Integrating Naturalistic Insights in Objective Multi-Vehicle Safety Framework

2024-08-19

2024 IEEE 27th International Conference on Intelligent Transportation Systems (ITSC)

AI Summary

Addressing the challenge of autonomous vehicle safety assessment in mixed traffic—where both objective risk metrics and human subjective perception must be jointly considered—this paper proposes a multi-vehicle cooperative safety assessment framework. Methodologically: (1) it integrates multiple objective safety indicators, including Time-to-Collision (TTC), Post-Encroachment Time (PET), and Deceleration Rate to Avoid Crash (DRAC); (2) it establishes a full-surround vehicle cooperative modeling mechanism, overcoming the limitations of pairwise vehicle assessment; and (3) it develops a parametric safety perception calibration model grounded in naturalistic driving data, enabling the first tunable coupling between objective safety metrics and subjective perception. Empirical evaluation in highway car-following scenarios demonstrates that multi-metric fusion significantly improves assessment accuracy, while full-surround modeling further enhances performance—validating the framework’s effectiveness and robustness on real-world data.

Technology Category

• Natural Language Processing: Safety and Robustness
• Intelligent Robots: Multimodal Perception & Sensor Fusion
• Planning, Routing, and Scheduling: Model-Based Reasoning

Application Category

• Security and Privacy: Large-scale security measurements
• Search and Retrieval-Augmented AI: Web evaluation methodologies and metrics
• Systems and Infrastructure for Web, Mobile and WoT: Web performance, measurement, and characterization

Abstract

As autonomous vehicle technology advances, the precise assessment of safety in complex traffic scenarios becomes crucial, especially in mixed-vehicle environments where human perception of safety must be taken into account. This paper presents a framework designed for assessing traffic safety in multi-vehicle situations, facilitating the simultaneous utilization of diverse objective safety metrics. Additionally, it allows the integration of subjective perception of safety by adjusting model parameters. The framework was applied to evaluate various model configurations in car-following scenarios on a highway, utilizing naturalistic driving datasets. The evaluation of the model showed an outstanding performance, particularly when integrating multiple objective safety measures. Furthermore, the performance was significantly enhanced when considering all surrounding vehicles.

Problem

• Assess safety in mixed autonomous-human traffic scenarios
• Integrate objective and subjective safety metrics
• Evaluate multi-vehicle safety in car-following situations

Innovation

• Integrates diverse objective safety metrics
• Adjusts model for subjective safety perception
• Uses naturalistic driving datasets evaluation