Ruihao Zeng

Hi, I am Ruihao (Andrew) Zeng, a Ph.D. student in TransportLab, School of Civil Engineering, The University of Sydney, supervised by Assoc. Prof. Mohsen Ramezani and Prof. David Levinson. My Ph.D. (2024 – 2027) is supported by the Australian Government Research Training Program (RTP) Scholarship, and my thesis is on Distributed Intelligence in Systems of Autonomous Vehicles.

My research focuses on Distributed / Decentralised Autonomous Vehicles, Deep Reinforcement Learning for Multi-agent Vehicle Control, 3D Perception and Multi-object Tracking with LiDAR, and Spatio-temporal Traffic Prediction. I like building end-to-end systems, from data pipelines and large-scale simulation to real-time deployment in the cloud. I hope to use intelligent methods to solve problems in transportation and society.

Before my Ph.D., I completed an M.Phil. in Transportation at TransportLab (2022 – 2023) with High Distinction. My M.Phil. thesis was Online Multi-Object Tracking Using LiDAR, and I received the Paulette Isabel Jones Career Development Award. I received my B.E. degree in Digital Media Technology & Computer Science in 2022, graduating first in my cohort as the valedictorian.

From November 2020 to December 2022, I worked as a research assistant at the Institute of Transportation Big Data of Digital Fujian, co-advised by Xing Wang and Fumin Zou, developing deep-learning models for traffic flow prediction, trajectory prediction and traffic-anomaly detection on city-wide sensor data (7M+ records). I also worked for a summer research program on GAN-based line art colorization at Massachusetts Institute of Technology in 2021, supervised by Mark Vogelsberger.

E-mail  |  Google Scholar  |  GitHub  |  LinkedIn

• [Aug 2026] Oral presentation at the 23rd IFAC World Congress, Busan, South Korea.
• [Aug 2026] Our paper "Deep Reinforcement Learning for Communication-Free Distributed Control of Autonomous Vehicles in Unstructured Intersection" was published in IEEE Transactions on Cybernetics!
• [2026] Oral presentation at the IEEE ITSS NSW Chapter Inaugural Meetup, Sydney, Australia.
• [2026] Co-authored paper "FedTETP: Federated learning with topology enhancement for traffic prediction" was accepted by Future Generation Computer Systems.
• [2025] Our paper "A dynamic-confidence 3D multi-object tracking method based on spatio-temporal association" was published in Measurement!
• [Jul 2025] Oral presentation at the TRB Traffic Flow Theory and Characteristics Committee (TFTC) Summer Meeting – Symposium on Advancing Traffic Flow Theory for a Sustainable Future, Cairns, Australia.
• [May 2025] Poster presentation at the Transportation Research Symposium (TRS 2025), Rotterdam, Netherlands.
• [2025] Poster presentation at the Transport Research Association for NSW (TraNSW) Annual Symposium, Sydney, Australia.
• [2025] Co-authored papers were published in Applied Intelligence (DSTSPYN), Neurocomputing (DSTF) and Expert Systems with Applications (FedCTPM).
• [2024] Started my Ph.D. at TransportLab, The University of Sydney, supported by the Australian Government RTP Scholarship.
• [Jan 2024] Poster presentation at the Transportation Research Board Annual Meeting (TRBAM 2024), Washington, D.C., USA.
• [2024] Oral presentation at the TraNSW Annual Symposium, Sydney, Australia.
• [2023] Completed my M.Phil. with High Distinction and received the Paulette Isabel Jones Career Development Award.
• [2023] Most Innovative Award at the SCATS Student Traffic Data Innovation Challenge (NSW Government & iMOVE).
• [2023] Oral presentation at the TraNSW Annual Symposium, Sydney, Australia.
• [2023] Poster presentation at the HDR Conference, Faculty of Engineering, The University of Sydney.
• [Oct 2022] Our paper "STTF: An efficient transformer model for traffic congestion prediction" was accepted by International Journal of Computational Intelligence Systems.
• [Oct 2022] START my research journey in TransportLab, The University of Sydney!
• [Aug 2021] Our paper "A highly efficient framework for outlier detection in urban traffic flow" was accepted by IET Intelligent Transport Systems, Congrats :)

Full list on Google Scholar. Representative papers are highlighted. ^* denotes the corresponding author.

A continuous-PPO framework with curriculum learning and a reference-action-guided loss that controls acceleration and steering of autonomous vehicles in signal-free, lane-free intersections without any inter-vehicle communication. A rule-embedded hybrid policy supports long-term deployment safety: over 100M+ km of cumulative simulation, only 36 collisions were recorded.

TransportLab, The University of Sydney.

An online LiDAR-based 3D multi-object tracking method for moving sensors: a constant-acceleration motion model with orientation variation, a spatio-temporal feature-based data association model, and a dynamic confidence threshold to keep tracking temporarily occluded objects. Evaluated on KITTI and nuScenes.

Resulted from my M.Phil. thesis at TransportLab, The University of Sydney.

Institute of Transportation Big Data of Digital Fujian.

Supported by the Foreign Cooperation Project of Fujian Provincial Department of Science and Technology.

