Title Page
Abstract
Contents
Chapter 1. Introduction 17
1.1. Background 17
1.2. Problem Statement 32
1.3. Research Objective 34
1.4. Research Questions 35
1.5. Research Outline 36
Chapter 2. Logistics Transportation Network to Maximize Transportation Utilization and Efficiencies 40
2.1. Introduction 40
2.2. Related Literature 42
2.3. Methodology 44
2.4. Results 50
2.4.1. Scenario I: The Initial Route 51
2.4.2. Scenario II: The Route of Combine Trucks 58
2.4.3. Scenario III: The Route of Electric Vehicles 69
2.4.4. Scenario IV: The Route of Drone 78
2.4.5. The Comparison Between The Modified Saving Algorithm, The Nearest Insert, and The Farthest Insert Algorithm 85
2.5. Validation 106
2.6. Discussion and Conclusion 111
Chapter 3. Artificial Neural Network of Hub and Spoke Network for Logistics Efficiencies 117
3.1. Introduction 117
3.2. Related Literature 119
3.3. Methodology 128
3.4. Results 131
3.5. Validation 142
3.6. Discussion and Conclusion 147
Chapter 4. Reinforcement Modelling for Logistics Network Automation 151
4.1. Introduction 151
4.2. Related Literature 152
4.3. Methodology 159
4.4. Results 163
4.5. Validation 170
4.6. Discussion and Conclusion 174
Chapter 5. Conclusion 177
5.1. Summary and Contribution 177
5.2. Limitations and Future Research 180
Bibliography 182
초록 206
Table 1.1. The LPI score of Indonesia in 2018 21
Table 2.1. The comparison of the truck, electric vehicle, and drone 41
Table 2.2. The comparison of vehicle routing problem approaches 43
Table 2.3. The geographical locations of MPC and MPO 50
Table 2.4. The specification of gran max blind van 52
Table 2.5. The initial route for one shift in one day 53
Table 2.6. The vehicle utilization 54
Table 2.7. The cost of the initial route for one shift in one day (in IDR) 54
Table 2.8. The fuel use of vehicles 55
Table 2.9. The initial route for three shifts in one day 56
Table 2.10. The distance matrix 60
Table 2.11. The saving matrix 60
Table 2.12. The route of combine trucks for one shift in one day 64
Table 2.13. The vehicle utilization 64
Table 2.14. The cost of combine truck route for one shift in one day (in IDR) 65
Table 2.15. The fuel use of vehicles 65
Table 2.16. The route of combine trucks for three shifts in one day 66
Table 2.17. The Efficiency of Combine Truck Route and Initial Route 67
Table 2.18. The specifications of e-Canter 70
Table 2.19. The distance matrix 71
Table 2.20. The saving matrix 72
Table 2.21. The route of electric vehicles for one shift in one day 73
Table 2.22. The vehicle utilization 73
Table 2.23. The cost of electric vehicle routes for one shift in one day (in IDR) 74
Table 2.24. The charge use of electric vehicles 75
Table 2.25. The route of electric vehicles for three shifts in one day 75
Table 2.26. The Efficiency of Electric Vehicle Route and Initial Route 76
Table 2.27. The route of drones for one time delivery in one day 80
Table 2.28. The cost of drone route for one time delivery in one day (in IDR) 80
Table 2.29. The fuel use of vehicles 81
Table 2.30. The route of drones for one shift in one day 81
Table 2.31. The vehicle utilization 82
Table 2.32. The route of drones for three shifts in one day 82
Table 2.33. The Efficiency of Drone Route and Initial Route 83
Table 2.34. The route of combined trucks using nearest index for one shift in one day 87
Table 2.35. The vehicle utilization 87
Table 2.36. The cost of combine route using nearest index for one shift in one day 88
Table 2.37. The fuel use of vehicles 88
Table 2.38. The route of combined trucks using nearest index for three shifts in one day 89
Table 2.39. The route of electric vehicles using nearest index for one shift in one day 92
Table 2.40. The vehicle utilization 93
Table 2.41. The cost of electric vehicle using nearest index for one shift in one day 93
Table 2.42. The charge use of electric vehicles 94
Table 2.43. The route of electric vehicle using nearest index for three shifts in one day 94
Table 2.44. The route of combined trucks using farthest index for one shift in one day 97
Table 2.45. The vehicle utilization 98
Table 2.46. The cost of combine route using farthest index for one shift in one day 98
Table 2.47. The fuel use of vehicles 99
Table 2.48. The route of combine truck using farthest for three shifts in one day 99
Table 2.49. The route of electric vehicles using farthest index for one shift in one day 102
Table 2.50. The vehicle utilization 103
Table 2.51. The cost of electric vehicle using farthest index for one shift in one day 103
Table 2.52. The charge use of electric vehicles 104
Table 2.53. The route of electric vehicles using farthest index for three shifts in one day 104
Table 2.54. The distance matrix 109
Table 2.55. The saving matrix 109
Table 2.56. The route of the combine truck 109
Table 2.57. The vehicle utilization 110
