Presentation
Linking Team Situation Awareness Through Verbal Communication Patterns
SessionThursday Poster Session
DescriptionTeam Situation Awareness (TSA), a collective understanding of environmental changes, is critical for effective human-human and human-robot teaming in dynamic settings. Existing TSA measurement techniques, such as the Situation Awareness Global Assessment Technique (SAGAT), cannot continuously capture real-time fluctuations in TSA. Communication patterns offer a promising nonintrusive alternative for continuously assessing TSA through naturally occurring verbal interactions. This study examined the relationship between TSA and communication patterns in a construction setting. A two-worker–one-cobot collaborative pipe task was conducted with 21 dyads (42 participants) in a simulated VR construction site. Participants performed interdependent pipe-installation tasks while responding to dynamic hazards, requiring continuous monitoring to ensure safety, productivity, and quality. Communication was analyzed using a speech act coding framework, while TSA was assessed using SAGAT. TSA scores were decomposed into three dimensions of environment, task, and team, and three SA levels.
Results showed that high-TSA teams generated more level-2 environment-related communication than low-TSA teams. Additionally, high-TSA teams demonstrated more balanced and coupled communication patterns. These findings provide insights into communication-based TSA assessment and support future development of algorithms for tracking TSA and identifying at-risk workers in human-AI teaming environments.
Results showed that high-TSA teams generated more level-2 environment-related communication than low-TSA teams. Additionally, high-TSA teams demonstrated more balanced and coupled communication patterns. These findings provide insights into communication-based TSA assessment and support future development of algorithms for tracking TSA and identifying at-risk workers in human-AI teaming environments.
Event Type
Poster
TimeThursday, October 22nd5:30pm - 6:30pm PDT
Location