*Alphabetical order

This study examines the optimal timing for truck fleet operators to adopt green technologies amidst high upfront costs and competitive market pressures. We employ a Mean Field Game (MFG) framework using discrete-time dynamic programming to model how individual transition decisions interact with the aggregate industry distribution. The model quantifies a "green-advantage" reward for early adopters who outpace the market average. We provide rigorous proof for the existence and uniqueness of a Mean Field Equilibrium using fixed-point theory and monotonicity conditions. Online mirror descent is implemented to computationally solve the problem. Numerical results reveal how dynamic investment upfront costs, market advantage, and length of planning horizon influence the strategic, forward-looking behavior of fragmented fleet operators.

Firms face increasing pressure to transition toward greener technologies, yet early investment often entails high costs, long lead times, and substantial uncertainty about future demand. This creates a dilemma: whether to invest early to secure market position or wait until technologies mature and market conditions become clearer. Green technologies offer no functional or operational differentiation beyond sustainability and can coexist with conventional options. We analyze this transition problem for two market structures under demand uncertainty and competitive interaction. In fragmented markets, a focal firm chooses its investment level while anticipating decentralized, uncertain investment by many heterogeneous competitors. In concentrated markets, two dominant firms make strategically interdependent decisions. Across both settings, firms trade off early commitment against the option value of waiting as technology costs and market acceptance evolve. We show that optimal early investment falls into a small number of distinct regimes, from postponing investment to aggressive overinvestment aimed at capturing competitors’ green customers. Early investment can be optimal even when initially unprofitable, driven by the long-term value of protecting and expanding market share. Smaller firms invest more aggressively relative to their size than larger firms, while dominant firms behave more cautiously due to their greater exposure to overinvestment. Demand uncertainty and overinvestment costs shape these incentives, disciplining aggressive behavior in equilibrium while amplifying firm asymmetries. Finally, overestimating green demand typically leads to lower regret than underestimating it, though collective overestimation can result in inefficient industry‑wide overinvestment.

Full manuscript available upon request via email

Fleet electrification is essential to decarbonizing transportation. However, commercial adoption, particularly in heavy-duty segments, remains limited and falls short of zero-emission targets. As the primary decision makers, fleet operators face complex trade-offs that place this transition squarely within the scope of operations management (OM). This review adopts an OM perspective to synthesize and classify the expanding literature on fleet electrification, drawing also from operations research (OR), transportation, and energy domains. The contributions of this review are threefold:

 (i) Literature synthesis: We identify key decision factors at the strategic, tactical, and operational levels, connecting them to relevant academic work and supporting them with recent data and practical examples.

 (ii) Practice-oriented insights: We extract lessons learned to inform decision making by fleet operators and related stakeholders.

 (iii) Research agenda: We propose open research questions across decision levels to guide future studies in the OM and OR communities.

Publication

Begun in the early 2000s, mean field games have been a widespread research interest bothfor the theoretical difficulties concerning stochastic control and numerical analysis and for the promising application in economics, finance, and machine learning. The objective of this thesis is two-folded: firstly, to demonstrate the usage of two mainstream tools for solving mean field games, namely, the stochastic approach based on Pontryagin’s maximum principle and the analytic approach based on the coupled system of Hamiltonian-Jacob- Bellman (HJB) equation and Kolmogorov-Fokker-Planck (KFP) equation; secondly, to showcase the mean field games’ application in nature, including emergent behaviors and resource exploitation. To achieve this goal, we study the flocking model of birds and take the stochastic approach to compute explicitly the equilibria for both finite games and the linear quadratic mean field game, in order to prove that the convergence is well-established. Also, we provide Monte Carlo simulation for trajectories and velocities at the equilibrium. Moreover, since the analytic form of linear quadratic mean field games can be decomposed into Ricatti equations, we implement fixed point and Newton algorithms to tackle the forward-backward structure in the equation system. Following the analytic approach, we also review and compare several models of exhaustible resource production, such as oil. Furthermore, we extend mean field games to modeling renewables by incorporating the stochastic growth of the resource. To exemplify this application, we propose a mean field game model for the exploitation of fish stock and derive the associated HJB and KFP equations to characterize the mean field equilibrium.

Access on SSRN