Supply chain performance improvement: the role of information technology

SUPPLY CHAIN PERFORMANCE IMPROVEMENT:

THE ROLE OF INFORMATION TECHNOLOGY

By Bibhushan

Mechanical Engineering Department

Submitted in fulfillment of requirements of the degree of DOCTOR OF PHILOSOPHY

To the

INDIAN INSTITUTE OF TECHNOLOGY DELHI

FEBRUARY 2008

C cm: I V\ , 7r yy,

I. I. T. DELHI.

unr./

Acc. No 711-3(60----

CERTIFICATE

This is to certify that the thesis entitled 'Supply Chain Performance Improvement:

The Role of Information Technology' submitted by Bibhushan to the Indian Institute of Technology, Delhi, India, for the award of degree of Doctor of Philosophy, is a record of bonafide research work carried out by him under our supervision and guidance. The results obtained in thesis have not been submitted to any other University or Institute for the award of any degree or diploma.

a#1

Dr. Subhash Wadhwa Dr. • noop Chawla

Professor Professor

Department of Mechanical Engineering Department of Mechanical Engineering Indian Institute of Technology Delhi Indian Institute of Technology Delhi

New Delhi- 110016 New Delhi- 110016

ACKNOWLEDGEMENTS

This research work involved extensive time and efforts from my side. However, this research would not have been possible without valuable guidance, support and encouragement provided by my supervisors. I take this opportunity to express my sincere thanks to them. My last six and half years at Indian Institute of Technology Delhi have been a great learning experience. Even before the start of this research, I have been immensely benefited by the experienced, dedicated and dynamic faculty in the Industrial Engineering Group of Mechanical Engineering Department. Their devotion to industry relevant research is contagious. This devotion motivated me to choose research as a career five years ago.

First and foremost, I would like to express my gratitude to Professor Subhash Wadhwa and Professor Anoop Chawla, my Ph.D. supervisors at Indian Institute of Technology Delhi. Professor Wadhwa initiated me into the multiple entity flow view of the world and the associated simulation modeling concepts, which not only formed a backbone of this thesis but also brought about a paradigm shift in my thinking. His dedication to research along with his skill to visualize the value preposition in research output and his diligence for innovative leads in research have been the points of motivation that I admire the most. His ideas on simulation of manufacturing system as entity flows with time, cost and value attributes formed the backbone of this research. I am indebted for his priceless supervision that enabled me to conclude my research. Moreover, his domain knowledge and expertise, kindness and love for the students are but a few attributes that I would like to emulate. I will forever cherish my close association with him.

iii

I would like to express my sincere thanks to all my professors viz Professor S. Wadhwa, my supervisor, Professor Anoop Chawla, my co-supervisor, Professor S.G. Deshmukh, my Student Review Committee (SRC) chairman, and Professor D.K. Banwet and Mr.

A.D. Gupta, members of my SRC who have made this research a possibility. I am indebted to all the SRC members for providing direction and encouragement to initiate this research and their continuous feedback and guidance to complete the same. I am also thankful to Professor Arun Kanda, Head of Department at Mechanical Engineering Department, IIT Delhi and Professor M.P. Gupta from Department of Management Studies, IIT Delhi for providing support, guidance, direction and encouragement in times of need. I am also thankful to the 1-IRD project led by Prof S. Wadhwa and Prof.

S.G. Deshmukh that provided us with the research facilities in the CIM lab. Last, but not the least, I am very grateful to Mr. M.K. Bhatnagar of IE Lab, who was always more than happy to understand and fulfill all the requirements needed for this research.

Bi hushan Dated:

iv

I dedicate this research to

T'arents

Who compromised on their needs but never compromised on my wants

And

My Gurus

Who have always been a source of inspiration for me

ABSTRACT

This thesis aims to enrich some insights into inventory based performance improvement in supply chains by optimization of inventory policy parameters and judicious selection of information technology (IT) for information sharing (IS) across the supply chain.

