7v m.v 7v 0.». 0.0 0.0 N0 0.0 0.0 ms 0 N-.. 0.0 0.0 0.0 :V fv N.v 7v TV TV NV 0.? 0% 5v 0.0 0.0 OS r... 0.0 v.0 0.x. 0 0.0 0.0 ad fv :V Nd 7v :V :V N.v m.v 0.v NV 0.0 0.0 OR is 0.0 v.0 0.x- < 0N0 0.0«-H oaOu-vamcorxaoo: Y-Y‘T-V—V‘T—T‘T'T—Y— Ov-NC‘OVLOCONCD 0.0 0.0 0v 0v ad N0 v.0 0.0 0.0 OK 0.5 0.0 for QNF v.9 m._.N n.0N o.NN QVN m.\-N 0.0 0.0 0v 0..» WV N0 v.0 0.0 0.0 0...- wk 0.0 70.. 0N? v.0? meN «.mN CNN QVN mKN 0 Yr 0.0 0.0 0v QV 0.0 ND 0.0 0.0 0.0 0.x- 05 0.0 for 0.N.. 0.0« NeN 0.0N meN QMN F.0N m. 0.0 0.0 0v 0.v 0.0 N0 0.0 0.0 0.0 o..- 0.5 0.0 «.0? 0N? 0.0« NéN 0.0N 0...N m.mN ...wN < 0') ‘. OY-NWVLDIONQ FV‘FV—V‘T‘FPP O) OFNOVLOCDNQ 30.05.00» 000005 .050. 00.9.0 05009.0 139 015 .2507. 05 5.3 3.00000 00..080.0< .0 00.0.m.000< :0 20...”. m2: mr w— MP w— mp .1 m— NF :0? m m N. m m v m N P o Trigll» 0 T « . .r w A L 0 0 fi 0 .1 0 0 0 NM 0.0 Q/QIQ!QIQ b.\ 0.00 m «1 \\Q\.\Q on! i - I: it 070. a O o .0 0.00— .5 m U |, O l. . m < III. 0 0.00« m; wU.OIU U.Ow..00..02100.. 00000.: ..<.....2. 00 00— 00. 000 10:1an :10 SUN“ memo :10 SUN!) 148 00.. 8023. 2: 5E 0:303:00 802.02 2.0 230.. m >004021005 000005. ...<.5.z. 0.2.5 0.2.5 0.0.5.0.m.¢.0.0..:o.mm5om....00.o m.m.~..0.m.30.0.:o—mm50mcmm.o vl-lh I-rv|l.llollu ol'loull.l§l||.|lrlltlllo.l... [-oll~ln|9| blI-Ylol'l: 0.0 .DI'...I.'9||I$ » . > y .||I.|-I|o L r r p p L > 0.0 -i. : I I I1 1.. \OiCa...T-C\C-lC-l /C.../.. 1 I- | Iii-05D}?- ij/C--.Cil 4.! I’m-fa Ola/G \\m.. OIICllfllflt‘lalHllfluulfllu IU/U/ 0\\5.. I/.- i I/ Glzi III-l 0.0. m al./ui- 0.0_ m °\O.I|I“-I\la//-L “ \\AVI&‘HO.//. an“ ..x s S \o\ w w \- 0.00 0. 0.00 ./-.-/. f \\O- \O -- o o I \ , A -/ I. \\ \Q\ \\o. 9./- -. . D/ A 1100K. f.-./ .//.m/ w -543... .- .l/ \A..., w O .l I... «N, oz.-. ,/ \ /, . I: :| o/ 3 .1. 0.00 m 2 o .1. 0.0m u 0.0 . - oil: 0.0 1.10.1 - .N.N nilOli N-NII .II 0.0.. c 0.0.. .-N I|I-l.- ..N 11.1: n. >004021005 000005... ..<.5.z. 0 5004021005 000005... ..<.5.z. II‘I I]! II I IIIIL llln I1 0.2.5 0.2.5 0.0—5.0.m.v.m.0...o.mm5omv00.0 0.0—5.30.30.2205005000.00.o L p > L[0.1l|?.l.lel-f.olllnl0ll|¢.lrlbyll0|"[ 0.0 r > > ’ r b y p h p > > . p p [ 0.0 . . DiuiOlqulDImT/O/O: ltuUlCl:UI/I/l O/G n n I [I . m I l/ 00— m I. - ... S S w .... 0.00 ,/ 1 \O A A 15- m... ,-_...-..--.-.\ m m «N '1- o. .---- o.00 m 0.00 m n-N II?! N N ii-D I! 0.0... 0.0« < >00..021005 000005... 455.2. 149 0v05 5005 0.05 550 000 000 000 0.0 .55 055 000 000 000 00v VNV 050 000 050 .00 00— 005 0:0 000 050 500 #00 000 #00 005 0V5 005 500 0V0 550 00v 05v 000 000 000 00— 00— 005 00. MVN 0vv0 50¢0 0500 0500 0050 0000 5005 5005 050— 000— 0005 000— 0055 000 050 505 000 050 000 050 00. W0 5.05 000— Nev— 0005 5v0. 005— .00. 000 000 005 000 500 0V0 00¢ 00v 000 5V0 v00 0Q0 50— 005 T0 OFNM¢WQNQGO v—I—v-v-v-v—I—v-v—F'N OFNNVLOCDI‘CDO) 0cm 500 050 550 000 005 005 0.5 550 0.0 .00 000 00v 000 #00 050 000 05. 505 00 00 0:0 000 050 500 #00 005 #05 000 0V0 000 500 0v0 55¢ 000 050 000 005 005 00 00 00 00 mV0 0v00 5000 0500 0550 0000 000— .005 505— 050. 00v. 000— 00.— 000— 000 055 500 000 05¢ 000 055 00 W0 0 55¢. movp 0V0. 0005 5¢—5 000— 500 000 005 0v0 000 50v 5V¢ 000 000 000 FVN ¢05 0v— 50 00 70 00500v005000 v-v-v-v-v-v-I—r-v-v—N 050000050 0v05 5005 0505 550 .00 .00 000 050 005 055 000 000 000 00v v~¢ 050 000 050 .00 00— 00— V0 000 050 500 #00 000 V00 005 0¢5 005 500 0V0 550 00v 05¢ 000 000 000 00. 005 00. 00. 00 0vv0 50V0 0500 0500 00.0 0000 5005 500. 0505 000— 000— 0005 005— 000 050 505 000 0.0 000 050 005 N0 5505 0005 0va 0005 5V05 005— —005 000 000 005 000 500 0V0 00v vmv 000 5V0 v00 0VN 505 00— 70 PNMVLOCDNQGO v-v—v-v-r-t-v—v-v-N Or-NMQLDCDINQGO ,. 0V0 500 0.0 550 500 505 005 055 000 0.0 000 00v 00v 000 #00 050 000 055 .0— 00 00 0:0 0)... 00002.00 05 5.3 50.00000 00.00803... 050.0800 00.0 0.00..- 000 050 500 VNO 005 «05 000 0v0 000 500 0V0 55v 000 050 000 00. 005 00 00 00 00 MYO 0v00 5000 0500 0550 0000 0005 500— 5055 050. 00v— 000— 005. 000— 000 055 500 000 0.? 000 055 00 90 55v— 00v5 0V0. 0005 5v—5 0005 .00 000 005 0V0 000 50v 5v¢ 000 000 000 PVN v05 0v. 50 00 T0 v-NMfiI-OCDINQOO v-v-v— N O CFNNVLOCOINGO, PFPF F'- 00.o..0 0.00.8.0 30.05.00... 3002.. .00.... 150 000 000 000 050 000 500 0.0 000 00. 05. .0. .0. .#. .0. 00. 50. 00 00 .0 00 0. #-0 .00 050 000 0#0 000 0.0 000 050 .00 000 0 .0 50. 05. 00. .#. 0.. 00 05 #0 00 0. 0-0 00. 00. 00 #0 00 #0 05 05 05 00 .0 50 00 0# 0# .# 00 00 00 0. 0. #.0 ..0 500 000 50. .0. 00. 05. 05. 00. 50. 0#. 0#. 00. 00. 00. 00 00 .0 00 o. 00-00#005000 v-t—I—v-r-v-v-v-r-r-N Ot—NCOVLDQNQ 000 000 000 050 000 500 0.0 000 00. 05. .0. .0. .#. .0. 00. 5o. 00 00 .0 00 0. #-0 .00 050 000 0#0 000 0.0 000 050 .00 000 0.0 50. 05. 00. .#. 0. . 00 05 #0 00 0. 0-0 00. 00. 00 #0 00 #0 05 05 05 00 .0 50 00 0# 0# .# 00 00 00 0. 0. 0-0 0 #.0 ..0 500 000 50. .0. 00. 05. 05. 00. 50. 0#. 0#. 00. 00. 00. 00 00 .0 00 0. .-0 00«-00#005000 PFv-v-I—v—v—I-v-r-N O—NMVLDCOBG) 000 000 000 000 0#0 500 000 000 050 000 .00 .#0 .00 .00 0.0 50. 00. 00. 0#. 00. 00. #-0 .5# 00# 00# 00# 0.# 00# 000 000 .