DONGHO KANG

Legged robotics, character animation, optimal control, and reinforcement learning

profile.jpg

Wasserwerkstrasse 12

8092 Zurich

Switzerland

I am a doctoral student in computer science based in Zurich, Switzerland 🇨🇭. From 2019, I am a scientific assistant of the Computational Robotics Lab, ETH Zurich, under the supervision of Prof. Stelian Coros.

The goal of my research is to create legged robots that exhibit natural and animal-like behaviors. Thus, my research interests are broad ranging from legged locomotion control to computational models of character animation.

Apart from my professional bio, I am originally from Seoul, South Korea 🇰🇷. Describing myself, I am a video game fan, and a classical music lover.

I am currently looking for self-motivated Master students (from ETH or students from any other institutes who can visit ETH Zurich by an available exchanging program) who are passionate to join our journey to build a bio-inspired legged robot. If you are intested in participating in one of our 3-month (14-weeks, part-time) or 6-month (full-time) research projects at Computational Robotics Lab, please see more details here if you are interested.


The Computational Robotics Lab is looking for a talented mechatronics engineer who has a strong background in designing, and building a robotic system. If you are interested in joing CRL as a part-time or a full-time research engineer, please contact us through crl-jobs (at) inf (dot) ethz (dot) ch: more details are availabe here.

news

Oct 16, 2023 My recent collaborative work titled “Tuning Legged Locomotion Controllers via Safe Bayesian Optimization,” co-authored with Daniel Widmer, Bhavya Sukhija, and Jonas Hübotter, under the guidance of Prof. Andreas Krause and Prof. Stelian Coros, has been accepted to the 7th Annual Conference on Robot Learning (CoRL 2023)! Visit the project website to learn more.
Aug 23, 2023 We present our latest work, “RL + Model-based Control: Using On-demand Optimal Control to Learn Versatile Legged Locomotion”. The paper is published to the IEEE Robotics and Automation Letters (RA-L) journal. To learn more about the paper, please visit the project website.
Jun 30, 2022 I am happy to announce my paper, "Animal Motions on Legged Robots Using Nonlinear Model Predictive Control" has been accepted to 2022 IEEE/RSJ International Conference on Intelligent Robots and Systems (IROS 2022).
Jul 9, 2021 I am happy to announce my two papers have been accepted to 2021 IEEE/RSJ International Conference on Intelligent Robots and Systems (IROS 2021):
Mar 7, 2020 I am happy to announce that I am starting my Ph.D. at ETH Computational Robotics Lab. with the guidance of Prof. Stelian Coros. I participate in legged robot projects of CRL, and work with GhostRobotics Vision 60, Unitree Laikago robots.

selected publications

2023

  1. widmer2023tuning.png
    Tuning Legged Locomotion Controllers via Safe Bayesian Optimization
    Daniel Widmer, Dongho Kang, Bhavya Sukhija, and 3 more authors
    2023
  2. kang2023rl.png
    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

2022

  1. kang2022animal.png
    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
  2. kang2022nmpc.png
    Nonlinear Model Predictive Control for Quadrupedal Locomotion Using Second-Order Sensitivity Analysis
    Dongho Kang, Flavio De Vincenti, and Stelian Coros
    2022

2021

  1. kang2021animal.png
    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
  2. devincenti2021control.png
    Control-Aware Design Optimization for Bio-Inspired Quadruped Robots
    Flavio De Vincenti, Dongho Kang, and Stelian Coros
    In 2021 IEEE/RSJ International Conference on Intelligent Robots and Systems (IROS), 2021