DONGHO KANG

Elias-Canetti-Strasse 2

8050 Zurich

Switzerland

I am a Research Scientist at the Robotics and AI Institute (RAI Institute) based in Zurich, Switzerland 🇨🇭.

Since joining the RAI Institute, I have been part of the collaborative humanoid robotics research initiative with Boston Dynamics (BD), developing control algorithms for natural, fully steerable locomotion that was demonstrated live on stage during the launch of the new BD Atlas at CES 2026.

Prior to this, I completed my PhD in Computer Science in 2025 at the Computational Robotics Lab, ETH Zurich, under the supervision of Prof. Stelian Coros. My doctoral research focused on developing control algorithms that enable legged robots to move like real animals by leveraging biological motion capture data.

Today, my research centers on developing embodied intelligence for humanoid robots, enabling natural, effective, and adaptive whole-body loco-manipulation in real-world environments at human cadence. To achieve this, my core work bridges optimal control, reinforcement learning, and data-driven animation, with a keen interest in extending these frameworks through vision-language models (VLMs) to link semantic scene understanding with low-level motor control.

I am originally from Seoul, South Korea 🇰🇷, where I spent the first 25 years of my life. Outside of research, I am a huge fan of video games and classical music, and I also enjoy animated and sci-fi films.

news

Sep 04, 2026	Excited to announce that two collaborative papers with Taerim Yoon have been accepted to CoRL 2026: “Teaching Robots Like Dogs: Learning Agile Navigation from Luring, Gesture, and Speech” and “Learning Unions of Convex Sets via Invertible Latent Decomposition for Path Planning.” Thank you to all the co-authors, and huge congratulations to Taerim for leading these amazing projects!
Feb 10, 2026	“Teaching Robots Like Dogs,” a collaborative study with Taerim Yoon from Korea University, has received media attention from Tech Xplore (English) and Yonhap News (Korean). The paper explores agile navigation through luring, gesture, and speech. Read the full paper and see the demonstrations on the project webpage.
Nov 27, 2025	Check out Lukas Molnar’s recent work, “Whole-body Inverse Dynamics MPC for Legged Loco-Manipulation,” to see how we address the challenge of loco-manipulation tasks. This paper introduces a Whole-Body MPC framework that unifies motion and force planning within a single control layer. The paper has been accepted to IEEE Robotics and Automation Letters (RA-L) for the special issue on “Advancement on MPC and Learning Algorithms for Legged Robots.”
Oct 07, 2025	I’m thrilled to share that a collaborative project I was a part of, “SPiDR: A Simple Approach for Zero-Shot Safety in Sim-to-Real Transfer” has been accepted to NeurIPS 2025. Huge congrats to Yarden As for leading this excellent work!
Aug 28, 2025	After an incredible journey completing my PhD, I’m starting my next chapter as a research scientist at the RAI Institute. I am excited to join the amazing team and contribute to the mission of creating the next generations of intelligent robots. Looking forward to the exciting challenges and collaborations ahead.

selected publications

HuMBLE: Human Motion-Driven Behavior Learning for Embodied Locomotion

Mike Zhang, Dongho Kang, Kevin Bergamin , and 19 more authors

2026

arXiv Bib

@misc{zhang2026humble,
  title = {HuMBLE: Human Motion-Driven Behavior Learning for Embodied Locomotion},
  author = {Zhang, Mike and Kang, Dongho and Bergamin, Kevin and Burger, Nicola and Deits, Robin and Foster, Jonathan and Hammoud, Bilal and Hughes, Katie and Iacobelli, Francesco and Koolen, Twan and Mungai, M. Eva and Nobles, Zach and Rozen-Levy, Shane and Sleiman, Jean Pierre and Yu, Fangzhou and Zhang, Yunbo and Rizzi, Alfred and Hodgins, Jessica and Kuindersma, Scott and Abe, Yeuhi and Bertrand, Sylvain and Farshidian, Farbod},
  year = {2026},
  eprint = {2610.10489},
  archiveprefix = {arXiv},
  primaryclass = {cs.RO},
  url = {https://arxiv.org/abs/2610.10489},
}

Teaching Robots Like Dogs: Learning Agile Navigation from Luring, Gesture, and Speech

Taerim Yoon, Dongho Kang, Jin Cheng , and 5 more authors

2026

arXiv Bib Website

@misc{yoon2026teaching,
  title = {Teaching Robots Like Dogs: Learning Agile Navigation from Luring, Gesture, and Speech},
  author = {Yoon, Taerim and Kang, Dongho and Cheng, Jin and Zargarbashi, Fatemeh and Huang, Yijiang and Ahn, Minsung and Coros, Stelian and Choi, Sungjoon},
  eprint = {2601.08422},
  archiveprefix = {arXiv},
  primaryclass = {cs.RO},
  url = {https://arxiv.org/abs/2601.08422},
  year = {2026}
}

Multi-Domain Motion Embedding: Expressive Real-Time Mimicry for Legged Robots

Matthias Heyrman, Chenhao Li, Victor Klemm , and 3 more authors

2025

arXiv Bib

@misc{heyrman2025multi,
  title = {Multi-Domain Motion Embedding: Expressive Real-Time Mimicry for Legged Robots},
  author = {Heyrman, Matthias and Li, Chenhao and Klemm, Victor and Kang, Dongho and Coros, Stelian and Hutter, Marco},
  eprint = {2512.07673},
  archiveprefix = {arXiv},
  primaryclass = {cs.RO},
  url = {https://arxiv.org/abs/2512.07673},
  year = {2025}
}

