Energy-Aware Coordination of Electric Vehicle Fleets for Resilient Transport Systems during Emergency Evacuations

Research Project, 2027 – 2028

Climate change is increasing the frequency and severity of natural disasters such as floods, wildfires, and extreme weather events, making efficient and large-scale evacuation planning increasingly critical. Societal conflicts can lead to new forms of tension and potential crises. At the same time, the transport sector is undergoing a rapid transition toward electrification. While electric vehicles (EVs) are central to achieving climate-neutral mobility, most existing evacuation planning methods still assume conventional fuel vehicles and do not account for EV-specific constraints such as limited battery capacity, energy consumption variability, charging infrastructure availability, time-consuming recharging, and interactions with the electricity grid. This gap creates a critical challenge for designing evacuation strategies that remain effective in future electrified transport systems. This project aims to develop a new computational framework for coordinating fleets of EVs during large-scale emergency evacuations under energy and infrastructure constraints.

The project formulates the Multi-Agent Electric Vehicle Evacuation Orienteering Problem (MAEV-EOP) and models evacuation as a multi-agent, multi-objective decision process. Using advanced AI methods such as multi-agent reinforcement learning (MARL), the framework will enable coordinated decision-making among multiple EVs while balancing objectives such as evacuation coverage, evacuation time, energy consumption, charging requirements, and impacts on the electricity grid.

The project will deliver scalable models and open-source simulation tools for evaluating EV-based evacuation strategies under various conditions. The results can support transportation planners, emergency management authorities, and energy system operators in designing large-scale resilient and energy-aware evacuation systems for electrified transport networks.