Institute of Transportation Big Data of Digital Fujian.

Supported by the Natural Science Foundation of China and the Foreign Cooperation Project of Fujian Provincial Department of Science and Technology.

• FedTETP: Federated Learning with Topology Enhancement for Traffic Prediction
  Xing Wang, Chunxia Chen, Jixiang Chen, Biao Jin, Fumin Zou, Lyuchao Liao and Ruihao Zeng
  Future Generation Computer Systems, 2026


• FedCTPM: A Federated Cross-Regional Collaboration Learning Approach for Traffic Flow Prediction
  Xing Wang, Jixiang Chen, Feifei Chen, Biao Jin, Fumin Zou, Lyuchao Liao and Ruihao Zeng
  Expert Systems with Applications, 2025


• DSTF: A Diversified Spatio-Temporal Feature Extraction Model for Traffic Flow Prediction
  Xing Wang, Xiaojun Wang, Faliang Huang, Fumin Zou, Lyuchao Liao and Ruihao Zeng
  Neurocomputing, vol. 621, 129280, 2025


• DSTSPYN: A Dynamic Spatial-Temporal Similarity Pyramid Network for Traffic Flow Prediction
  Xing Wang, Feifei Chen, Biao Jin, Mingwei Lin, Fumin Zou and Ruihao Zeng
  Applied Intelligence, vol. 55, 237, 2025

A triple-agent reinforcement-learning architecture (a TD3-based decision agent, a convex-optimisation solver agent and a market-observer agent) coupled in a loosely-coupled pipelining model, with Jensen–Shannon-divergence-based reward shaping to prevent policy collapse under volatile markets. Achieved a 22.49% annual return and a Sharpe ratio of 0.92 on the S&P 500, and outperformed 10 baselines (incl. EIIE, RAT, PPN) across CSI 300, DJIA and S&P 500 over a 10-year horizon.

A real-time LiDAR system that detects parking occupancy and visualises results through AWS-hosted web applications. Built EC2-based scripts for point-cloud parsing, map construction and automated parking-space recognition, and integrated the SENSR perception software with AWS into a complete data-acquisition-to-cloud-deployment pipeline.

Stereo-camera computer vision detects Lego block layouts and reconstructs a virtual city in real time; road-network generation algorithms convert the physical layout into a digital graph, and a real-time traffic simulation visualises traffic flow on the user-created road structure.

Learned a lot :)

Honors & Awards:

• Most Innovative Award of SCATS Student Traffic Data Innovation Challenge (NSW Government & iMOVE). (2023)
• Second Prize of National Student Mathematical Modeling Competition. (2021)
• Bronze Medal of Kaggle HuBMAP Competition. (2021)
• Honorable Mention of Interdisciplinary Contest in Modeling (ICM) & Mathematical Contest in Modeling (MCM). (2021)
• First Prize of the 10th MathorCup College Mathematical Modeling Challenge. (2020)

Conferences:

• Oral presentation, 23rd IFAC World Congress – Busan, South Korea. (2026)
• Oral presentation, IEEE ITSS NSW Chapter Inaugural Meetup – Sydney, Australia. (2026)
• Oral presentation, Summer Meeting of the Traffic Flow Theory and Characteristics Committee (TFTC) – Symposium on Advancing Traffic Flow Theory for a Sustainable Future – Cairns, Australia. (2025)
• Poster presentation, Transportation Research Symposium (TRS) – Rotterdam, Netherlands. (2025)
• Poster presentation, Transport Research Association for NSW (TraNSW) Annual Symposium – Sydney, Australia. (2025)
• Poster presentation, Transportation Research Board Annual Meeting (TRBAM) – Washington, D.C., USA. (2024)
• Oral presentation, Transport Research Association for NSW (TraNSW) Annual Symposium – Sydney, Australia. (2024)
• Oral presentation, Transport Research Association for NSW (TraNSW) Annual Symposium – Sydney, Australia. (2023)
• Poster presentation, HDR Conference, Faculty of Engineering, The University of Sydney – Sydney, Australia. (2023)

• Exam Supervision Officer, The University of Sydney. (Jun. 2026 – Present)
• Higher Degree Research (HDR) Student Representative, School of Civil Engineering, The University of Sydney. (Jan. 2025 – Present)
• Tutor, Marker and Thesis Assessor, School of Civil Engineering, The University of Sydney. (Feb. 2023 – Present)
• Research Project Mentor (Capstone & VRI), School of Civil Engineering, The University of Sydney – LiDAR-based pedestrian detection and parking-occupancy mapping. (Dec. 2023 – Oct. 2024)
• Algorithm and Application Developer, Lego ARCity Project, Outreach Office of School of Civil Engineering, The University of Sydney. (May 2023 – Dec. 2023)
• Researcher & Developer, Parking-Lot Occupancy LiDAR Detection (POLD), TransportLab & Seoul Robotics. (Jan. 2023 – Jul. 2023)
• Urban Traffic Data Analyst, Service-Oriented Trajectory Prediction in Traffic Network, Institute of Transportation Big Data of Digital Fujian, Fuzhou, China. (Nov. 2020 – Dec. 2022)

Template modified from here.