U-AMC

Hi, my name is Jason Kim, and I work on state estimation for legged robots in URobotics. I am currently under the supervision of Prof. Yonggun Cho at Inha University. My focus is building estimators that are geometrically consistent and robust in the real world. By combining sensor-fusion framework based on invariant EKF formulations and GP-style motion priors—with learning-based modules that handle non-linear IMU behavior and disturbances. I validate the system on realistic SLAM datasets and push it to run in real time on embedded hardware. My final goal is field-ready state estimation that lets legged robots navigate and collaborate reliably in complex environments, which will lead us to the era of legged-robot-society.