Research

Our group draws diverse tools from control, optimization, machine learning, and economics to enable advanced applications for the development of sustainable Smart Cities. Specifically, our recent research focuses on (a) the design and analysis of intelligent transportation systems; (b) the economic operations of modern smart grids with deep penetration of renewable energy; and (c) the integration between distinct urban infrastructure systems for synergistic operation. Overall, my research interest can be summarized into the following themes:

Emerging Mobility: ride-sourcing platform, multi-modal transport, on-demand delivery
Autonomous Vehicles: autonomous mobility-on-demand, control of mixed fleet, human machine interaction
Transport Electrification: control of EV fleet, charging infrastructure, EV policies
Smart Grids: ancillary services, renewable integration, distributed energy resources, transactive control
Transport-Energy Nexus: joint planning and control of transportation and power infrastructures.
Machine Learning: deep reinforcement learning, multi-agent reinforcement learning, large language models. etc.
Optimization and Control: non-convex optimization, network optimization, optimal control,
Economics: game theory, mechanism design, power economics, transportation economics

Selected Projects:

Figure Courtesy: https://www.careserviceproject.eu/new-mobility-services/

Regulations on Emerging Mobility

Emerging mobility systems (e.g., ridesharing, micromobility) have fundamentally changed the way people travel in the city. However, their rapid growth has also created public concerns on the aggravated traffic congestion, the competition with public transit, and the working conditions of their employees. This prompted various cities to make regulations to curb these negative externalities. We develop economic equilibrium models to quantify the impacts of these policies, and use the empirically calibrated model to derive regualtory guidance for policy makers.

Control and Operation of Autonomous Mobility-on-demand Systems

Autonomous vehicles will revolutionize urban mobility. It may reduce vehicle ownership, eliminate the need for parking in dense areas, and enable affordable mobility-on-demand services around the clock. Our group develops multi-agent reinforcement learning algorithms to investigate how to efficiently operate autonomous mobility-on-demand systems, how they interact with human drivers over the transport network, and how they compete with other transport modes in a multimodal transportation system.

img-connected-automated-vehicles-public-transportation.jpg

Figure Courtesy: https://www.wsp.com/en-SG/insights/connected-automated-vehicles-and-public-transportation

Figure Courtesy: https://www.nytimes.com/2021/02/12/technology/electric-cars-batteries.html

Promoting Electric Vehicles with Innovative Business Models

Transport accounts for a significant portion of carbon emissions. Therefore, transport electrification can play an important role in carbon reduction. However, in many large cities, inconvenience of charging still remains a bottleneck for large-scale EV adoption. Our group investigates how to plan charging infrastructures to ease the charging anxiety, how to operate EV fleet to improve efficiency while reducing carbon emissions, and how to enact thoughtful policies to promote a smooth transition to an electrified urban mobility future

Balancing Power Grids under Deep Renewable Penetration

Integrating renewable energy is crucial for building a sustainable urban future. However, renewable energy (such as wind and solar) is uncontrollably and uncertain, which poses significant challenges for grid operations. Our group explores various technical solutions from both demand and supply side to enable large-scale integration of renewable energy at an affordable cost.

Mapping Residents’ Perception of Safety through Street View Images and Interpretable AI

Leveraging advancements in Geographic Artificial Intelligence (GeoAI) and urban big data, this study aims to design and train a GeoAI model using Google Street View images and satellite imagery to measure human visual perception of safety. The project will generate a comprehensive safety perception map of Hong Kong and explore factors influencing safety perception and "perception bias" with interpretable AI techniques. The findings will provide valuable insights for urban design and policy-making, supporting the development of safer, greener, and more livable environments, particularly in major projects like Lantau Tomorrow Vision and the Northern Metropolis.

Works related to this project:

(1) Hou, Ce, Fan Zhang, Yuhao Kang, Song Gao, Yong Li, Fábio Duarte, and Sen Li. "Transferred Bias Uncovers the Balance Between the Development of Physical and Socioeconomic Environments of Cities." Annals of the American Association of Geographers 115, no. 1 (2025): 148-166.

(2) Hou, C., Zhang, F., Li, Y., Li, H., Mai, G., Kang, Y., Yao, L., Yu, W., Yao, Y., Gao, S. and Chen, M., 2025. Urban sensing in the era of large language models. The Innovation, 6(1).