向未曾抵达之境Toward the unknown
Co-founder & Chief Center Director · Freljord (Beijing) AI Laboratory
I’m a final-year Automation undergraduate at Northeastern University (new tab), currently in research collaboration and joint training at Beihang University (new tab). I work with my senior labmate, Prof. Jiarong Lin (new tab) (Associate Professor at Beihang, ranked among the world’s top 2% of scientists, and a National Young Talent), on UAV world models, spatial intelligence, and aerial agents.
In Fall 2027, I’ll begin a PhD in robotics with a full scholarship at the University of Hong Kong (new tab), joining Prof. Fu Zhang (new tab)’s MaRS Laboratory (new tab) and continuing the HKU–Beihang research collaboration and joint training. It’s a new challenge, and I look forward to spending more time on difficult research problems.
Automation · Final-year undergraduate
Undergraduate advisor: Prof. Yong Zhao (new tab)
Project 985 / 211 · Double First-Class university. President of NEU at Qinhuangdao · Distinguished Young Scholar · Changjiang Scholar
Control Science and Engineering: A+ and a national Double First-Class discipline; ranked second in the 2006 and 2012 national assessments. Read more (new tab)
Current research collaboration and joint training
Collaborating advisor: Prof. Jiarong Lin (new tab)
GAMMA-LAB / γ-LAB. Associate Professor at Beihang · World’s Top 2% Scientists · National Young Talent · 2025 IEEE T-RO Best Paper Award winner
Working closely with my senior labmate, Prof. Jiarong Lin (Associate Professor at Beihang, ranked among the world’s top 2% of scientists, and a National Young Talent), on world models for aerial robots. Read more (new tab)
Incoming Robotics PhD (Full Scholarship) · Fall 2027
PhD advisor: Prof. Fu Zhang (new tab)
Robotics · MaRS Laboratory. Director of the MaRS Laboratory
Aerial robotics, perception, navigation, and control. Read more (new tab)
My research collaboration and joint training at Beihang are based at the Institute of Unmanned Systems, led by Prof. Xiwang Dong (new tab), its dean, a National Leading Talent and a Grade-II Professor.
How can robots learn useful models of the physical world through interaction, and use that understanding to act?
My current work with my senior labmate, Prof. Jiarong Lin (Associate Professor at Beihang, ranked among the world’s top 2% of scientists, and a National Young Talent), focuses on world models for aerial robots. I’m also interested in spatial intelligence, reinforcement learning, agents, and recursive self-improvement, and how these ideas might contribute to artificial general intelligence (AGI).
Multimodal Physical Agent Algorithms Intern
Leading the development of multimodal agents for the physical world, connecting environment understanding, task planning, and action execution.
AIGC / Large-model Algorithms Intern
Tencent Games · Game AI algorithms
Worked on multimodal agents for game NPCs, combining visual input, text, and scene state for task planning and execution.
Autonomous Driving Systems Architecture Intern
Autonomous & Intelligent Systems R&D · L3 driving
Worked on a layered L3 driving architecture and multi-sensor fusion, connecting sensor interfaces, temporal alignment, and edge inference.
Robotic Perception Algorithms Intern
Centre for Frontier AI Research, Singapore
Studied visual world models based on latent dynamics, learning environmental representations and future states from RGB-D observations.
Research Intern · Multimodal Perception
Optoelectronic sensing & artificial intelligence
Worked on a multimodal driver-fatigue perception system, combining visual, audio, and time-series signals in a working application.
Nav Combat Workspace
A unified ROS workspace connecting localization, navigation, vision, chassis and gimbal control, referee communication, and behavior-tree decisions.
Abot Cruise Workspace
A complete ROS 1 system for visual autonomous navigation, integrating chassis drivers, LiDAR SLAM, localization, vision, and task scheduling.
During my undergraduate studies: 15 national competition awards, including 9 national first prizes, and 38 software copyright registrations as first author.
2023–2024 academic year · First in the school’s comprehensive ranking
Robot Task Challenge · Team captain
National final · Software lead
AI+ special track · Algorithm lead
Outdoor unmanned vehicle track · Team captain
Smart delivery track · Team captain
Software copyrightsProject approvalsUniversity honorsMedia coverage
A moment from a technical talk I gave at DJI. I also plan to share paper readings and engineering notes on Bilibili. Videos coming soon.
Find me on Bilibili (new tab)If you’re working on robotics, world models, or agents, I’d be happy to hear from you.