#0 000 000 500 000 000 .00 0.0 00. 00. 0#. #0. 00. 0-0 #0. .0. 00. 00. 55. 05. 00. 00. 00. #0. 00. 0#. .#. 50. 00. 50. 00. 0.. ... 00. 00. 0-0 #00 .00 500 000 500 .00 050 000 000 000 5#0 000 000 000 0.0 50. 00. 00. 0#. 00. 00. ..0 00-00#005000 v-v-v-v-q—v-v—v—I—c—N OFNC‘OVLOCOINQ 000 000 000 000 0#0 500 000 000 050 000 .00 .#0 .00 .00 0.0 50. 00. 00. 0#. 00. 00. #-0 Our. .5233. o... 5.3 32.0.5.0 $20300 3.3.3620 00.0 0.03. .5# 00# 00# 00# 0 .# 00# 000 000 .#0 000 000 500 000 000 .00 0.0 00. 00. 0#. #0. 00. 0-0 #0. .0. 00. 00. 55. 05. 00. 00. 00. #0. 00. 0#. .#. 50. 00. 50. 00. 0.. ... 00. 00. 0-0 #00 .00 500 000 500 .00 050 000 000 000 5#0 000 000 000 0.0 50. 00. 00. 0#. 00. 00. ..0 00-00#005000 PI—v-r-F-v-v-v-v-v-N OFNMQLDCDINQ 3.96 0.03950 .6225: $32.“. .3.... UH. .5333. 05 5.3 3.2230 :o..an.=< 9.8.2820 5.0 230.”. 151 m<£5 00 0. 0. 5. 0. 0. #. 0. 0. .. 0. 0 0 5 0 0 T 4 4 f 0(55 O0 0. 0. 5. 0. 0. #. 0. 0. .. o. 0 I .t‘ \ 1:11.. A y o | .11 0 >00402I005 000000.“. 45.5.2. v 0 ‘ O 0.0 o aninGt . ..0 III. 512 # 0 0 2 l ior! 3 [ill 000 000. 000. 0000 0000 000 000. 000. 0000 0000 lndlnCDdOEflJNn .Lfldan :IO smm < >00402100... 00000.... ..<_..._2_ mgs. ON a. m. n. o. m. e. m. a. .F o. m m n o m e n N o cow 0 m I. ooo. s O .1 0 com. m 2 u E. ---- . .\C\. 0.0 I. \aa\aq .1:4u::. ooou U\.,r~\ ..0 .III oil DxIC\ comm o >00..oz:005 000000.. 455.2. Tn“- I n-lnHttl|l .llllnuillu l l 0‘ I 1.. It: 1 HI.” 1| irl ll. mgsp ow m. o. .— m. m. v. m. ~. .. o. m m n m m e n N .l‘ -11'IIY. ‘I$.l-.ou||..l.-lf .ll Tlol|+ll viii. 0 00m HDWOR n «unannmuuouuuunonuunum\x m I. \\-\\-\\- ooo. s I\l m \.\ o uuxn oom. n a fiH\ 0.0 IIIIO'I. m ¢~|IlmTII 0000 ..0 II' oOmN 152 017. 08.33. 2.. .23 3:53.07. @8230 3:33:50 2.0 2%.“. m5... 92.5 ON 0— m— 5.. 0— 0. V. 0.. N— Z 0— m m 5 0 m v n N — 0 0a m— 0— 5. 0. 0— Q, m. N— .... 0— 0 m 5 0 0 v n N — 0 > f C Ill Fr 0 \Dxto‘iu‘MWVmVa 20 QiO‘LulleLuquTfi E-i¢\0\\o\0\\01 O‘U‘E‘hv: \-Q\O.\u OO— Q‘Ot‘O‘iTxO \ \ GDP \0. \ n \b\\¢\\o\o n \o\¢\¢\m\§\ N \o o\\ N .\ Ho \ .\ I n. Ma \ U \-\--\--\-\a. b.\. o N m. Lani..- ML. .\. oom s T. ..--- \b\o . .\ o o I[-...:._ \o\ .\.\ 0 \._Q \o\ .1 ‘0\\0 ‘0- .... Q\°\Q \§ A \\AV\ .\».\ M o o\ \.\. 08 w ?-o\o \.\¢ oom m 0\ nl §\ \\ §\ .1!’\\!. m .‘IQ\§\ m 00v 00v YNIIIT nfi lotl -IIDI TNI'IIIH— v.~|¢|| n.~ Il¢| «.mllT Fault-III. 000 000 0 5004021005 00w00mm .5#—z. o >00..02:Uw... mwmoOmn. ..<_._._z. ME... m5: o~2822£39~_:o.mmnmmemN.o 822:2m.22§:9mmnomenN.o r b y b! o b I . o 00— m m ... O \ H. s a-¥O\c\o\o\ \ c s 02 m. ix 02 o a: M l‘l‘litl‘ “a“ 0\0 M 000 k‘ \sluilxi \\ 000 W I I O\§ m m Q\O\O\O\O\O\\o\\ox w 8.. o!‘ \.\.\ oov Q‘\\ 000 000 m >0040210m5 00m001n. ..<_....2_ < >0040210wh 00m00mm ...<:..2. 153 0 mm mm mv 0v mm 0v 0v mm mm .0 .0 mm mm mm Om 5v 3‘ we mm 0 mm mm .3 0v 0m 0v m¢ mm mm .0 .0 mm 00 mm Om 5v 3 NV mm #-N 0 mm mm mfi .v 00 .¢ 0... 00 50 00 00 mm 00 N0 0v 5.» vv NV 0? 0 mm mm 5‘ .¢ mm .v 0? 00 50 00 00 mm mm mm Om 5v 3 we 0.» m. (DOV-NMVLDCDINQ r-r-v-v-v—v-r-r-v- Ov-NCOVLDCDNQ 0 0 MN mm mm mm «.3 me. 0v 0¢ vv 3» .v .v 5. 5. mm mm N0 N0 50 50 00 00 05 05 V5 #5 .5 .5 00 00 0v 0.. 0. 0. 0 0 0 0 O 0 m-w 00A. .5333. o... 5.3 3.2230 5.8.53. ..o 5:52.00... 5N6 05“.; 0 mm mm me. 0v 3 ..V 5. mm mm 50 00 05 v5 .5 00 0V 0. mo—wmvwco5oocn u—I-v-v—r—v—I—v-v-v— OPNMVLDCDINGJ 00 00 00. 0.. .N. VN. mm. 00. 0#. mm. N0. Om. 0N. .N. 0.. 0.. mm. QN. ON. 0 00 mm 00. 0.. 8. cm. mm. mm. 0.... mm. mm. 00. 0N. .N. 0.. 0.. mm. VN. ON. 0 N-N 00 mm 00. 0.. .N. ew. mm. mm. 0#. mm. mm. om. 0N. .N. 0.. 0.. MN. ¢N. ON. 0 00 mm 00. 0.. .N. QN. mm. mm. 0?. mm. am. On. 0N. .N. 0.. 0.. mm. #N. ON. 6. m. m. 5. mOf-NCOVLOCD v—v-r-v-f-I—I— OPNCOV'LOCDNQ m 00 N0 00. 00. 00. mo. .0. 00 00 mo 5m #0 .m 0v 0.. 5v me me 5v 0 00 N0 00. 00. 00. 00. .0. 00 00 mo 50 v... .m we 0.» 5e 0v 0.» 5v .-N m 00 N0 00. 00. 00. mo. .0. 00 00 mm 50 V0 .m 0.. 0? 5V 0v mv 5v 0 mm N0 00. 00. 00. m0. .0. 00 00 no 5m #0 .0 0v 0#. 5v 0v 0v 5v 0. 0. 5. OOFNNVLOCO r-v—v—v—v—I—v- OFNC‘OVLDQONQ 30.05.00. 380:. 3...... 3.9.0 c.0228 154 v m. .N NN MN NN NN ON 0. v. M. M. N. v. 5. NN 0N NN ..N 5N c m. .N NN MN NN NN ON 0. .... M. M. N. ... 5. NN 0N NN ..N 5N V m. .N NN MN NN NN ON 0. v. M. M. N. v. 5. .N 0N NN VN 0N .V m. .N NN MN NN NN ON 0. v. M. M. N. v. 5. .N 0N NN VN MN m. GOPNMVLDCOINCD v—v-v-c-v-I—v—v—r- O—NM¢mm5m v 0. NN .VN mN mN 0N QN MN NN NN MN MN .VN 5N 0N 0N VN 0N 0N .V v 0. 0. NN NN VN ..N mN 0N mN mN 0N mN .VN ¢N MN MN NN NN NN NN VN MN MN MN vN VN 5N 5N 0N 0N 0N 0N .vN MN 0N 0N 0N 0M 0 MN 0.. 8233. 2.. 5.3 3.3.0.230 5:633... ¢ 0. NN VN 0N 0N 0N ..N MN NN NN MN MN .VN 5N 0N 0N MN 0N 0M 00-NM¢00500 v-I-r-v-v-v-v-r-v-v— OFNMVLDQDBQ 5000500000000000000¢ 5mmm5mmmmmmmwwcocococom¢ wmd 053.. N-N I’NCDCONINCDCDLOLOLOLOCOCOCOCDCOCOCDLOVP [\(DCDNINOCDLDLDLDLDCDCDQCDCOOCDLOV ODOPNMQ’LOCDINQO') v-u—I-v-v-r-v-v—v-I— Ov-NCOVLOCDfiw 00055500-3 QCDBNNCOLDV O) .-N QQNNNCDLD? mao555com0Z 0.: ......) ..050:.0.3<..0 Comummmswo< 0.6 0....w_n— m5... m2... 0. M. 5. 0. m. v. M. N. .. o. M M 5 o 0 v M N . 0 M. 0. 5. o. m. c. M. N. .. 