The overall theme of this research is to enhance and enrich the research on inventory management and information sharing within the supply chain. This research considers inventory policy parameters like expected service quality (ESQ), ordering constraints and capacity constraints, and IT based IS as means to improve the cost based performance of different inventory control policies.

This research is motivated by contemporary research on simulation based modelling and analysis of numerous problems within the broader domain of supply chain. This literature is collated, enhanced and synthesized to develop an object-oriented framework for modelling supply chains with decision flexibility. The synthesis is based on a multiple entity flow perspective emphasizing the concepts of action points and decision points that control the flow of entities in desirable directions. The proposed conceptual framework is used to develop a simulation based supply chain modelling environment, which can model multiple decisions in the supply chain like inventory management, sourcing, transport selection and production planning. Subsequently, simulation models are developed using this modelling environment to analyze the performance of four different inventory control policies.

The studies in this research have two major objectives. First, they aim to optimize the 'inventory policy parameters for each node in the supply chain. Second, they also seek to determine the optimal IS level for each node in the supply chain. While fulfilling

vii

these two objectives, these studies also compare the performance of four different inventory control policies. The studies in this research have been carried out with three different levels of variability within the supply chain. In the first level, impulse demand is considered to represent the controlled variations in the demand. At this level, the performance of different nodes in the supply chain is compared along multiple performance metrics for four different inventory policies. At the second level, the demand with constant coefficient of variation (COV), i.e. Poisson process, is used to induce variability in the demand. In the studies with this level, three different parameters of inventory typical policies: ESQ, minimum order quantity (MnOQ) and (MxOQ) have been optimized for all nodes in the supply chain. In addition, optimal IS level is also determined for each node in the supply chain. The effect of changes in the COV is considered at the third level. At this level, only the optimal. IS level is determined for each node, since COV is an uncontrollable variable for the supply chain and cannot be optimized.

This research work highlights some important findings for inventory management in supply chains. Firstly, the findings suggest that mathematically less complicated inventory policies can perform better than the more sophisticated policies, just by optimally seleting its parameters. Secondly, the research also highlights that IS can either improve or deteriorate the performance of supply chain, depending on the level of IS and inventory policy used. Hence, there is a need to select the level of IS judiciously. Another important finding is that different inventory policies may perform better for different types and of variability in the supply chain. Finally, the results also show that for most of the inventory policies concerned, the optimal IS level remains same for different COV levels.

viii

TABLE OF CONTENTS

Certificate

Acknowledgements ⁱⁱⁱ

Abstract vii

Table of Contents ix

List of Figures xix

List of Tables xxiii

List of Abbreviations xxvii

CHAPTER 1. INTRODUCTION 1

1.1 RESEARCH CONTEXT AND MOTIVATION 1

1.1.1 Simulation for Supply Chain Modeling and Analysis 2

1.1.2 Object-Oriented Simulation Modelling 3

1.1.3 Multiple Entity Flow Perspective 5

1.1.4 Inventory Management for Supply Chain Performance Improvement 6

1.2 RESEARCH OBJECTIVES AND METHODOLOGY 7

1.3 ORGANIZATION OF THE THESIS 11

CHAPTER 2. LITERATURE REVIEW 15

2.1 INTRODUCTION 15

2.2 SUPPLY CHAIN SIMULATION MODELLING AND ANALYSIS 16 2.2.1 Application of Simulation to Strategic Problems in Supply Chain 17 2.2.2 Application of Simulation to Operational Problems 20 2.2.2.1 Application of Simulation to Planning in Supply Chains 20 2.2.2.2 Application of Simulation for Inventory Management in SCs 21 2.2.2.3 Application of Simulation for Distribution in Supply Chains 22

ix

2.2.2.4 Application of Simulation for Sourcing in Supply Chains 24

2.2.3 Simulation Based Tools and Models 26

2.2.4 Agent Oriented Simulation Modelling 30

2.2.5 Simulation-based Hybrid Modelling 35

2.2.6 Application of Simulation for Model Validation 39

2.3 INVENTORY MANAGEMENT IN SUPPLY CHAINS 42

2.3.1 Classification of Inventory Policies 43

2.3.1.1 Optimization goal: Local Vs Global 43

2.3.1.2 Control Type: Decentralized Vs Centralized 43 2.3.1.3 Inventory Control: Continuous Versus Periodic 44 2.3.1.4 Temporal Information Requirements: Time Phased Versus

