Hae Won Park

Personal Robots
  • Research Scientist

Hae Won Park is a Research Scientist at the Personal Robots Group leading the long-term personalization of interactive AI systems in domains that help human flourishing. She oversees and closely works with students on many projects including early childhood education, healthcare, eldercare,  family interaction, and emotional wellness. Before, she was a PhD student at the Institute of Robotics and Intelligent Machines (IRIM) at Georgia Tech, where Hae Won was a member of the Human-Automation Systems (HumAnS) Laboratory advised by Prof. Ayanna Howard. While doing her PhD, Hae Won co-founded Zyrobotics, a spin-off from Georgia Tech that is licensing the three patents from her research. 

Long-term Personalization of Technology
Hae Won’s work focuses on developing interactive social machines that deeply personalize to their users over a long-term interaction, attending to each individual’s unique needs and goals. Systems she develops are always deployed and tested in real-world, in high societal impact areas that support people’s learning, emotional and social well-being, as well as behavior changes and de… View full description

Hae Won Park is a Research Scientist at the Personal Robots Group leading the long-term personalization of interactive AI systems in domains that help human flourishing. She oversees and closely works with students on many projects including early childhood education, healthcare, eldercare,  family interaction, and emotional wellness. Before, she was a PhD student at the Institute of Robotics and Intelligent Machines (IRIM) at Georgia Tech, where Hae Won was a member of the Human-Automation Systems (HumAnS) Laboratory advised by Prof. Ayanna Howard. While doing her PhD, Hae Won co-founded Zyrobotics, a spin-off from Georgia Tech that is licensing the three patents from her research. 

Long-term Personalization of Technology
Hae Won’s work focuses on developing interactive social machines that deeply personalize to their users over a long-term interaction, attending to each individual’s unique needs and goals. Systems she develops are always deployed and tested in real-world, in high societal impact areas that support people’s learning, emotional and social well-being, as well as behavior changes and decision making. This is achieved by formulating theories of mental processes as computational models, which an agent infers through users’ verbal/nonverbal cues, and a reward given to the agent for the actions it takes in response to user states. The challenge lies in the sparsity of available interaction data. For instance, reinforcement learning with data acquired from an interaction with just a single agent either requires a lot of time, or the state and action space must be significantly simplified. To tackle such problems, Hae Won’s research agenda is centered around developing personal robots and agents that are capable of transfer learning from one interaction to another, as well as utilizing collective knowledge and sensing across intelligent and intuitive interfaces embedded in everyday objects.

 
Personal website: http://haewonpark.com