0. M 5 0 n v M . 0 2-11.]. .I o m/ , » ll {Tibia o o\o\.u , e \ ,{\\O\\O 00 n D/U D 0\ 0m 0 N N w “@110 a». w 0 O I 00. .1. O IAVIII 00. M m n U, 0 ¢ nu M n of 1 m- .7! on. I. (It-'1 cow ooN «1N m0.010 0_Owh O on 0 0m 0 m m 1. ... S S m m a L 8. m FZL. 2. m I1O/ \ D.\ ’0 .°|.°\\v .ror’g. \\g\°\g & M .0\0 n of ... om. 1 oo~ ooN N-N 00.010 0.0m... > . o o u A v i 1 0.0 u p . . o p L p L . L . r p . . > > “V 0.0 0".0! 0 ..II m 0.0. m m 0.9 w w IT 19\ o w <--i-i-- . M ... a 0.8 A // \ .§ - / \§A/ . , . 0.0N ....I._ 0 . -.0. - // .l ,Q/{Qx . u .4 3.8 m C .‘ 7 ‘ v 0.0 . 4 o I 0 o 4 o o y ' ‘ o o 4 IA . 0.0 Q/dT/O!D|Q.|n'oi:c||: Q @1103le / E/D/D/Q é.“ . . . n 00. m 00. m n. 1 S S m m . I! a] 0.8 GILT! 08 M o a ., M 0 m 0 II‘Il} - .I/ m "|!.\V"I. \ 0.8 m T--¢! .02 m ma‘-.0||l (19'... (Ill!!! III 0.0.. 0.0? N-N 00.010 0.005(050 ..N 00.010 0.00...0N_ 0.: .._._>> \AzowawU :0_.0.:0....< 03.00—35.30 _N.© 0....w_5— mi: 05.: 8 m. 0. 5. m. 0. 3 n. N. .. o. m 0 5 0 0 v n N . 0 ON m. m. 5. 0. 0. v. n. N. .. o. m m 5 m 0 ¢ n N . o - { , . \u 0 v r - _\ In! H.» 0 lemumxa LxVTTmA ' \ \I\ \\0\0 .\"\\A'U\\D\\o \\.\o. \‘I\ Q\ oom \I..\'\° oom \\I\I\Q\O\ n \,l‘-l\.lHO\o n ..\..\..\\0\\0 N ..\.. 0030.-.... N . |l\l . \1 I 1 ;I \l\ lt\0\0\ I I I'll. \0‘0 II. I . I‘D|\\° 1 0-0 0:0. 000. 5 0x00 000. s O O J: I: m m 000. 000. n n 1 1 , _ 000w . ; 000m 0|I¢ll ululoi! alfill (lutl. o'ol ulbll alas! (lllll 0000 003 ¢.N 00.010 0.0053950 MN 00.010 05055.50 us: 0.2.. 000.220.230.329.“ 0 n 0 0 v m u . 0 0N0.0.D0.m.32~.:0.0 0 n 0 m c n N . 0 v r > b > ..I'V > ¥ 0 p > I b r u..||o r ? r\‘-\. o l\IH.IuHM\O olAYlI I\ID\\\IH\I\ 000 n ..\\IWmHC\ n 0 ----.Il m \nflb\-o m ..\a 000. u. ..\\m\0 000. s I\\5Iu\0 .1 \Im .1 < c- . ..\o\ o . l. -m\9 o i I .._w\o\ 000. 010 . 000. m \me00. n ..1 1 ... ui 000m L 0000 n\ \ all oullol allOlll (lull \n\o\o ul-_.\0\ N.N 00.010 0.005(050 .-N 00.010 0.0055,...0 160 000 8223. 20 a? 0:590: 052% 202205 00.0 2%: ..v N.N 00.010 0.005OOJOZIUwh 000000.; 43.52. 05:5 0—0—550—0F¢FOFN—:OP00500¢0N..0 . . o . L4 “ l o A» 0 l o o 0 u « « I» y 0.0 MV/ - . n Tall/o oom m l 1 u/A/ S O WNTSWDx cow 3 1 \CwCluC m . - \ .\C I. ...“CWC\ m .l O\\ -. 0.00 l m; Ilol N; -l El 3 lull 0.00 0 >OOJOZIUmk mmmUOmm ._ 9 o 0 o 2 o a o 0 . 0 v 0.0 l . n 00N W l S ...J 0 00¢ .: O n l. m 0.00 I. fimélol N-Fl---:ll Tplll l- .000 U >OOJOZIUmh 00000.: 43.52. 22: 3828239329 m m h m m c m N _ o f» 0 L‘ .l » .1 v 0 “ll. ¢ 0 l. .1 0 ¢ TlLIlla 0.0 if D/ I”? D/C/ LY 0.0« n D/ m l. S \ A3 . o . l . cow 5 n l m 0.00 1. m4lol- .2 0.8 < >OOAOZIUmk 000000“; 23.52. llllll.\ 169 SN u 00.213 0‘: 22502 05 :23 3:508?” 00 0222000< 0N0 800:0 0.2.5 0.2: 0P0—5—0~0~¢~m—NF:0—0050m¢mNFO 0—0—5F0pm—¢wmpN—:0—00500¢0N—0 lel-IIIIIIIIIIII ImTwTEI$I I . 0.0 pHIllllllllllIll IEIEIQI 0.0 DIDIDIDIDID 0 - 0.0 m l 0.0 m I I. 2 m... S O O 0.9 a 0.0. a A A v V B N I 3 I I I -- I --l I l I fill -ll 0.? m 0.3 M fin; lol «4 lDIl Z llll‘ 0; lol 3 l,D.l E lll .r O.ON .2 0.0N 0 >OO._OZIUmH 000000.”. 1.42:2. U >OOJOZIUwh 000000.". ._<:._Z_ 0.2:. NEE. 0—0F5P0F0P¢F0FNFZOF00500¢0NP0 0FOF5F0—mP¢—MPN9:0900500¢0N—0 pHnllllIll-IIHGVEIEI§I$I$I§I$I§IEISIEI$ 0.0 .H-lllllnlllll IEIQIBIEIalemleQIwTfi 0.0 D\O\U\D\D\U\C o n 0.0 m t 00 m m a O 0 0.0— M 0.0— M w m . --llll-.-lIl m 3 0.3 1 ‘04: M. 0; lol N; lDlli T— lll A 0; llol «4 lDlI T— lll LY O.ON 2 0.0N 0 >OOAOZIUwh mmmUOmm 4.3.52. < >OO._OZIUm._. 000000m ASH—Z. 170 emu No Is DIE 20::02 0:0 :23 2200000 020050210N 02223200 5N0 0.5022 0.2:. 0NOFOP5p0—0—¢—0PNF :0_. 0 0 5 0 0 ¢ 0 N F 0 .ll OllTI. . L r 4 » . ... v 2- 1W 0.000 I: II\ E II 2 2 22 F2. .. E.. E E ..-.E LI L.II.I 3 . o I 0.000. S llDlI :llll 2- 0.000. 0 >800sz» 0800:: 05:2. 22; gem—2333023 :0, 0 0 5 0 0 v m m _ 0 I. .I o .l » v 0 v 2 .I 0 0 0 0 o . 0.0 . 0.000 Ill _ 2 3 I .oII- 0.009 2 Bl ... Ill .2 0.00..“— 0 5002021002. 000000m 242:2. 1.0:!an :JO 8.1.an .Lfidan :JO $1.an E < >002OZIUmH mmmUOmm 25.2.22. 22: 0~0_0:—0_m;—0.~_:0_00n0000a 0 Tll¢lrl|+l.lll1l.l.ll¢ . > w I o .lI+.l? OO. O n . N IQ\Q\§\§ 00.000 u Q. S . I ISI. 0 IIII 3I3 : Z.) 3 . w m Silo II 00009 m NI. lDl I. T— lIl 2, 00.000; 0 5000253 000005 ._OO2OZIQwh 00000002 .2<;_z_ QIIIIWWIIGNBI- 0.09 m U S . 0. 000m A V U L 0.000 M 0. .IO... 2 I C IIIIIIIIl .2 0.00« 0 5000218: 000005 .322. 0.2:. 0090—2023329 :0, 0 0 n 0 0 v m N F 0 .r l 0 o o I. l. v u o 0 o o o 0 0 0 . o l 0.0 QMGHBHEIIBIQIQIQIEIEI alt- EIQIEIiIIiIQIEIaIJ! 0.09 m m . m 0000 A V U m . I-IIIIII--III: --. I -ll 0.000 M figIlioll «..lIICIIII :llllI L . 0.00". 172 000N em I 00.213 0: 3502 05 :23 awe—00:00.2. 000005 .20 022230 0N0 0000202 lllllIl I I ll 00:202.? 000~ 000 . 000 . 000 0 . 9\Ol.|l\l.0\o o DIIIIOIIlOI-ull oo— 00N 000 all ulIlI mlél -vrt»]e-rt, (al-u M(t) = - 5H/5m = 7&2(t)[¢1f(t) - 492] - [205(Q(t) - m(t)) - 2620100) - M(t))le'“, (211-2) l3'(t) = - 6H/5b1 = K3(t)0ta(t) + p(t)e-rt, (a1-3) M0) = - 5W5b2 = ' M(t)l35'1(t)(91f(t) - sz(t)) + p(t)“, (211-4) X5'(t) = — 5H/5k = 0, (al-S) K6'(t) = - arr/5x = o, (a1-6) M(t) = - arr/6s = M(tmbzromrfrt) - pza(t))5'2(t)- (a1-7) APPENDIX II 180 02.... mam... Emmfi mam... 03m... 03E 03m... 03m... mam... 03E. 03m... 03E mam... 8.. Q0000..