Using game theory & GIS to optimize supply chain management in Mumbai for multiple sectors

Abstract

Efficient goods transportation is essential for sustaining economic growth in Mumbai, especially within its congested and spatially complex urban logistics environment. This thesis aims to optimize supply chain management across the mid-mile by analysing transportation routes used by multiple logistics providers. The focus is on cost and route optimization through total freight consolidation, rather than targeting any specific sector. However, the companies from which data was obtained primarily operate in FMCG, temperature-controlled goods, medical supplies, and e-commerce. GIS is used to map existing freight routes and visualize the optimized routes generated through a Linear Programming (LP) model in Python. The study employs the Transportation Problem framework using Linear Programming (LP) in Python to optimize routes under different strategic scenarios, minimizing total transportation costs while satisfying real-world supply, demand, and capacity constraints. Five strategic interaction models are developed using Game Theory, each representing different configurations of collaboration and competition among logistics players. These include one fully competitive model where all three players compete, one fully collaborative model where all three collaborate, and three hybrid strategies where two players collaborate while the third competes. Building upon these models, a total of eighteen strategic scenarios are evaluated to explore varied cooperation-competition dynamics and their operational outcomes. Freight consolidation is operationalized by identifying overlapping optimized routes and evaluating infrastructure sharing potential across providers. Payoffs are calculated for each strategy across all eighteen scenarios based on costs, revenues, and operational performance. These are compiled into payoff matrices, and strategic decisions are evaluated using Nash Equilibrium, solved via Python. The resulting equilibrium strategies are visualized using GIS to produce actionable decision maps, highlighting optimized routes and shared usage under collaborative setups. Sensitivity analysis is employed to assess the impact of varying operational parameters and strategic configurations on outcomes, enhancing the robustness of the model. The most effective strategy across all evaluated scenarios involves partial collaboration between two logistics players while the third competes independently. This configuration results in cost savings of up to 28.4% and profit gains ranging from 28% to 36%, demonstrating the clear advantage of selective collaboration in high-density urban logistics. Notably, even the non-collaborating player benefits significantly, reinforcing that coordination among some players can yield substantial system-wide improvements. This integrated approach equips logistics providers with data-driven, spatially grounded tools for cost reduction, strategic coordination, and improved sustainability in urban freight systems. Keywords: Game Theory, GIS, Supply Chain Management, Supply Chain Optimization, Urban logistics, Mumbai, Freight Consolidation, Transportation, Decision-making.