Donghyun Kim

Assistant Professor

My overarching research goal is to develop legged robotic systems to enhance human life quality. Legged robotic systems have been highlighted as a strong mobile system that can go anyplace where humans can go. As a robotics researcher with expertise in legged locomotion study, I would like to use the technology to help people who need continuous care, monitoring, and assistance. Currently, I am extending my field from locomotion control to creating intelligent mobile systems by integrating vision/tactile sensing and autonomous navigation on legged robots. Moreover, I recently initiated a robotic guide dog research with a collaboration with Prof. Ivan Lee, a core faculty of the Center for Personalized Health Monitoring.

Developing quadruped robots that can guide blind or visually impaired people will make a remarkable impact for blind or visually impaired people by providing them with an efficient, economical, and sustainable solution for their independent mobility. We are approaching this problem in a human-centric manner rather than developing technology by speculating users’ needs. Therefore, we will involve extensive surveys and human studies in the project.

Current Research

My research interest lies in developing mobile robotic systems assisting people who have difficulties in independent mobility. My prior research has
focused on developing control methods to enhance legged robots' agility, stability, and robustness. I am extending my field to monitoring, guiding,
and assisting people with integrated vision/tactile sensing, advanced intelligence, and interactive control algorithms. One of my major projects is
creating a robotic guide dog. The research involves survey studies and human-robot interaction. Also, real-time perception, decision-making, and
interactive walking control are critical research subjects.

Another project that I initiated recently is developing a mobile motion capture system by utilizing depth cameras built on a quadruped robot. I am
also interested in creating new exoskeleton robots by devising new actuators and drivetrain mechanism. Our group is developing novel actuators having high torque density while maintaining clear transparency between the motor input and the actuator output.

groups.cs.umass.edu/daros

Academic Background

Post-docs (2019.01 - 2020.12) MIT, U.S.
Post-docs (2018.01 - 2018.12) University of Texas at Austin, U.S.
PhD (2012.09 - 2017.12) University of Texas at Austin, U.S.
MS (2010.03 - 2012.02) Seoul National University, South Korea
BS (2003.03 - 2007.07) Korea Advanced Institute of Science and Technology,
South Korea

expand

Publications

Kim, Sangbae, Meng Yee Chuah, Lindsay Epstein, Donghyun Kim, and Juan Romero. "Sensing system." U.S. Patent Application 16/409,537, filed September 24, 2020.

Md Mahmudur Rahman, Tauhidur Rahman, Donghyun Kim, Mohammad Arif Ul Alam. Knowledge Transfer across Imaging Modalities Via Simultaneous Learning of Adaptive Autoencoders for High-Fidelity Mobile Robot Vision, In IEEE-RAS International Conference on Intelligent Robots and System (IROS), 2021.

Donghyun Kim, Daniel Carballo, Jared Di Carlo, Benjamin Katz, Gerardo Bledt, Bryan Lim, and Sangbae Kim. Vision Aided Dynamic Exploration of Unstructured Terrain with a Small-Scale Quadruped Robot. In IEEE-RAS International Conference on Robotics and Automation (ICRA). IEEE, 2020.

Thomas Dudzik, Matthew Chignoli, Gerardo Bledt, Bryan Lim, Adam Miller, Donghyun Kim, and Sangbae Kim. Robust Autonomous Navigation of a Small-Scale Quadruped Robot in Real-World Environments. In IEEE-RAS International Conference on Intelligent Robots and System (IROS), 2020.

Donghyun Kim

Institute for Applied Life Sciences