- or 0 00 NOTE 0.0 Rim. v.0 (...mm 08.0 (In...< 86.0 HE $60 :In. 8F 00 om 8 Q0 8 QM 5 07m: >> 9+0? I 0.0 .v. 8? m2 9. ”.2 8m .0.). ON mm. or E ON 00 8? N0 00 6 830cm> 368$ rgmmd v05fi9 50558.9. 8503.0? N5059N0 $5 859.50 F§0 gwd v8N.m.. 558.9 $0.8 $5.9 Bfifivv 305.0". 05.8 §.8 0w gNF gdr ..550N50 850.80 00000.08 @5000 0905.80 008.85 850.85 5008.005 $.95 m5vwm. .00 0.08. 500 5..N50.0N.V 008?. NM 0990K 5805.59 9.85va 8300.... 930.8? $58.. §.8 n. OOOOOOOOOOOOOOOOOOOO 0: OOOOOOOOOOOOOOOOOOOO LL 05850.0 80850.0 $§.m 50 {00. C 00005.}; 080.20.. 088.8 03,30. 5 0580.00 maven 0588.00 8.00530 5%. V0 850.? any”? $003.3 $0.? Q505Q. 3. 9.80.9 $.00 < 000500003 0. .0002 00 208.3%. :< S. 58. N 000.08. R 80050.8 085.8 0808.8 85.8 8545.8 {090.8 058. ..N 0..?v0...0m 058.. F.0N 0538.00 mwgmw 0w 55.50 5QdN8 53.00 §.B 05.8 0. 5 N05 OP .2.. 80999 005N090 000968 858.90 5N05N. r 5 §.80 8580.8 :5N. ..00 B50085 $0.35 FN505.050 $000 #3809 905.08 §FOVN 8.094.505 :80.va 809.0: 83 .50.. EN: Br >.. 53.8.. 05.8? 0050.50? 808.8? §5Q¥ 50.50.02. .008. w: 5005.05 9550.02. g5“; 5000.8? $.8N 885% 0.3.5? 008. 3% 00.59 0.500 RM 8.. ON? Br .2} aflg 0050.085 058.5505 3.0055 0F50.S5 590N055 9.8.905 «$5ng $.8N5 8.45% 88.850 NF5N.§U v~8058 §.§N 080.08? 00GB: 00.8? r mgr 08.. gr >> OOPNWVIDDNCDQg Y—V'FY-V—T‘FFT‘V" Ox-NMVIDIONQ *— 181 NF+HN50 8.730 P5§N50 Nw+§0 0..+m0500 8§ § Ne+w§0 03.0800 vamg NF+§0 mtgw «$085 NF+§5 Nw+§5 mimmmfi «$090.0 mv+m=mm «$058.». «709% Nigw «$950.? «$083 «$08: ~P+wmde mimmbe S+wmmmm C¢S 3.7996 3.0.30 3.003.... 805% :+wm$.v $05 :+va $050.85 rw+wSFm @005 3mg...” $.95 5+“va 05.0098 rigm 30va0 530030 308 240590 80.5w ww+mv8€ 0.03.9 ..Immfiv 585.500 3.030...» 83.590 rr+w§¢ $568 5+wm5wé {EQQ 5+9» gflN S+m58.m 8? “.¢.. 0.2.x 350580 3.5800 5.0 mvwvfid g0 gvd g0 gd 9 F830 @8500 555.0 3580 $050.0 $050 $350 Q5550 9050.00 g0 V500v80 grad em OOOOOOOOOOOOOOOOOOOOO 8 0 E500 £0500 50am 555.: 35v: @009 859 0330.5 Egan mwflmuém 0558.90 80530 $60 85.0.9 30.? gm? $0.9. gm...“ 998.9 gflu X0 0 Ed E0500 5085.0 55$. 5 £053; @005 gm 0530.5 05908 05530 RES 80508 $60 850.0”V g5». $3.9 $0.0». 0505N5.:V 998.9 $00 vfi. OOOOOOOOOOOOOOOOOOOOO g 8550. $36. $50. gmd. 59080. 3680. go. 830. g? 533$. $0. 885.? 8N8? 9305.0. .5900. $80. 55%.? 05.8.0? ngW S50..- $4.820. gwd. 805020. $20. 858.0. 0583.0. 53rd. 85:0. $0. 3%”? E5580. $5.. 9050..- 850.? $3.5. 00590 0500 000 0.9 .28 @8000 $50 $0 05.8.0 grdw ego? BFfidr 58.8.0? 5§.R gwkev Ed. 5550: @585 §.8 955805 $50.50 g3. QENNVN 05590.: £030.». §N >8 8 §§§§§§§§§§§§BBBBBBBBB BBBBBBBBNNNNRRNNR 50. 5.0. 5.0. 50. Ed. 5Q. 50. 5.0. 8.0. ...0. NO. 00. v0. .. 0. .00. ..00 5.0 00 50 00 N0 .00 5.0 50 5.0 5.0 50 5.0 8.0 8.0 8.0 w .0 N0 00 0.0 0.0 0.0 00 N0 v.0 3.0 8.0 (DOT—NWVIDCONQQR Fr-v-v-w-v-v-FFF OV-NOOVLDONCD *- 182 gr. Nth 8N8“? QB? 32‘8”. wawz‘m- 05051 v5.55 90.9KB 38.38 055mm“ 0 $068 mumavvv 9.36m 88t§ QMBCN $.90 fig BNQO gem? RHZMRN 82.5% gamgm QQHUFNG NngNw 85% 36.5 $.98? 8393—. NQQMN 8.VNDVN $8me ggw §N.5 gag 9mg $8.95 hvfig gflfi 98% Rag v§.§ §.§N 85.3% gnaw mEHNEN 52mg Raw—0.80 wag 3.9..va 90% CNNRZ. 85$ mmmwwnvr @gw tug gnaw gdvm 5.8? R363? 9% gNg game N—umvwfim g5 Ema—8 Egg E? 8.83: vmmEbflN wenmv§ $.55 gx or.mm:.r mfim Sugar gag Ngd 03386 §.8 Swarm? $.85 gN GYMNNV NmKBNU www.cgw mun—09$. memmme NTmeM gig mBfimFNw BESS. gm 07$?“ gNN @58er RmMmdmN (gm; Nvaofi $5.3 mfg: N§NN $656 ammummuN E34. w$immw o o 078va Egmmm @8va 352% $3,959.26an E590 gage O OOOOOOOOOOOOOOOOOOOOO $87 39% 0 BF $523. 8_. ESQBN 8F FEMNBF BF mNva—h 8F 8mg 8? £285 8F gamw 8F nag 8P avg 8—. 03.9% 8P mgr 8_. Kg 8F gag 8F RWRNS Br 858% 8r £9963 8F Saga 8F mag 8F SQBBL 8w ENE 8.. $30 m 9593.9 85.2% mQHddE B mXNFNQ. 98:5 8 vaudfi vmvvotb B NSKEK $68 8 gmg «Swag 8 gag gnaw B @NK8 BVNNB B mag QRN.Bm B QNB SE B erlen. F§.B B 98.9%. Qumg 8 NENQ ago 8 mKBQV 0E8.@ 8 g.§ gflm B ggm GERMBV B ENQQ 55.3 B 8vmmdwm 859mg 8 mayo? gfimu B {KNRF KENS 8 $8 gm“ 8 8 RN 8 x x mm §NHN 9mg é: N939 $8.8m 8 grad; egg gdflm 0.. g: 839 §.§ 9 5.3 gov §.g t FAR—.99 $5.9 gmwvm 9 82.3.9 55.: 858mm 9 Ear Ever 55% 3 KKQVGF 5N5: BSVQw 9 E55. $8.: Fgémm Nw Ear grmNMr gate“. 5 ENNN gov v—aNB or gag gm BKNFB m Emma gun?” BBQ—NV m MNKBMM vgflv mag n XMEFWU grew. NEOQN m 38.9 Bug vg.8_. m gum m58 mega; v $58 98 $.03 m nag or 9%wa N 89.8 NF gduw F 8v or 8? o 5 2m >w ._. BIBLIOGRAPHY 183 BIBLIOGRAPHY Abernathy, W.J., "Production Process Structure and Technological Change," Decision Sciences, Vol.7, 1976, pp.607-619. Abernathy, W.J., "Pattern of Industrial Innovation," Technology Review, 80(7), June-July, 1978a. Abernathy, W.J., The Productivity Dilemma, Baltimore:Johns Hopkins University Press, 1978b. Abernathy, W.J., K.B.Clark, and A.M.Kantrow, Industrial Renaissance: Producing a Competitive Future for America, Basic Books, New York, NY, 1983. Adam BE, and PM. Swamidass, "Assessing Operations Management from a Strategic