Walk Like Dogs: Learning Steerable Imitation Controllers for Legged Robots from Unlabeled Motion Data

Dongho Kang, Jin Cheng, Fatemeh Zargarbashi , and 3 more authors

2026

arXiv Bib Supp Website

@misc{kang2026walk,
  title = {Walk Like Dogs: Learning Steerable Imitation Controllers for Legged Robots from Unlabeled Motion Data},
  author = {Kang, Dongho and Cheng, Jin and Zargarbashi, Fatemeh and Yoon, Taerim and Choi, Sungjoon and Coros, Stelian},
  year = {2026},
  eprint = {2507.00677},
  archiveprefix = {arXiv},
  primaryclass = {cs.RO},
  url = {https://arxiv.org/abs/2507.00677},
}

Spatio-Temporal Motion Retargeting for Quadruped Robots

Taerim Yoon, Dongho Kang, Seungmin Kim , and 4 more authors

IEEE Transactions on Robotics, 2025

arXiv Bib HTML Supp Website

@article{yoon2025spatiotemporal,
  author = {Yoon, Taerim and Kang, Dongho and Kim, Seungmin and Cheng, Jin and Ahn, Min Sung and Coros, Stelian and Choi, Sungjoon},
  journal = {IEEE Transactions on Robotics},
  title = {Spatio-Temporal Motion Retargeting for Quadruped Robots},
  year = {2025},
  volume = {41},
  number = {},
  pages = {5471-5490},
  keywords = {Robots;Robot kinematics;Legged locomotion;Trajectory;Quadrupedal robots;Artificial intelligence;Training;Kinematics;Videos;Pipelines;Legged robots;learning from demonstration;motion control;motion retargeting},
  doi = {10.1109/TRO.2025.3600123}
}

Tuning Legged Locomotion Controllers via Safe Bayesian Optimization

Daniel Widmer, Dongho Kang, Bhavya Sukhija , and 3 more authors

In Proceedings of The 7th Conference on Robot Learning , 2023

arXiv Bib HTML Supp Website

@inproceedings{widmer2023turning,
  title = {Tuning Legged Locomotion Controllers via Safe Bayesian Optimization},
  author = {Widmer, Daniel and Kang, Dongho and Sukhija, Bhavya and H\"{u}botter, Jonas and Krause, Andreas and Coros, Stelian},
  booktitle = {Proceedings of The 7th Conference on Robot Learning},
  pages = {2444--2464},
  year = {2023},
  editor = {Tan, Jie and Toussaint, Marc and Darvish, Kourosh},
  volume = {229},
  series = {Proceedings of Machine Learning Research},
  publisher = {PMLR}
}

RL + Model-Based Control: Using On-Demand Optimal Control to Learn Versatile Legged Locomotion

Dongho Kang, Jin Cheng, Miguel Zamora , and 2 more authors

IEEE Robotics and Automation Letters, 2023

arXiv Bib HTML Supp Website

@article{kang2023rl,
  author = {Kang, Dongho and Cheng, Jin and Zamora, Miguel and Zargarbashi, Fatemeh and Coros, Stelian},
  journal = {IEEE Robotics and Automation Letters},
  title = {RL + Model-Based Control: Using On-Demand Optimal Control to Learn Versatile Legged Locomotion},
  year = {2023},
  volume = {8},
  number = {10},
  pages = {6619-6626},
  doi = {10.1109/LRA.2023.3307008}
}

Animal Motions on Legged Robots Using Nonlinear Model Predictive Control

Dongho Kang, Flavio De Vincenti, Naomi C. Adami , and 1 more author

In 2022 IEEE/RSJ International Conference on Intelligent Robots and Systems (IROS) , 2022

Bib HTML PDF Supp Website

@inproceedings{kang2022animal,
  title = {Animal Motions on Legged Robots Using Nonlinear Model Predictive Control},
  author = {Kang, Dongho and De Vincenti, Flavio and Adami, Naomi C. and Coros, Stelian},
  booktitle = {2022 IEEE/RSJ International Conference on Intelligent Robots and Systems (IROS)},
  pages = {11955-11962},
  year = {2022},
  organization = {IEEE},
  doi = {10.1109/IROS47612.2022.9981945}
}

Animal Gaits on Quadrupedal Robots Using Motion Matching and Model-Based Control

Dongho Kang, Simon Zimmermann, and Stelian Coros

In 2021 IEEE/RSJ International Conference on Intelligent Robots and Systems (IROS) , 2021

Bib HTML PDF Supp Website

@inproceedings{kang2021animal,
  title = {Animal Gaits on Quadrupedal Robots Using Motion Matching and Model-Based Control},
  author = {Kang, Dongho and Zimmermann, Simon and Coros, Stelian},
  booktitle = {2021 IEEE/RSJ International Conference on Intelligent Robots and Systems (IROS)},
  pages = {8500--8507},
  organization = {IEEE},
  year = {2021},
  doi = {10.1109/IROS51168.2021.9635838}
}