Junhyeok Ahn

I am a software engineer at Boston Dynamics. I received my Ph.D. in Mechanical Engineering from The University of Texas at Austin under the supervision of Dr. Luis Sentis. My research interests lie broadly in planning, control, optimization, and machine learning algorithms for legged robots. At Boston Dynamics, I am currently working on a behavior foundation model for Atlas.

Email  /  CV  /  Google Scholar  /  LinkedIn  /  GitHub

Large behavior model that lets Atlas use vision, proprioception, and language cues to perform long-horizon tasks combining locomotion + whole-body manipulation

Control and evaluation of Draco3 humanoid robot.

Learning-based safety verification tool for legged robots.

Novel network representation for hybrid dynamical systems.

Robust gain adaptation method in whole-body controller.

Trajectory optimization framework with a centroidal inertia network and whole-body controller with an implicit task hierarchy for humanoid robots.

Trajectory generation based on sampling-based reachability analysis and optimal control.

Time-to-velocity reversal planner and whole-body locomotion controller for passive-ankle bipeds.

Safe and efficient learning of locomotion policy combining a capturability, control barrier function, and hierarchical reinforcement learning.

Design, control, and evaluation of a new human-scaled biped robot with liquid cooled viscoelastic actuators.

Whole-body controller that considers centroidal momentum dynamics, operational task priorities, contact reaction forces, and internal force constraints.

Kino-dynamic RRT planning for humanoids in maze environemnts.

Design, control, and evaluation of a new viscoelastic liquid cooled actuator technology.

Design, control, and evaluation of a new viscoelastic liquid cooled actuator technology.

Human-aware control policy that incorporates social cognitive theory and model predictive control with mixed integer constraints.

Website source from Jon Barron and Chaitanya Malaviya