Trajectory-based indicators to determine the local character of intersection conflicts: A micro-spatial analysis.

Area of Science:

• Traffic Engineering and Road Safety
• Transportation Data Analysis
• Behavioral Analysis in Transportation

Background:

• Collecting real-world road user behavior data for safety assessment is challenging due to data protection and logistical issues.
• Existing methods struggle with rapid, portable data collection for road safety research.
• The Mobility Observation Box (MOB) offers a flexible solution for data collection and video analysis.

Purpose of the Study:

• To advance road safety research by analyzing naturalistic video data using the MOB.
• To model the likelihood of critical conflict occurrence using random parameters binary modeling.
• To identify factors influencing conflict angles in critical situations using Gaussian generalized additive spatial modeling.

Main Methods:

• Utilized over 51 hours of naturalistic video data from a busy Vienna intersection.
• Employed object detection and tracking to derive movement trajectories and Surrogate Safety Measures (SSMs) like PET, TCA, and TTC.
• Applied random parameters binary modeling and Gaussian generalized additive spatial modeling to analyze conflict occurrence and influencing factors.

Main Results:

• Identified specific speed, acceleration, and road user type effects on conflict likelihood and angles.
• Cars were found to be the benchmark for safety, with most other road users being less likely to be involved in critical conflicts.
• Cyclists in a leading role were an exception, showing a higher likelihood of involvement in safety-critical conflicts.
• Kinematic parameters (speed, acceleration, deceleration), interaction duration, and leading road user position significantly influenced conflict angles.

Conclusions:

• The MOB system effectively enables the derivation of quantitative road safety indicators from real-world behavior data.
• Road user type significantly impacts safety-critical conflict involvement, with cars serving as a baseline.
• Understanding micro-spatial and kinematic factors is crucial for analyzing and mitigating critical conflict angles at intersections.