Composite Safety Potential Field for Highway Driving Risk Assessment

TL;DR

This paper addresses the need for real-time highway risk assessment that combines human-like risk perception with objective collision probability. It introduces Composite Safety Potential Field (C-SPF), a two-field framework comprising a Subjective Field (S-field) that models proximity-based risk using probabilistic spacing and a Generalized Gaussian formulation, and an Objective Field (O-field) that quantifies collision risk from relative motions via $r_{o,ij} = P_{ij} T_{ij}$. The S-field is calibrated with abundant two-dimensional spacing data from naturalistic trajectories, while the O-field relies on straightforward motion-based calculations, enabling data-efficient calibration and broad generalization. Across highD data and case studies, C-SPF demonstrates superior explainability for both longitudinal and lateral driver maneuvers—such as braking and lane-change abandonment—compared with baselines like TTCi and RDSI, offering practical potential for pre-collision warnings and ADAS integration. The framework thus provides a human-centric, interpretable risk assessment tool with real-world applicability and clear pathways for future enhancements, including road geometry and speed-variant modeling.

Abstract

In the era of rapid advancements in vehicle safety technologies, driving risk assessment has become a focal point of attention. Technologies such as collision warning systems, advanced driver assistance systems (ADAS), and autonomous driving require driving risks to be evaluated proactively and in real time. To be effective, driving risk assessment metrics must not only accurately identify potential collisions but also exhibit human-like reasoning to enable safe and seamless interactions between vehicles. Existing safety potential field models assess driving risks by considering both objective and subjective safety factors. However, their practical applicability in real-world risk assessment tasks is limited. These models are often challenging to calibrate due to the arbitrary nature of their structures, and calibration can be inefficient because of the scarcity of accident statistics. Additionally, they struggle to generalize across both longitudinal and lateral risks. To address these challenges, we propose a composite safety potential field framework, namely C-SPF, involving a subjective field to capture drivers' risk perception about spatial proximity and an objective field to quantify the imminent collision probability, to comprehensively evaluate driving risks. The C-SPF is calibrated using abundant two-dimensional spacing data from trajectory datasets, enabling it to effectively capture drivers' proximity risk perception and provide a more realistic explanation of driving behaviors. Analysis of a naturalistic driving dataset demonstrates that the C-SPF can capture both longitudinal and lateral risks that trigger drivers' safety maneuvers. Further case studies highlight the C-SPF's ability to explain lateral driver behaviors, such as abandoning lane changes or adjusting lateral position relative to adjacent vehicles, which are capabilities that existing models fail to achieve.

Figures (14)

• Figure 1: One-dimensional proximity risk distribution based on GGD
• Figure 2: Distribution of velocity and steering angle of highD vehicles
• Figure 3: S-field vehicle proximity risk parameter inference results with bootstrapping sampling. The filling colors indicate the standard deviations obtained from bootstrapping.
• Figure 4: Visualization of S-field with parameters calibrated with highD data
• Figure 5: Visualization of O-field with different inter-vehicle motions