Interaction Lab · Dr. Maja Matarić Socially Assistive Robots for Personalized Cognitive Behavioral Therapy An NIH-funded project exploring how Blossom, an in-home socially assistive robot, can support college students with elevated anxiety in practicing cognitive behavioral therapy (CBT) exercises between therapy sessions. The project investigates how personalization through direct user feedback and cues from natural interaction can help Blossom adapt to individual preferences and evolving needs. The goal is to sustain engagement with CBT practice and build supportive robot–user relationships that promote well-being over time. Current work focuses on longitudinal interaction with Blossom over a 6 week period.
Rice University · University of Southern California Cross-cultural Personalization of Robot and AI-mediated Well-being Support This project investigates how individual cultural orientations and national contexts shape acceptance of and preferences for well-being support through AI chatbots and social robots. It explores preferences for support types, such as peer support and psychoeducation, alongside how that support is communicated. Building on an initial survey with more than 200 U.S. participants, current work focuses on an in-person study in the Global South to understand preferences for interaction styles and develop predictive models that inform personalization across cultural and linguistic contexts.
University of Washington · Dr. Maya Cakmak Enabling End Users to Program Robots with Shared Autonomy Selected as an NSF REU Fellow ($9,000) at the University of Washington to contribute to an Amazon-funded project on shared autonomy in human-robot interaction in a home environment. I developed a web-based teleoperation and shared autonomy system for the Hello-Robot Stretch, enabling users to specify control handoffs via end-user programming in long-horizon tasks. I evaluated system performance across 10 in-home tasks, identifying technical limitations, and led a 24-participant user study with novice and experienced users to analyze human-in-the-loop decisions and inform future approaches to personalized robot learning.
Rice University · Dr. Vaibhav Unhelkar · Honors thesis Caregivers’ Perceptions of Robot Capabilities and Errors Honors thesis on how informal and formal caregivers perceive and respond to robot errors, guiding the design of intuitive training methods and trust-aware robotic systems. I designed and proposed the methodology, analytic plan, and survey measures to examine error perception, recovery strategies, and trust outcomes across caregiver groups in both home and clinical care contexts. I implemented robot-assisted caregiving tasks with error injection via YOLO training and state-based task control.
Computer-Human Interaction Lab · Dr. Michael Byrne Computational Modeling for Ballot Design Implemented and conducted 7 experiments in Python with 240 participants on visual grouping, analyzed the data, and used it to enhance a computational model for evaluating voting ballot design. I selected study objectives and research methods to build on those findings and establish a visual grouping hierarchy.
Computer-Human Interaction Lab · Dr. Michael Byrne Effect of AI on Information Gathering Supervised a 60-participant study comparing AI-based and traditional web-based information gathering, evaluating task performance and cognitive strategies in health and consumer contexts.