Perspective," Journal ofManagement, 15(2), 1989, pp.181-203. Adler, P.S., "Managing Flexible Automation," California Management Review, Spring 1988, pp.34-56. Amit, R. and Y. Ilan, "The Choice of Manufacturing Technology in the Presence of Dynamic Demand and Experience Effects," IIE Transactions, 22(2), June 1990, pp. 100-1 1 1. Atkinson, J ., "Flexibility: Planning for an Uncertain Future," Manpower Policy and Practice, 1, Summer 1985. Ayres, R.U. and S. Miller, "Robotics, CAM, and Industrial Productivity," National Productivity Review, Winter 1981-1982, pp.42-60. Ayres, R. and Wilbur A. Steger, "Rejuvenating the Life Cycle Concept," The Journal of Business Strategy, 1983, pp.63-76. Baranson, 1., Robots in Mamifacturing-Key to International Competitiveness, Lomond Publications Inc., Mt. Airy, MD, 1983. 184 Bass, F.M., "The Relationship between diffusion rates, experience curve, and demand elasticities for consumer durable technological innovations," Journal of Business, 1980, 53(3), 551-367. Bergstrom, G.L. and BE. Smith, "Multi-item Production Planning - An Extension of the HMMS Rules," Management Science, 16(10), June 1970, pp.B614-B629. Bobrowski, PM. and VA. Mabert, "Altemate Routing Strategies in Batch Manufacturing: An Evaluation," Decision Sciences, 19, 1988, pp.713-733. Boothroyd, G., "Economics of Assembly Systems," Journal of Manufacturing Systems, 1(1), 1982, pp.111-127. Browne, J ., D. Dubois, K. Rathmill, S.P. Sethi, and KB. Stecke, "Classification of Flexible Manufacturing Systems," The EMS Magazine, April 1984. Buffa, E. 8., "Research in Operations Management," Journal of Operations Management, 1(1), 1980, pp.1-8. Buffa, E.S., Meeting the Competitive Challenge, Dow J ones-Irwin: Homewood, 11.,1984. Burstein, M.C., "Finding the Economic Mix of Rigid and Flexible Automation for Manufacturing Systems," Proceedings of the Second ORSA/TIIVIS Conference on Flexible Manufacturing Systems, Ann Arbor, MI, 1986, pp.43-54. Butler, J E "Theories of Technological Innovation as Useful Tools for Corporate Strategy," Strategic A/Ianagement Journal, 9, 1988, pp.15-29. Buzacott, J .A., "The fundamental principles of flexibility in manufacturing systems," Proceedings of the First International Conference on FMS, Brighton, 1982, pp. 13 -22. Buzacott, J.A. and Mandelbaum, M., "Flexibility and Productivity in Manufacturing Systems," Proceedings of the [IE Conference at Chicago, IE and Management Press, Atlanta, CA, 1985, pp.404-413. Buzacott,J.A.and D.D.Yao,"Flexible Manufacturing Systems:A Review of Analytical Models,"Management Science, 32(7), 1986,890-905. Buzzell, RD. and FD. Wiersema, "Modeling Changes in Market Share: A Cross- Sectional Analysis," Strategic Management Journal, 1981, 2, pp.27-42. Chang, RH. and CM. Jones, "Production and Workforce Scheduling Extensions," AIIE Transactions, 2(4), December 1970, pp.326-333. 185 Chatterjee, A, M. Cohen, W. Maxwell, and L. Miller, "Manufacturing Flexibility: Models and Measurements, " Proceedings of [st ORSA/ T IMS Special Interest Conference on FMS, Ann Arbor, MI, 1984, pp.49—64. Chew, W.B., "No-Nonsense Guide to Measuring Productivity," Harvard Business Review, January-February 1988, 110-118. Chung, CH. and 1.]. Chen, "A Systematic Assessment of the Value of Flexibility for an FMS," Proceedings of the Third ORSA/TIMS Conference on Flexible Manufacturing Systems, 1989, pp.27-32. Cohen, Morris A. and Robert M. Halperin, "Optimal Technology Choice in a Dynamic— Stochastic Environment," Journal of Operations Management, 6(3), May 1986. Cohen,M.A. and H.L.Lee, "Manufacturing Strategy: Concepts and Methods," The Management of Productivity and Technology in Manufacturing, Plenum Publishing Co, 1985. Cox, Taylor,Jr., "Toward the Measurement of Manufacturing Flexibility," Production and Inventory Management Journal, 1st Qtr, 1989. DeMeyer,A.,and others, "Flexibility:The Next Competitive Battle-The Manufacturing Futures Survey," Strategic Management Journal, Vol. 10, 1989, pp. 13 5-144. Devinney,T.M., "Entry and Learning," Management Science,33(6), 1987, pp.706-724. Dhalla, N.K. and S. Yuspeh, "Forget the Product Life Cycle Conceptl," Harvard Business Review, January-February 1976, pp.102—112. Drucker,P.F., "What We Can Learn from Japanese Management,"Harvard Business Review, March-April 1971, pp. 1 10-122. Ettlie, J E WP. Bridges, and RD. O‘Keefe, "Organization Strategy and Structural Differences for Radical Versus Incremental Innovation," Management Science, 30(6), 1984, pp.682-695. Ettlie IE. and SA. Reifeis, "Integrating Design and Manufacturing to Deploy Advanced Manufacturing Technology," Interfaces, 17(6), November-December 1987, pp.63- 74. Falkner, C.H., "Flexibility in Manufacturing Plants, " Proceedings of 2nd ORSA/ T IMS Special Interest Conference on FMS: 0R Models and Applications, Ann Arbor, MI, 1986, pp.95-106. 