Instantaneous 45

2.3.1.5 Spatial Information Requirements: Installation Stock versus Echelon Stock 46

2.3.2 Coordination within Supply Chain 46

2.3.3 Vendor Managed Inventory (VMI) 49

2.3.4 Inventory Management under Uncertainty 50

2.3.5 Integration of Inventory Management with Other Functions 51 2.4 INCREASED VISIBILITY THROUGH INFORMATION SHARING (IS) 53

2.5 RESEARCH GAPS 57

2.6 CHAPTER SUMMARY AND KEY CONCLUSIONS 60

CHAPTER 3. AN OBJECT ORIENTED SIMULATION-BASED

FRAMEWORK FOR MODELLING SUPPLY CHAINS 63

3.1 INTRODUCTION 63

3.2 A GENERIC MODEL OF SUPPLY CHAIN 64

3.3 OBJECT ORIENTED SIMULATION MODELLING 68

x

3.4 MODELLING OF ELEMENTARY SUPPLY CHAIN CONSTRUCTS 73

3.4.1 Classification of Objects 73

3.4.2 Multiple Entity Flow Perspective 74

3.4.3 Action Points as Processes in the System 77

3.5 HIERARCHY OF OBJECT USED IN SUPPLY CHAIN MODELLING 80

3.5.1 Modeling of a Supply Chain Network 81

3.5.2 Modeling of Supply Chain Nodes 84

3.5.3 Modeling of Supply Chain Operations 86

3.5.4 Modeling the Manufacturing System 88

3.6 CHAPTER SUMMARY AND KEY CONCLUSIONS 89

CHAPTER 4. DEVELOPMENT OF SIMULATION MODELING

ENVIRONMENT AND DEMONSTRATION MODELS 91

4.1 INTRODUCTION 91

4.2 MODELLING THE SUPPLY CHAIN BUILDING BLOCKS 92

4.2.1 Modelling the Manufacturing System 93

4.2.2 Modelling the Transports 96

4.2.3 Modeling the Player Role 99

4.2.4 Modeling the Supply Chain Node 104

4.2.5 Modelling the Inter-Node Interactions 107

4.2.5.1 Defining Inter-Node Relationships 108

4.2.5.2 Defining Inter-Node Lead Times 108

4.2.5.3 Defining Inter-Node Speeds 109

4.2.5.4 Defining Inter-Node Distances 110

4.2.5.5 Defining Product Demands 111

4.3 MODELLING SUPPLY CHAIN DECISIONS 111

4.3.1 Source Selection Policies 112

xi

4.3.2 Inventory Control Decisions 116

4.3.3 Transportation Decisions 123

4.3.4 Production Planning Decisions 126

4.4 PERFORMANCE METRICS 130

4.4.1 Inventory Related Performance Metrics 130

4.4.2 Demand Related Performance Metrics 132

4.4.3 Service Related Performance Metrics 133

4.5 MODELLING THE SUPPLY CHAIN 134

4.6 MODEL VERIFICATION AND VALIDATION 136

4.7 CHAPTER SUMMARY 141

CHAPTER 5. EFFECT OF IMPULSE DEMAND VARIABLES ON THE

PERFORMANCE OF SUPPLY CHAIN 143

5.1 INTRODUCTION 143

5.2 EXPERIMENTAL SETUP 144

5.2.1 Demand Impulses 145

5.2.2 Simulation Parameters 147

5.2.3 Balancing the Policies 149

5.2.4 Metrics Considered for Performance Measurement 151 5.3 EFFECT OF TRANSFORMED RELATIVE IMPULSE AMPLITUDE (TRIA) ON THE

