A Unified Probabilistic Approach to Traffic Conflict Detection

TL;DR

The paper addresses the lack of a unified, consistent approach to traffic conflict detection by framing conflicts as context-dependent extreme events and proposing a Bayesian risk model $p(c|s,X)$ that integrates interaction context $\theta$ and proximity distribution parameters $\phi$. It decomposes conflict detection into three learning tasks—context representation, proximity distribution inference via Gaussian Process Regression, and EVT-based conflict probability and intensity evaluation—yielding a data-driven, generalizable framework. Demonstrations on highD and 100-Car NDS data show competitive collision-warning performance relative to traditional surrogates and reveal the framework's ability to capture two-dimensional lane-change conflicts and a long-tailed distribution of conflict intensity. The approach supports consistent safety evaluation, infrastructure assessment, and autonomous-driving applications, while highlighting avenues for representation learning, cross-modal extension, and explicit severity modelling. Overall, the unified probabilistic framework advances robust, context-aware conflict detection with potential to improve road safety analytics and automated-vehicle safety systems.

Abstract

Traffic conflict detection is essential for proactive road safety by identifying potential collisions before they occur. Existing methods rely on surrogate safety measures tailored to specific interactions (e.g., car-following, side-swiping, or path-crossing) and require varying thresholds in different traffic conditions. This variation leads to inconsistencies and limited adaptability of conflict detection in evolving traffic environments. Consequently, a need persists for consistent detection of traffic conflicts across interaction contexts. To address this need, this study proposes a unified probabilistic approach. The proposed approach establishes a unified framework of traffic conflict detection, where traffic conflicts are formulated as context-dependent extreme events of road user interactions. The detection of conflicts is then decomposed into a series of statistical learning tasks: representing interaction contexts, inferring proximity distributions, and assessing extreme collision risk. The unified formulation accommodates diverse hypotheses of traffic conflicts and the learning tasks enable data-driven analysis of factors such as motion states of road users, environment conditions, and participant characteristics. Jointly, this approach supports consistent and comprehensive evaluation of the collision risk emerging in road user interactions. Our experiments using real-world trajectory data show that the approach provides effective collision warnings, generalises across distinct datasets and traffic environments, covers a broad range of conflict types, and captures a long-tailed distribution of conflict intensity. The findings highlight its potential to enhance the safety assessment of traffic infrastructures and policies, improve collision warning systems for autonomous driving, and deepen the understanding of road user behaviour in safety-critical interactions.

Figures (10)

• Figure 1: Illustration examples of context-dependent proximity-based conflict probability.
• Figure 2: Proximity-characterised conflict hierarchy varies in different interaction contexts.
• Figure 3: Interaction spectrum described with context-dependent proximity distribution and extreme value theory.
• Figure 4: The unified framework and statistical learning tasks for conflict detection.
• Figure 5: Evaluation of SVGP training and selection of the model to apply.