186 Farley, J.U., B. Kahn, D.R. Lehmann, and W.L. Moore, "Modeling the Choice to Automate, " Sloan Management Review, Winter 1987, pp.5-15. Fine,C.H.and AC. Hax, "Manufacturing Strategy:A Methodology and an Illustration," Interfaces, 15(6), 1985, pp.28-46. Fine, C .H. and RM. F reund, "Economic Analysis of Product-Flexible Manufacturing Systems Investment Decisions, " Proceedings of the Second ORSA/TIMS Conference on FMSs, Ann Arbor, MI, 1986, pp.55-68. Fine, CH. and RM. Freund, "Optimal Investment In Product-Flexible Manufacturing Capacity," Management Science, 36(4), April 1990, pp.449-466. F rohman, A.L., ”Technology as a Competitive Weapon,"Harvard Business Review January-February 1982, pp.97-104. Frohman, A.L., "Putting Technology Into Strategic Planning," California Management Review, 27(2), Winter 1985, pp.48-59. Gaimon, C., "The Acquisition of Automation Subject to Diminishing Returns," [IE Transactions, June 1985, pp.147-156. Gaimon, C ., "The Optimal Acquisition of Automation to Enhance the Productivity of Labor,"Management Science, 31(9), September 1985, pp.1175-1190. Gaimon, C ., "The Strategic Decision to Acquire Flexible Technology," Proceedings of the Second ORSA/TIMS Conference on FMSs, Ann Arbor, MI, 1986, pp.69-81. Gaimon, C., "Dynamic Game Results of the Acquisition of New Technology," Operations Research, 37(3), May-June 1989, pp.410-425. Gaimon, C., "Incremental Capacity Expansion: The Effect of Lead Time and Learning," Working Paper, Georgia Institute of Technology, April 1992. Garrett, S.E., "Strategy First: A Case in FMS Justification," Proceedings of the 2nd ORSA/TIMS Conference on FMSs, Ann Arbor, MI, 1986, pp. 17-30. Gerwin,D.,"An Agenda For Research on the Flexibility of Manufacturing Processes," International Journal of Operations and Production Management, 7(1), 1987, pp. 3 8-49. Gold, 8., "CAM:New roles for production," Harvard Business Review, 60(6), 1982, pp.88-94. 187 Goldhar,J.D.and M.Jelinek,"Plan for Economies of Scale,"Harvard Business Review, November-December 1983, pp.141-148. Groover,M.H., Automation, Production Systems, and C omputer—A ided Manufacturing, Prentice-Hall, Englwood Cliffs, New Jersey, 1980. Gupta, Y.P. and S. Goyal, "Flexibility of manufacturing systems: Concepts and measurements," European Journal of Operational Research, 43, 1989, pp. 1 19- 135. Gupta, Y.P. and S. Goyal, "Flexibility Trade-offs in a Random Flexible Manufacturing System: A Simulation Study," International Journal of Production Research, 30(3), 1992, pp.527-557. Hambrick, D.C., I.C. MacMillan, and DC. Day, "Strategic Attributes and Performance in the ECG Matrix-PIMS Based Analysis in Industrial Based Product Businesses," Academy of Management Journal, 25, 19823, pp.510-531. Hambrick, et al, "The Product Portfolio and Profitability - A PIMS Based Analysis of Industrial Product Businesses," Academy of Management Journal, 25, 1982b, pp.733-755. Hayes,R,H., "Why Japanese Factories Work," Harvard Business Review, July-August 1981, pp.57-66. Hayes,R.H.and K.B. Clark, "Explaining Observed Productivity Differentials Between Plantszlmplications for Operations Research,"Inteifaces, 15(6), 1985, pp.3-14. Hayes,R.H. and R. Jaikumar, "Manufacturing‘s Cn'sis: New Technologies, Obsolete Organizations, " Harvard Business Review, September-October 1988, pp.77-85 Hayes,R.H. and S.G. Wheelwright, "Link Manufacturing Process and Product Life Cycle," Harvard Business Review, J anuary-February 1979a, pp. 133-140. Hayes,R.H. and S.G. Wheelwright, "The Dynamics of Process—Product Life Cycles," Harvard Business Review, March-April 1979b, pp. 127-136. Hayes, RH. and S.G.Wheelwright, Restoring Our Competitive Edge: Competing through Manufacturing, John Wiley&Sons;New York, NY, 1984. Hernot, Scott K, Daniel Levinthal, and James G.March, "Learning from Experience in Organizations," American Economic Review, 75(2), 1984, pp.298-302 188 Hill, Charles, "Differentiation Versus Low Cost or Differentiation and Low Cost: A Contingency Framework," Academy of Management Review, 13(3), 1988, pp.401- 412. Hill, Terry, Manufacturing Strategy: Text and Cases, Irwin, 1989. Hofer, CW. and D. Schendel, Strategy Formulation: Analytical Concepts, West Publishing Co., St.Paul, MN, 1978. Holt, C .C ., F. Modigliani, J .F . Muth, and HA. Simon, Planning Production, Inventories, and Work Force, Prentice-Hall, Englewood Cliffs, NI, 1960. Hough, Paul G., "Cost Estimating For The Factory of The Future, " Proceedings of the Third ORSA/ T IMS Conference on Flexible Manufacturing Systems: Operations Research Models and Applications, 1989. Jaikumar,R., "Postindustrial Manufacturing,"Harvard Business Review, November - December 1986, pp.69-76. J elinek M. and JD. Goldhar, "The Interface between Strategy and Manufacturing Technology," Columbia Journal of World Business, Spring 1983, pp.26-36. Jelinek M. and ID. Goldhar, "The Strategic Implications of the Factory of the Future," Sloan Management Review, Summer 1984, pp.29-37. Kaplan,R.S., "Measuring manufacturing Performance: A New Challenge for Managerial Accounting Research, "The Accounting Review, No.4, 1983a, pp.686-707. Kaplan, R.S., "Must CIM be justified by faith alone?" Harvard Business Review, 58(4), 1986, pp.87-95. Kantrow,A.M., "The Strategy-Technology Connection,"Harvard Business Review, July-August 1980, pp.6—21. Karmarkar, U.S.and S.Kekre, "Manufacturing Configuration and Mix Flexibility,"working paper, University of Rochester, 1985. Kegg, R.L., "Flexible Manufacturing Trends and Data Base Requirements, " Proceedings of the 5th International Conference on Production Engineering, ISPE, Tokyo, 1984. Kim, J .8., "Manufacturing Strategy and the Environment: An Analytic Study," Unpublished Dissertation, The Ohio State University, 1989. 