Yi Cheng

Yi Cheng

chengy22@mails.tsinghua.edu.cn

I am a Master student in Artificial intelligence at Tsinghua University, advised by Houde Liu and Linqi Ye. Previously, I completed my Bachelor's in Mechanical design and automation at XiDian University, advised by Jiaji Wu.

My research area include Legged/Humanoid Robot. I care about and continue to think about robots and embodied intelligence in real application scenarios to achieve commercial closed loop.

Robots might not be ready to take over the world, but they're on their way to mastering the coffee run. If you're interested in diving deeper into my research or swapping entrepreneurial insights, let's connect!

Publications

	Multi-Brain Collaborative Control for Quadruped Robots Hang Liu, Yi Cheng, Rankun Li, Xiaowen Hu, Linqi Ye, Houde Liu CoRL, 2024, submitted webpage \| pdf \| video
	Agile and versatile bipedal robot tracking control through reinforcement learning Jiayi Li, Linqi Ye,Yi Cheng*, Xianglong Li, Houde Liu, Bin Liang ICRA, 2024, late breaking results pdf \| video
	Quadruped Robot Traversing 3D Complex Environments with Limited Perception Yi Cheng, Hang Liu, Guoping Pan, Linqi Ye, Houde Liu IROS 2024 ( Oral pitch ) webpage \| pdf \| arXiv \| video
	Structural Optimization of Lightweight Bipedal Robot via SERL Yi Cheng, Chenxi Han, Yuheng Min, Linqi Ye, Houde Liu, Hang Liu IROS 2024 ( Oral pitch ) webpage \| pdf \| video
	From Knowing to Doing: Learning Diverse Motor Skills through Instruction Learning Linqi Ye, Jiayi Li, Yi Cheng, Xianhao Wang, Bin Liang, Yan Peng arxiv, 2023, IEEE Transactions on Automation Science and Engineering( T-ASE ) submitted pdf \| arXiv \| video

Selected Projects

	27 Dofs Humanoid robot and Dual arm robot with dexterous hands 2024 ZereX, Founder The project is fully self-developed, including structure, motor, hardware and algorithms., tailored to flexible manufacturing and education. Robots operate on a fully data-driven learning algorithms, and the team achieved closed-loop validation from algorithms to hardware in just two months.
	Robot operating system —— AliothOS 2021 - now, Founder The os integrates both hardware and software systems, focusing on core innovations in kernel design, communication mechanisms, and deployment processes to address market issues like unstable communications and complex setups. For developers, it provides a built-in algorithm library and constructs the Alioth robotics community across hardware, cloud services, and algorithm ecosystems. \| Promotional Video \| IDE Video \|
	Deep Power——Quadruped Robot of Multi-Terrain Locomotion 2019 - 2020, Founder We built a quadruped robot with upper-mounted motors and timing belt-driven legs, using aluminum and 3D-printed parts. It features MPC+WBC for motion control and uses Intel's D435i and RoboSense's 32-line LiDAR for perception. Applied in unstructured special scenarios \| Video1 \| Video2 \| Video3 \|

Website template from here and here