Stochastic Modeling of Road Hazards on Intersections and their Effect on Safety of Autonomous Vehicles

2025-06-03

AI Summary

Addressing the challenge of quantifying holistic safety for autonomous vehicles (AVs) encountering road hazards at intersections, this paper proposes an Operational Design Domain (ODD)-driven stochastic probabilistic modeling framework. It is the first to jointly model environmental uncertainty (e.g., dynamic traffic and occlusions) and AV functional deficiencies (e.g., perception and monitoring failures) via fault tree analysis and scenario parameterization, enabling formal derivation of a quantitative, probabilistic safety metric—namely, the fatality probability. The method supports both single- and multi-condition safety assessment under varying ODD constraints and yields verifiable, scalable, and interpretable risk values. Key contributions include: (1) a unified environment-system joint uncertainty model; (2) explicit analytical derivation of fatality probability across diverse operational conditions under ODD constraints; and (3) a theoretically rigorous, engineering-deployable quantitative foundation for AV safety certification and regulatory validation.

Technology Category

• Reasoning under Uncertainty: Uncertainty Representations
• Machine Learning: Calibration & Uncertainty Quantification
• Planning, Routing, and Scheduling: Planning under Uncertainty

Application Category

• Security and Privacy: Large-scale security measurements
• Search and Retrieval-Augmented AI: Web evaluation methodologies and metrics
• Responsible Web: Machine-in-the-loop, human agency and autonomy

Abstract

Autonomous vehicles (AV) look set to become common on our roads within the next few years. However, to achieve the final breakthrough, not only functional progress is required, but also satisfactory safety assurance must be provided. Among those, a question demanding special attention is the need to assess and quantify the overall safety of an AV. Such an assessment must consider on the one hand the imperfections of the AV functionality and on the other hand its interaction with the environment. In a previous paper we presented a model-based approach to AV safety assessment in which we use a probabilistic model to describe road hazards together with the impact on AV safety of imperfect behavior of AV functions, such as safety monitors and perception systems. With this model, we are able to quantify the likelihood of the occurrence of a fatal accident, for a single operating condition. In this paper, we extend the approach and show how the model can deal explicitly with a set of different operating conditions defined in a given ODD.

Problem

• Modeling road hazards at intersections for AV safety
• Assessing safety impact of imperfect AV functions
• Extending model to diverse operating conditions (ODD)

Innovation

• Stochastic modeling of road hazards
• Probabilistic AV safety assessment
• Handling multiple operating conditions