Consensus-Aware AV Behavior: Trade-offs Between Safety, Interaction, and Performance in Mixed Urban Traffic

TL;DR

The paper investigates consensus among safety, interaction quality, and traffic performance in mixed urban traffic with automated vehicles, using high-resolution TGSIM trajectory data at signalized intersections. It introduces a three-dimensional consensus framework and a set of metrics (TTC, PET, headways, string stability, and pedestrian hesitation) to quantify trade-offs between AV and HDV behaviors. The study finds full consensus is rare (1.63% of AV–VRU frames) and highlights systematic differences: AVs prioritize safety margins but can slow throughput and reduce interaction clarity, while HDVs show more varied, sometimes more flow-friendly behavior. The work provides empirical guidance and an open-source codebase to inform the design of consensus-aware AV agents intended to improve safety, interaction, and efficiency in real-world mixed-traffic environments.

Abstract

Transportation systems have long been shaped by complexity and heterogeneity, driven by the interdependency of agent actions and traffic outcomes. The deployment of automated vehicles (AVs) in such systems introduces a new challenge: achieving consensus across safety, interaction quality, and traffic performance. In this work, we position consensus as a fundamental property of the traffic system and aim to quantify it. We use high-resolution trajectory data from the Third Generation Simulation (TGSIM) dataset to empirically analyze AV and human-driven vehicle (HDV) behavior at a signalized urban intersection and around vulnerable road users (VRUs). Key metrics, including Time-to-Collision (TTC), Post-Encroachment Time (PET), deceleration patterns, headways, and string stability, are evaluated across the three performance dimensions. Results show that full consensus across safety, interaction, and performance is rare, with only 1.63% of AV-VRU interaction frames meeting all three conditions. These findings highlight the need for AV models that explicitly balance multi-dimensional performance in mixed-traffic environments. Full reproducibility is supported via our open-source codebase on https://github.com/wissamkontar/Consensus-AV-Analysis.

Figures (9)

• Figure 1: Framework for deriving consensus trajectory-derived safety, interaction, and performance metrics.
• Figure 2: Left: Kernel Density Estimates (KDE) of TTC values for AVs and HDVs interacting with VRUs, with a 3s threshold shown. Right: Empirical Cumulative Distribution Function (ECDF) of total time spent with TTC $<$ 3s.
• Figure 3: Distribution of TTC with VRUs by Detection Zone.
• Figure 4: Left: KDE of PET values for AVs and HDVs interacting with VRUs. Right: ECDF of the number of PET $<$ 5s interactions per vehicle.
• Figure 5: Deceleration heatmap for AVs (left) and HDVs (right) during a selected left-turn approach.