Trust-Aware Embodied Bayesian Persuasion for Mixed-Autonomy

2025-09-18

AI Summary

This study addresses trust degradation in mixed autonomy traffic—arising from strategic interactions between autonomous vehicles (AVs) and human-driven vehicles (HVs)—by proposing a trust-aware embodied Bayesian persuasion framework. To ensure transparency and non-manipulativeness, the framework introduces two theoretical guarantees: the Trust Threshold Theorem and the Optimal Signal Magnitude Theorem. It employs physically grounded, interpretable signals—such as subtle forward nudges—in continuous action space to model human-AV communication at intersections. Extensive simulation experiments in mixed-traffic scenarios demonstrate that the approach significantly enhances human drivers’ situational caution, eliminates collisions entirely, and improves overall traffic throughput. The core contribution lies in the first integration of embodiment, interpretability, and Bayesian persuasion theory for human–AV coordination—yielding a solution that is simultaneously safe, trustworthy, and traffic-efficient.

Technology Category

• Humans and AI: Human-Aware Planning and Behavior Prediction
• Philosophy and Ethics of AI: Safety, Robustness & Trustworthiness
• Reasoning under Uncertainty: Sequential Decision Making

Application Category

• Responsible Web: Machine-in-the-loop, human agency and autonomy
• Economics, Online Markets and Human Computation: Trust and reliance of crowd workers and data experts on GenAI
• User Modeling, Personalization and Recommendation: Accountability, Transparency, and Ethics for personalization

Abstract

Safe and efficient interaction between autonomous vehicles (AVs) and human-driven vehicles (HVs) is a critical challenge for future transportation systems. While game-theoretic models capture how AVs influence HVs, they often suffer from a long-term decay of influence and can be perceived as manipulative, eroding the human's trust. This can paradoxically lead to riskier human driving behavior over repeated interactions. In this paper, we address this challenge by proposing the Trust-Aware Embodied Bayesian Persuasion (TA-EBP) framework. Our work makes three key contributions: First, we apply Bayesian persuasion to model communication at traffic intersections, offering a transparent alternative to traditional game-theoretic models. Second, we introduce a trust parameter to the persuasion framework, deriving a theorem for the minimum trust level required for influence. Finally, we ground the abstract signals of Bayesian persuasion theory into a continuous, physically meaningful action space, deriving a second theorem for the optimal signal magnitude, realized as an AV's forward nudge. Additionally, we validate our framework in a mixed-autonomy traffic simulation, demonstrating that TA-EBP successfully persuades HVs to drive more cautiously, eliminating collisions and improving traffic flow compared to baselines that either ignore trust or lack communication. Our work provides a transparent and non-strategic framework for influence in human-robot interaction, enhancing both safety and efficiency.

Problem

• Modeling communication between autonomous and human-driven vehicles at intersections
• Addressing trust decay and manipulation in game-theoretic vehicle interactions
• Translating abstract Bayesian signals into continuous physical action space

Innovation

• Bayesian persuasion models communication at intersections
• Introduces trust parameter for minimum influence level
• Grounds signals into continuous physical action space