Daehwan Yoo
Research Mentor(s): Christin Salley
Mentor Department: Michigan Institute for Data Science (MIDAS)
Authors: Daehwan Yoo, Christin Salley, Lu Wang
Session: Session 2 (10:00am – 10:50am)
Presentation Type: Poster 29
Abstract
Emergency telecommunicators are critical first responders who navigate high-pressure situations to ensure timely and effective emergency responses. Despite their pivotal role, research often overlooks the unique occupational stressors they face compared to other emergency personnel, such as paramedics or firefighters. Studies reveal that emergency telecommunicators frequently experience elevated levels of stress due to time-sensitive decision-making, limited control over outcomes, and insufficient organizational support, contributing to mental health challenges. While advancements in artificial intelligence (AI) have demonstrated potential in enhancing workplace efficiency and mental health support, there is limited research applying AI tools to analyze the emotional and psychological well-being of emergency telecommunicators. Current studies predominantly rely on qualitative interviews or self-reported data, which, while valuable, lack the objectivity and scalability required for real-time interventions. Furthermore, the integration of multimodal AI approaches—combining natural language processing (NLP) for transcript analysis and computer vision (CV) for facial emotion detection—remains underexplored in this field. This research aims to address these gaps by developing an AI-driven framework to analyze dispatcher stress levels using transcripts and audio recordings. This study seeks to compare various AI NLP techniques, such as sentiment analysis, topic modeling, and transformer-based models that can be utilized to foster improved mental health support and operational efficiency in this space. The findings aim to demonstrate the transformative potential of AI in critical, high-stress professions, while contributing to ethical AI development and emergency service innovation.