189 Klahorst, H. T., "Flexible manufacturing systems: Combining elements to lower costs and add flexibility, " Industrial Engineering, 31(11), 1981, 112-117. Kotabe, M. and J .Y. Murray, "Linking Product and Process Innovations and Modes of International Sourcing in Global Competition: A Case of Foreign Multinational Firms, " Journal of International Business Studies, 3rd Quarter 1990, pp.383-408. Kraj ewski,L.J. and L.p.Ritzman,Operations Management: Strategy and Analysis, Addison Wesley; Readings, Massachusetts, 1987. Kumar, V., "On Measurement of Flexibility in Flexible Manufacturing Systems: An Information-Theoretic Approach, " Proceedings of the 2nd ORSA/TIMS Special Conference on FMS, Ann Arbor, August 1986, pp.131—143. Lavington, F ., The English Capital Market, Methuen, London, UK, 1921. Levy, F.K., "Adaptation in the Production Process," Management Science, 11(6), April 1965, pp.B136-B154. Li, S.L. and Devanath T., "Technology Choice and Capacity Expansion with Two Product FamilieszTradeoffs between Scale and Scope, " International Journal of Production Research, 30(4), 1992, pp.887-907. Lim, S.H., "Flexible Manufacturing Systems and Manufacturing Flexibility in the United Kingdom," International Journal of Operations and Production Management, 7(6), 1987, pp.44-5. Madu, CN. and NC. Georgantzas, "Strategic Thrust of Manufacturing Automation Decisions: A Conceptual Framework," IIE Transaction, 23(2), 1991, pp. 138-148. Mamer,J.W.and K.F.McCardle, "Uncertainty,Competition,and the Adoption of New Technology,"Management Science, 33(2), 1987, pp.161-177. Meredith, J .R., "The Strategic Advantages of the Factory of the Future," California Management Review, 29(3), Spring 1987, pp.27-41. Meredith, J .R. and J. Camm, "Modelling Synergy and Learning Under Multiple Advanced Manufacturing Technologies," Decision Sciences, 20, 1989, pp.258—271. Meredith, JR. and M.M. Hill, "Justifying New Manufacturing Systems: A Managerial Approach," Sloan Management Review, Summer 1987, pp.49-61. Meredith, J .R. and Suresh, N.C., "Justification Techniques for Advanced Manufacturing Techniques," International Journal of Production Research, 24(5), 1986, pp. 1043-1057. 190 Meyer,M.H.and E.B.Roberts, "New Product Strategy in Small Technology-Based Firm,"Management Science, 32(7), 1986, pp.806-821. Milgrom, P. and John Roberts, "The Economics of Modern Manufacturing: Technology, Strategy, and Organization," American Economic Review, 80(3), 1990, pp.511- 528. Miller,J.G.and A.V.Roth, "Manufacturing StrategieszExecutive Summary of the 1988 North American Manufacturing Future Survey,"Boston University Manufacturing Roundtable Report, 1988. Miller, S.M., "Industrial Robotics and Flexible Manufacturing Systems," The Management of Productivity and Technology in Manufacturing, edited by PR. Kleindorfer, Plenum Press, NY. and London, 1985, pp.9-56. Monohan,G.E.and T.L. Smunt, "Optimal Acquisition of Automated Flexible Manufacturing Processes," Operations Research, 37(2), 1989, pp.288-3 00. Muth, J.F., "Search Theory and The Manufacturing Progress Function," Management Science, 32(8), 1986, pp.948-962. Nagarur, N., "Some Performance Measure of Flexible Manufacturing Systems," International Journal of Production Research, 30(4), 1992, pp.799-809. Naik, B. and Chakravarty, A.K., "Strategic Acquisition of New Manufacturing Technology: A Review and Research F ramework," International Journal of Production Research, 30(7), 1992, pp.1575-1601. Pearson, J .N., HD. F eldman, and J .S. Bracker, "The Impact of the Product/Production Process Interaction on the Evolution of Strategic Response," Journal of Management Studies, 28(2), March 1991, pp. 161-172. Phillips, L.W., D. Chang, and RD. Buzzell, "Product quality, cost position and business performance: A test of some key hypotheses," Unpublished manuscript, Harvard University, Boston, 1983. Porter,M.E.,Competitive Strategy, Free Press, New York, NY, 1980. Preece, D.A., "Organizations, Flexibility and New Technology, " Managing Advanced Manufacturing Technology, IFS, London, UK, 1986, pp.367-3 82. Raffi, F. and PM. Swamidass, "Toward A Theory of Manufacturing Overhead Behavior: A Conceptual and Empirical Analysis," Journal of Operations Management, 7(1), Oct. 1987, pp.121-137. 