Hae Won Park
Hae Won Park is a Postdoctoral Associate at the Personal Robots Group at the MIT Media Lab. Before, she was a postdoc at the Institute of Robotics and Intelligent Machines (IRIM) at Georgia Tech, where she was a member of the Human-Automation Systems (HumAnS) Laboratory working with Prof. Ayanna Howard. While doing her postdoc at Geogia Tech, Hae Won was the founding member and the principal engineer of Zyrobotics, a spin-off from Georgia Tech that is licensing the three patents from her PhD research.
Hae Won received her PhD and Master’s degree in 2014 and 2009 respectively from Georgia Tech and Bachelor’s degree from POSTECH in 2006. Before entering graduate school, she was a research scientist at the Cognitive Robotics Center at Korea Institute of Science and Technology (KIST), where she worked with Dr. Sung-Kee Park on multi-modal object register and recognition for visual SLAM.
Hae Won’s research, Inclusive Robotics, focuses on developing accessible and customizable robots for everyone, especially for the vulnerable populations. In her PhD dissertation, She used Bayesian case-based reasoning with neural networks to develop fast task instance retrieval and reuse algorithms for robot task learning from demonstration (LfD). Focusing on the fact that LfD can become prone to user demonstration errors, she modeled such error patterns by clustering task instances and finding the corresponding basis function to reveal skill deficits in a user’s cognitive and physical functions. Her robot learner was deployed in educational/therapeutic settings in which children became the robot’s teacher and provided task demonstrations. In real-time, the robot was able to adapt to different error patterns and change its behavior to maximize intervention benefits.