Robot Manipulation
Dexterous and mobile manipulation, long-horizon tasks, and task-and-motion planning.
PhD Researcher · Sydney, Australia
Intelligence,in motion.
I’m Zihao Li, a robotics researcher building systems that perceive, reason, and act in the physical world.
↓Interactive robot workspace. Move the pointer to direct the robot arm. Click to set a planned target.
PERCEPTION01REASONING02ACTION03
01Research direction
From pixels and words
to purposeful action.
My research sits at the intersection of robot manipulation, embodied intelligence, and multimodal learning. I develop generalizable robot policies for complex, long-horizon tasks.
I’m currently a Computer Science PhD student at The University of Sydney, advised by Prof. Weiming (William) Zhi.
Embodied Intelligence
Visuomotor policy learning, imitation and reinforcement learning, generalization, and adaptation.
Vision–Language–Action
Multimodal representations, language-conditioned control, and foundation models for robotics.
PERCEIVEREASONACTPERCEIVEREASONACT
CURRENT THESIS VECTOR
How can robots preserve what matters, ignore what doesn’t, and coordinate action when evidence arrives late?
02Trajectory
Experience & education
A path through mechanical systems, control, planning, and embodied AI.
The University of Sydney
PhD in Computer Science · Advisor: Prof. Weiming Zhi
Zeno AI
Embodied AI Researcher · Whole-body teleoperation and embodied AI
Computational Robotics Lab · ETH Zurich
Research Intern · Task-and-motion planning for autonomous scaffolding assembly
Zhejiang University
Control Science and Engineering · Multimodal perception and control for manipulation
Zhejiang University
Mechatronics Engineering & Automation · Chu Kochen Honors Program
03Selected publications
Ideas in print
Full record on Scholar ↗
01↗
02↗
03↗
04↗
05↗
2026ARXIVFIRST AUTHOR
TRACE: Trajectory-Routed Causal Memory for Delayed-Evidence Visuomotor Imitation
Zihao Li, Ranpeng Qiu, Yincong Chen, Guoqiang Ren, Weiming Zhi2026ARXIVFIRST AUTHOR
TriPilot-FF: Coordinated Whole-Body Teleoperation with Force Feedback
Zihao Li, Yanan Zhou, Ranpeng Qiu, Hangyu Wu, Guoqiang Ren, Weiming Zhi2025ARXIVFIRST AUTHOR
TAPOM: Task-Space Topology-Guided Motion Planning for Manipulating Elongated Objects
Zihao Li, Yiming Zhu, Zhe Zhong, Qinyuan Ren, Yijiang Huang2026CONSTRUCTION ROBOTICS
A Robot-Friendly Scaffolding System with Passive Error Correction
Pok Yin Victor Leung, Yijiang Huang, Yen-Ting Liu, Jakob Genhart, Zihao Li, et al.2026ARXIV
Robots that Collaborate: Sequential Asymmetric Imitation for Coupled Robot Policies
Yincong Chen, Ranpeng Qiu, Zihao Li, Yanan Zhou, Guoqiang Ren, Weiming ZhiMore publications
Vision-Force Guided Robotic EV Charging
Motion Control for Physical Human–Robot–Environment Interaction
Meta-RL Robust to Distributional Shift via Lifelong In-context Learning
A Head Stabilizer of a Snake Robot in Serpentine Locomotion
ZL / 202604Off the trajectory
Still curious
after hours.
When I’m not thinking about robots, you’ll find me playing basketball or badminton, swimming, or exploring the worlds of Hollow Knight and God of War.