SUPPLY CHAIN 152

5.3.1 Effect of TRIA on the Supply Chain using Demand Flow Policy (DFP) 152 5.3.2 Effect of TRIA on the Supply Chain using Order Q Policy (OOP) 156 5.3.3 Effect of TRIA on the Supply Chain using (s, S) Policy (sSP) 157 5.3.4 Effect of TRIA on the Supply Chain using (s, Q) Policy (sQP) 160 5.4 EFFECT OF BALANCE GAP (BG) ON THE SUPPLY CHAIN 163 5.4.1 Effect of BG on the Supply Chain using DFP 163

xii

5.4.1.1 Effect of Negative Impulse BG (NIBG) on the SC using DFP 163 5.4.1.2 Effect of Positive Impulse BG (PIBG) on the SC using DFP 165 5.4.2 Effect of BG on the Supply Chain using OQP 166 5.4.2.1 Effect ofNIBG on the Supply Chain using OQP 167 5.4.2.1 Effect of PIBG on the Supply Chain using OQP 168 5.4.3 Effect of BG on the Supply Chain using sSP 169 5.4.3.1 Effect of NIBG on the Supply Chain using sSP 169 5.4.3.2 Effect of PIBG on the Supply Chain using sSP 171 5.4.4 Effect of BG on the Supply Chain using sQP 172 5.4.4.1 Effect of NIBG on the Supply Chain using sQP 172 5.4.4.2 Effect of PIBG on the Supply Chain using sQP 173 5.5 EFFECT OF NUMBER OF IMPULSES (NI) ON THE SUPPLY CHAIN 177 5.5.1 Effect of NI on the Supply Chain using DFP 177 5.5.1.1 Effect of Number of Negative Impulses (NNI) on the SC using DFP 177 5.5.1.2 Effect of Number of Positive Impulses (NPI) on the SC using DFP 179 5.5.2 Effect of NI on the Supply Chain using OQP 181 5.5.2.1 Effect of NNI on the Supply Chain using OQP 181 5.5.2.2 Effect of NPI on the Supply Chain using OQP 182 5.5.3 Effect of NI on the Supply Chain using sSP 183 5.5.3.1 Effect of1■INI on the Supply Chain using sSP 183 5.5.3.2 Effect of NPI on the Supply Chain using sSP 185 5.5.4 Effect of NI on the Supply Chain using sQP 186 5.5.4.1 Effect of NNI on the Supply Chain using sQP 186 5.5.4.2 Effect of NPI on the Supply Chain using sQP 187 5.6 EFFECT OF IMPULSE WIDTH (IW) ON THE SUPPLY CHAIN 190 5.6.1 Effect of IW on the Supply Chain using DFP 191

5.61.1 Effect ofNegative Impulse Width (NIW) on the SC using DFP 191 5.6.1.2 Effect of Positive Impulse width (PIW) on the SC using DFP 191 5.6.2 Effect of IW on the Supply Chain using OQP 193 5.6.2.1 Effect of NIW on the Supply Chain using OQP 193 5.6,2.2 Effect of PIW on the Supply Chain using OQP 194 5.6.3 Effect of IW on the Supply Chain using sSP 195 5.6.3.1 Effect of NIW on the Supply Chain using sSP 195 5.6.3.2 Effect of PIW on the Supply Chain using sSP 196 5.6.4 Effect of 1W on the Supply Chain using sQP 198 5.6.4.1 Effect of NIW on the Supply Chain using sQP 198 5.6.4.2 Effect of PIW on the Supply Chain using sQP 199

5.7 CHAPTER SUMMARY AND KEY CONCLUSIONS 202

CHAPTER 6. EFFECT OF EXPECTED SERVICE QUALITY (ESQ) ON

THE PERFORMANCE OF SUPPLY CHAIN 209

6.1 INTRODUCTION 209

6.2 EXPERIMENTAL SETUP 210

6.2.1 Accommodating Random Demand in Inventory Policies 211

6.2.2 Response Surface (RS) Design 212

6.2.3 Modelling Information Sharing (IS) within Supply Chain 218 6.2.4 Nonparametric Statistical Tests used in Research 219

