RoSE - Learning in Routing Games for Sustainable Electromobility
Making traffic routing for commercial operators more sustainable by accounting for electromobility, operational costs, infrastructure condition deterioration, and environmental externalities.
- Simulation for quantifying travel delays and costs, impact of vehicle loading on infrastructure, electrified transportation energy consumption and charging potential
- Optimization for computing optimal traffic routes for commercial actors, such as logistics operators of heavy duty trucks, allowing us to investigate trade-offs between total cost, fairness and externalities
- Game theory and learning for modeling interaction among myopic travelers, engaging in repeated play and using heterogeneous data sources. Develop learning schemes that drive travelers towards the social routing optimum
Algorithms and methods to answer questions such as:
- How to route heavy duty electric and plug-in hybrid vehicle fleets to balance sustainability (infrastructure deterioration), operational costs (fuel costs and timeliness), and power grid reliability constraints?
- How to integrate heterogeneous data from vehicles (trajectories and charging times) and networks (traffic count and environmental measurements) to estimate key externalities?
- How to develop learning strategies to ensure proper exploration for estimating costs, but still limit the environmental footprint of the majority of vehicles?