AI-POWERED ANALYSIS

Cooperative In-Vehicle Intelligent Voice Agents in Level 4 Automation

Level 4 automated vehicles (AVs) are entering real-world urban traffic, where they must coexist and coordinate with human-driven vehicles. Although prior research has addressed automated driving scenarios, it largely focuses on isolated human-vehicle interaction or single-exposure studies, offering limited insight into cooperative behavior in mixed-traffic and multi-driver settings. This dissertation investigates voice-based interaction as a mechanism for communicating AV intent during cooperative driving situations. Co-operation is framed as a social and temporal process shaped by how AV passengers interpret voice-based communication and how this understanding manifests in vehicle behavior that surrounding human drivers must anticipate. Using a user-centered, mixed-methods approach, the research combines multi-driver driving simulation with a longitudinal simulator study. The work examines perceived safety, interpretation of cooperative intent, and coordination in mixed traffic, as well as how these processes evolve over repeated interaction. The dissertation contributes empirical evidence and methodological guidance for cooperative interaction in multi-driver simulation.