Research

My research focuses on Logistics Network Optimization, Reinforcement Learning, and Embodied AI.

Logistics Network Optimization

I advance the optimization of logistics networks through stochastic modeling and reinforcement learning for linehaul service networks.

• Focus: Studying routing policies, demand uncertainty, and dynamic decision-making in large-scale freight transportation networks.
• Impact: Applications include freight flow allocation, hub-and-spoke network design, and real-time linehaul optimization.

Online Reinforcement Learning

I develop reinforcement learning algorithms for sequential decision-making under uncertainty in complex transportation environments.

• Focus: Studying online exploration-exploitation trade-offs, regret minimization, and convergence guarantees in stochastic routing problems, with applications to mixed-autonomy traffic networks.
• Impact: Applications include online routing of autonomous vehicles in mixed-autonomy environments, real-time adaptive decision-making, and provably efficient online policy optimization in large-scale stochastic networks.

Embodied AI

I develop multi-agent reinforcement learning algorithms for embodied control in continuous physical environments.

• Focus: Studying cooperative multi-agent policy learning under partial observability, with model-based offline reinforcement learning to address data inefficiency in embodied locomotion tasks across MuJoCo benchmarks.
• Impact: Applications include coordinated locomotion of multi-joint robots, sim-to-real transfer for autonomous physical agents, and scalable offline training for embodied AI systems.