Dynamic Multi-Agent Operations

Research

Operational decision methods for coordinating heterogeneous agents and shared resources in dynamic logistics systems.

My research studies operational decisions in logistics systems where people, robots, vehicles, vessels, drones, and shared resources depend on one another. The central problems are synchronized operations and resource-efficient online decisions.

Research structure

The map shows how each project combines research areas and operating environments. Select a project, decision area, or environment to examine the connections.

Methods and evaluation

I use mathematical optimization, tailored heuristics, approximate dynamic programming, reinforcement learning, and data-driven decision models. The choice of method follows the operational problem and the time available for a decision.

Computational experiments and simulation measure performance. Interactive visualization supports diagnosis: it helps reveal fleet behavior, failure modes, and differences between methods, which can then guide revisions to the model or algorithm.

Model and solution method Computational experiment or simulation Measurement and interactive visual diagnosis

Research records