Towards LLM-Integrated Storytelling Robot to Improve College Student Mental Health

Tam Do, Chi Vo, Duy Linh Ha, Felix Mendez, Suyog Bam, Zhao Han

Keywords: LLM, storytelling robot, robotics, mental health

Abstract: The prevalence of mental health challenges among college students calls for innovative and accessible support systems. Traditional mental health resources are underutilized due to barriers like financial constraints and discomfort in seeking help. To mitigate these issues, recent research has explored robots for mental health interventions, but their empathy-evoking storytelling capabilities remain underdeveloped. In this work, we present preliminary results on the implementation of an LLM-integrated robot to improve student mental well-being through interactive storytelling. Unlike static approaches, the robot not only generates and adapts narratives based on user input in real time but also conveys emotions through facial expressions and gestures that correspond with the stories’ content. As future work, we are conducting a longitudinal study by deploying in student dormitories over one-week spans. Various validated psychological scales on psychological well-being and college-specific well-being factors will be used. Future results will offer valuable insights into how LLM-enabled interactive storytelling robots can improve student well-being.

Submission Number: 30