My research focuses on the modeling, analysis, and design of future transportation systems, with an emphasis on both short-term and long-term visions. For short-term objectives, I have developed data-driven methodologies for descriptive and prescriptive analysis of complex, large-scale transportation systems. However, what truly distinguishes my work is my long-term vision, where, in collaboration with many great partners, I have proposed innovative strategies for coordinating large-scale future traffic systems. The following highlights three key aspects of my research.
Rhythmic Control
Rhythmic Control (RC) is an innovative traffic control framework designed for connected and automated environments. RC introduces the concept of regularly and recurrently moving virtual containers within transportation facilities. Vehicles flow seamlessly within these pre-designed virtual containers, enabling trips to be completed without collisions or unnecessary stops at conflict points. The framework has undergone rigorous design and analysis, addressing both intersection-level and network-level implementations. RC marks a significant departure from traditional traffic management by transforming the current decentralized self-organized traffic systems into more structured and dynamically regulated systems. This shift represents a groundbreaking approach to managing traffic at the system level, offering a new school of thoughts for large-scale agent coordination.
Key materials on RC are provided below.
Rhythmic Control: Youtube Video (by Dr. Yafeng Yin)
We also have a book draft of RC (in Chinese) that is in press.
Orchestrated Transportation
Connectivity and automation are poised to transform transportation systems into more regularized and schedulized frameworks, as exemplified by the concept of Rhythmic Control (RC). Building on this foundation, we propose that future transportation systems should be "orchestrated" to maximize overall efficiency. Such an orchestrated system would seamlessly coordinate vehicle movements, human behaviors, and the allocation and matching of resources among participants within the system. My research focuses on addressing key questions critical to this vision: 1) Theoretical analysis for the benefits of system-level orchestration; 2) Performance evaluations of orchestrated transportation systems; 3) Development of effective mechanisms for managing participants behaviors; and 4) Exploration of implementation challenges and solutions for orchestrated systems in real-world scenarios.
Materials on this aspect will be available upon request.
With advancements in communication, data, and AI technologies, the modeling and analysis of transportation systems are undergoing a gradual transformation. Unlike fields such as computer vision or natural language processing, the transportation domain faces significant challenges due to limited and underrepresented data availability. Addressing these challenges requires a thoughtful integration of domain expertise, data engineering, machine learning, and decision-making techniques. My research in this area spans a range of topics, including traffic data estimation and imputation, predicting future traffic dynamics, and data-driven decision-making for large-scale transportation systems.
Some materials on this direction are provided below.
In addition to the three major research directions mentioned above, I have coauthored many papers on a variety of topics in transportation research. Below are a few selected works that I believe are worth highlighting. I welcome any suggestions or feedback—feel free to reach out.