Toward AI standardization: A triadic human-ai collaboration framework for multi-level autonomous mobility

The goal of the current study is to introduce a triadic human-AI collaboration framework that could be applied in transportation systems such as automated vehicles, micromobility systems, and vehicle teleoperation. Previous standards, such as SAE Levels of Automation, have focused on defining automation levels based on who controls the vehicle. However, it is still not clear how human users and AI should collaborate in real time, especially in dynamic driving contexts where roles can shift frequently. To fill this gap, this study proposed a triadic human-AI collaboration framework with three AI roles: Advisor, Co-Pilot, and Guardian. These roles can dynamically adapt to human needs based on real-time data, such as mental states and environmental conditions. The Advisor AI offers informational support without direct intervention. The Co-Pilot AI provides partial intervention when needed, with the goal of sharing control with humans. The Guardian AI performs emergency overrides if necessary. The use cases for these AI roles in micromobility devices, such as e-scooters, are presented to demonstrate how these roles can influence user preferences and trust. Overall, the study takes a first step toward a universal role-based collaborative framework for AI standardization and explores how AI technologies can be embedded in future transportation systems while considering human interactions.