191 Ramasesh, RV. and MD. J ayakumar, "Measurement of Manufacturing Flexibility: A Value Based Approach, " Journal of Operations Management, 10(4), October 1991, pp.446-468. Ranky, P., The Design and Operation of FMS, [FPS Ltd, North-Holland, 1983. Richardson,P.R.,A.J.Taylor,and J .R.M. Gordon, "A Strategic Approach to Evaluating Manufacturing Performance," 15(6), 1985, 15-27. Rosenberg, N., Inside the Black Box: Technology and Economics, Cambridge University Press, Cambridge, MA, 1982. Roth,A.V., "Strategic Planning for the Optimal Acquisition of Flexible Manufacturing Systems Technology," Unpublished Ph.D.dissertation, The Ohio State University, 1986. Roth,A.V., Gaimon Cheryl, and Lee Krajewski, "Optimal Acquisition of FMS Technology Subject to Technological Progress," Decision Sciences, 22, 1991, 308-334. Schonberger, RI, World Class Manufacturing, The Free Press, NY, 1986. Schroeder, D.M., "A Dynamic Perspective on the Impact of Process Innovation upon Competitive Strategies," Strategic Management Journal, 11, 1990, pp.25-41. Segev,E., "Strategy, Strategy Making,and Performance-An Empirical Investigation," Management Science, 33(2), 1987, pp.258-269. Sethi, AK. and SP. Sethi, "Flexibility in Manufacturing: A Survey," The international Journal of FA/ISs, 2, 1990, 289-328. Shipper, F., "The Production Process and the Technological Environment," Proceedings of the 41st Annual Meeting of the Academy of Management, San Diego, 1981, p.443. Simmonds, K., "The Fundamentals of Strategic Management Accounting," presented to the Institute of Cost and Management Accountants' Technical Symposium, Pembroke College, Oxford, January 6-8, 1981. Skinner,W., "The Focused Factory," Harvard Business Review, May-June 1974, pp. 1 14-121. Skinner,W., "Operations Technology: Blind Spot in Strategic Management," Interfaces, Vol.14, No.1, 1984, pp.167-176. 192 Slack,N., "Flexibility as a Manufacturing Objective," International Journal of Operations and Production Management, 2(3), 1983, pp.4-13. Smith, M.L., R. Ramesh, RA. Dudek, and EL. Blair, "Charateristics of US. Flexible Manufactuimg Systems - A Survey, " Proceedings of the Second ORSA/TIMS on FMSs, Ann Arbor, MI, 1986, pp.477-486. Smith, W.C., "Product Life Cycle Strategy: How to Stay on the Growth Curve," Management Review, January 1980, pp.8-13. Son, Y.K. and CS. Park, "Economic Measure of Productivity, Quality, and Flexibility in Advanced Manufacturing Systems," Journal of A/Ianufacturing Systems, 6(3), pp. 193 -206. Stalk, G., "Time-The Next Source of Competitive Advantage," Harvard Business Review July-August, 1988, pp.41-51. 3 Suresh, NC. and J .R. Meredith, "Achieving Factory Automation through Group Technology Pinciples," Journal of Operations Management, 5(2), 1985, pp.151- 182. Swamidass,P.M., "Manufacturing Strategyzlts Assessment and Practice,"Journal of Operations Management, 6(4), 1986, pp.471-484. Swamidass,P.M., Manufacturing Flexibility, Monograph No. 2, Operation Management Association, January, 1988. Swamidass,P.M.and W.T.Newell, "Manufacturing Strategy,Environmental Uncertainty and PerformancezA Path Analytical Model," Management Science, 29(4), 1987, pp.509-524. Thietart, RA. and R. Vivas, "An Empirical Investigation of Success Strategies for Business Along the Product Life Cycle," Management Science, 30(12), December 1984, pp.1405-1423. Urban,G.L.,and others, "Market Share Rewards to Pioneering Brands: An Empirical Analyses and Strategic Implications, " Management Science, 32(6), 1986, pp.645-659. Vickson, R.G., "Optimal Conversion to a New Production Technique Under Learning," IIE Transactions, June 1985, pp.175-181. Vidale, ML. and HE. Wolfe, "An Operations Research Study of Sales Response to Advertising," Operations Research, 5, 1957, pp.370-381. 193 Wasson,C.R, Dynamic Competitive Strategy and Product Life Cycle, Austin Press, Austin, TX, 1978. Wheelwright, S.C., "Reflecting Corporate Strategy in Manufacturing Decisions," Business Horizons, February 1978, pp.57-66. Wheelwright, S.C., "Japan-Where Operations Really Are Strategic," Harvard Business Review, July-August 1981, pp.67-74. Wheelwright, S.C., "Manufacturing StrategyzDefining the Missing Link," Strategic Management Journal, Vol.5, 1984, pp.77-91. Wheelwright, S.C.and RH. Hayes, "Competing through Manufacturing," Harvard Business Review, January-February 1985, pp.99-109. White, RE, "Generic business stategies, organizational context and performance: An empirical investigation," Strategic Management Journal, 7, 1986, pp.217-231. Wild, W.G. and 0. Port, "This Video Game is Saving Manufacturing Millions," Business Week, August 17, 1987, pp.82-84. Wright, T.P., "Factors Affecting the Cost of Airplanes," Journal of Aero. Science, 3, 1936, pp.122-128. Yelle, L., "The Leaning Curve-Historical Review and Comprehensive Survey," Decision Sciences, 10, 1979, pp.302-324. Zelenovic,D.M., "Flexibility - a Condition for Effective Production Systems," International Journal of Production Research, 20(4), 1982, pp.319-337.