6.2.4.1 Kruskall-Wallis Rank Sum Test 220

6.2.4.2 Mann-Whitney U Test 221

6.3 RS OPTIMIZATION FOR SUPPLY CHAIN WITH DFP 221

6.3.1 Effect of ESQ on the Supply Chain under DFP 223 6.3.2 RS Regression for TC Vs ESQ of Supply Chain Nodes under DFP 223 6.3.3 RS Optimization for TC Vs ESQ of Supply Chain Nodes under DFP 224

xiv

6.4 RS OPTIMIZATION FOR SUPPLY CHAIN WITH OQP 227 6.4.1 Effect of ESQ on the Supply Chain under OQP 227 6.4.2 RS Regression for TC Vs ESQ of Supply Chain Nodes under OQP 228 6.4.3 RS Optimization for TC Vs ESQ of Supply Chain Nodes under OQP 229

6.5 RS OPTIMIZATION FOR SUPPLY CHAIN WITH sSP 230

6.5.1 Effect of ESQ on the Supply Chain under sSP 231 6.5.2 RS Regression for TC Vs ESQ of Supply Chain Nodes under sSP 232 6.5.3 RS Optimization for TC Vs ESQ of Supply Chain Nodes under sSP 234

6.6 RS OPTIMIZATION FOR SUPPLY CHAIN WITH sQP 235

6.6.1 Effect of ESQ on the Supply Chain under sQP 235 6.6.2 RS Regression for TC Vs ESQ of Supply Chain Nodes under sQP 236 6.6.3 RS Optimization for TC Vs ESQ of Supply Chain Nodes under sQP 237 6.7 SELECTION OF OPTIMAL IS LEVEL FOR EACH NODE 238 6.7.1 Selection of Optimal IS level for Supply Chain with DFP 239 6.7.2 Selection of Optimal IS level for Supply Chain with sSP 241 6.7.3 Selection of Optimal IS level for Supply Chain with sQP 244

6.8 CHAPTER SUMMARY AND KEY CONCLUSIONS 247

CHAPTER 7. EFFECT OF ORDERING AND CAPACITY

CONSTRAINTS ON THE PERFORMANCE OF SUPPLY CHAIN 251

7.1 INTRODUCTION 251

7.2 EXPERIMENTAL SETUP 252

7.3 MNOQ SELECTION FOR EACH NODE OF SUPPLY CHAIN 254 7.3.1 MnOQ Selection for each Node of Supply Chain with DFP 255 7.3.1.1 Effect of MnOQ on the Supply Chain with DFP 255 7.3.1.2 RS Regression for TC Vs MnOQ of DFP-Based Supply Chain 255 7.3.1.3 RS Optimization for TC Vs MnOQ of DFP-Based Supply Chain 257