Table 2.58. The Efficiency of Initial Route, Route of Combine Truck, Route of Electric Vehicle, and Route of Drone 112
Table 2.59. The efficiency comparison results in combined truck 113
Table 2.60. The efficiency comparison results in electric vehicle 114
Table 3.1. The comparison of traditional and machine learning forecasting 118
Table 3.2. The combine route in January 2019 132
Table 3.3. The data normalization 133
Table 3.4. The data in January 2019 138
Table 3.5. The data normalization 139
Table 4.1. The multidimensional deep reinforcement learning 164
Table 4.2. The data on distance, time, cost, and utilization 166
Table 4.3. The data normalization 166
Table 4.4. The data on distance, time, cost, and utilization 167
Table 4.5. The data normalization 167
Table 4.6. The data on distance, time, cost, and utilization 171
Table 5.1. The summary of the initial route, route of combine truck, route of electric vehicle, and route of drone 177
Figure 1.1. Market landscape 20
Figure 1.2. The autonomous systems and intelligence technology 26
Figure 1.3. Research outline 37
Figure 2.1. The Mail Processing (MPC) and Main Post Office (MPO) in Jakarta. 45
Figure 2.2. The Modified Saving Algorithm 49
Figure 2.3. The initial route 52
Figure 2.4. The grand max blind van 52
Figure 2.5. The initial route simulation system 57
Figure 2.6. The initial route statistical chart simulation system 57
Figure 2.7. The Colt Diesel Engkel 58
Figure 2.8. The first route of combine truck 61
Figure 2.9. The second route of combine truck 61
Figure 2.10. The third route of combine truck 62
Figure 2.11. The fourth route of combine truck 62
Figure 2.12. The fifth route of combine truck 63
Figure 2.13. The combine route simulation system 68
Figure 2.14. The combine route statistical chart simulation system 69
Figure 2.15. The e-Canter 69
Figure 2.16. The electric vehicle routes simulation system 77
Figure 2.17. The electric vehicle routes statistical chart simulation system 77
Figure 2.18. The drone cargo 79
Figure 2.19. The drone route simulation system 84
Figure 2.20. The drone route statistical chart simulation system 84
Figure 2.21. The combine route nearest index simulation system 90
Figure 2.22. The combine route nearest index statistical chart simulation system 90
Figure 2.23. The electric vehicle route nearest index simulation system 95
Figure 2.24. The electric vehicle route nearest index statistical chart simulation system 95
Figure 2.25. The combine route farthest index simulation system 100
Figure 2.26. The combine route farthest index statistical chart simulation system 100
Figure 2.27. The electric vehicle route farthest index simulation system 105
Figure 2.28. The electric vehicle route farthest index statistical chart simulation system 105
Figure 2.29. The Fuso Truck 107
Figure 2.30. The combine route simulation system 110
Figure 2.31. The combine route statistical chart simulation system 111
Figure 3.1. Human Neural Networks and ANN 119
Figure 3.2. The artificial neural network 129
Figure 3.3. Artificial Neural Network 130
Figure 3.4. The architecture of the artificial neural network 131
Figure 3.5. The artificial neural network combined route 134
Figure 3.6. The artificial neural network combined route state 135
Figure 3.7. The artificial neural network combined route coefficient correlation 135
Figure 3.8. The result plot of neural network 136
Figure 3.9. The architecture of the artificial neural network 137
Figure 3.10. The performance of artificial neural network 139
Figure 3.11. The training state of artificial neural network 140
Figure 3.12. The artificial neural network coefficient correlation 141
Figure 3.13. The result plot of artificial neural network 142
Figure 3.14. The artificial neural network combined route 143
Figure 3.15. The artificial neural network combined route state 144
Figure 3.16. The artificial neural network combined route coefficient correlation 144
Figure 3.17. The The performance of artificial neural network 145
Figure 3.18. The training state of artificial neural network 146
Figure 3.19. The artificial neural network coefficient correlation 147
Figure 4.1. Reinforcement learning 153
Figure 4.2. Deep reinforcement learning 157
Figure 4.3. Deep Q Learning 161
Figure 4.4. The deep reinforcement learning 164
Figure 4.5. The reinforcement learning simulation system 168
Figure 4.6. The reinforcement learning simulation system 168
Figure 4.7. The reinforcement learning statistical chart simulation system 169
Figure 4.8. The reinforcement learning simulation system 169
Figure 4.9. The reinforcement learning simulation system 171
Figure 4.10. The reinforcement learning simulation system 172
Figure 4.11. The reinforcement learning statistical chart simulation system 172
Figure 4.12. The reinforcement learning simulation system 173