xv

7.3.2 MnOQ Selection for each Node of Supply Chain with OQP 258 7.3.2.1 Effect of MnOQ on the Supply Chain with OQP 258 7.3.2.2 1?S Regression for TC Vs MnOQ of OQP-Based Supply Chain 260 7.3.2.3 RS Optimization for TC Vs MnOQ of OQP-Based Supply Chain 260 7.3.3 RS Optimization for Supply Chain with sSP 262 7.3.3.1 Effect of MnOQ on the Supply Chain with sSP 262 7.3.3.2 RS Regression for TC Vs MnOQ of sSP-Based Supply 263 7.3.3.3 RS Optimization for TC Vs MnOQ of sSP-Based Supply Chain 264 7.3.4 RS Optimization for Supply Chain with sQP 265 7.3.4.1 Effect of MnOQ on the Supply Chain with sQP 265 7.3.4.2 RS Regression for TC Vs MnOQ of OQP-Based Supply Chain 267 7.3.4.3 RS Optimization for TC Vs MnOQ of sQP-Based Supply Chain 268 7.4 MXOQ SELECTION FOR EACH NODE THROUGH RS DESIGN 269 7.4.1 Optimal MxOQ Selection for Nodes of Supply Chain with DFP 269 7.4.1.1 Effect of MxOQ on the Supply Chain with DFP 270 7.4.1.2 RS Regression for TC Vs MxOQ of DFP-Based Supply Chain 271 7.4.1.3 RS Optimization for TC Vs MxOQ of DFP-Based Supply Chain 272 7.4.2 Optimal MxOQ Selection for Nodes of Supply Chain with sSP 273 7.4.2.1 Effect of MxOQ on the Supply Chain with sSP 273 7.4.2.2 RS Regression for TC Vs MxOQ of sSP-Based Supply Chain 274 7.4.2.3 RS Optimization for TC Vs MxOQ of sSP-Based Supply Chain 275 7.4.3 Optimal MxOQ Selection for Nodes of Supply Chain with sQP 277 7.4.3.1 Effect of MxOQ on the Supply Chain with sQP 277 7.4.3.2 RS Regression for TC Vs MxOQ of sQP-Based Supply Chain 278 7.4.3.3 RS Optimization for TC Vs MxOQ of sQP-Based Supply Chain 279 7.5 SELECTION OF OPTIMAL IS LEVEL FOR EACH NODE 281 xvi

7.5.1 Selection of Optimal IS level for Supply Chain with DFP 281 7.5.2 Selection of Optimal IS level for Supply Chain with sSP 284 7.5.3 Selection of Optimal IS level for Supply Chain with sQP 286

7.6 CHAPTER SUMMARY AND KEY CONCLUSIONS 290

CHAPTER 8. EFFECT OF COEFFICIENT OF VARIANCE (COV) ON

THE PERFORMANCE OF SUPPLY CHAIN 293

8.1 INTRODUCTION 293

8.2 PERFORMANCE OF DFP-BASED SUPPLY CHAIN UNDER COV 294 8.2.1 Performance of DFP-Based Supply Chain under Medium COV 295 8.2.2 Performance of DFP-Based Supply Chain under High COV 296 8.2.3 Performance of DFP-Based Supply Chain under Extreme COV 297 8.3 PERFORMANCE OF SSP-BASED SUPPLY CHAIN UNDER COV 298 8.3.1 Performance of sSP-Based Supply Chain under Medium COV 299 8.3.2 Performance of sSP-Based Supply Chain under High COV 300 8.3.3 Performance of sSP-Based Supply Chain under Extreme COV 302 8.4 PERFORMANCE OF SQP-BASED SUPPLY CHAIN UNDER COV 303 8.4.1 Performance of sQP-Based Supply Chain under Medium COV 304 8.4.2 Performance of sQP-Based Supply Chain under High COV 306 8.4.3 Performance of sQP-Based Supply Chain under Extreme COV 308

8.5 CHAPTER SUMMARY AND KEY CONCLUSIONS 309

8.5.1 Performance of Inventory Policies under Low COV 311 8.5.2 Performance of Inventory Policies under Medium COV 311 8.5.3 Performance of Inventory Policies under High COV 312 8.5.4 Performance of Inventory Policies under Extreme COV 313 CHAPTER 9. SUMMARY AND CONCLUSIONS 315

9.1 THESIS SUMMARY 315

xvii

9.2 KEY CONCLUSIONS 321

9.3 INDUSTRY IMPLICATIONS 329

9.4 SALIENT CONTRIBUTIONS 333

9.5 LIMITATIONS AND FUTURE DIRECTIONS 336

References 339

Brief Resume 359

BRIEF OVERVIEW 359

EXPERIENCE 359

EDUCATION 359

COMPUTER SKILLS 360

List of Publications 361

REFEREED JOURNALS 361

NATIONAL / INTERNATIONAL CONFERENCES 361

xviii