Modeling road user response timing in naturalistic traffic conflicts: A surprise-based framework.

Area of Science:

• Human-Computer Interaction
• Traffic Safety Engineering
• Cognitive Psychology

Background:

• Traditional perception-response time models are inadequate for naturalistic traffic conflicts.
• Existing methods fail to account for situation-dependency and stimulus definition in real-world driving.

Purpose of the Study:

• To present a novel framework for measuring and modeling human response times in naturalistic traffic conflicts.
• To address limitations of traditional response time models in dynamic driving scenarios.

Main Methods:

• Modeled response timing as a belief update process driven by surprising stimuli.
• Developed a heuristic model fit to naturalistic human response data from the SHRP2 dataset.
• Proposed computational implementation using evidence accumulation and machine learning.

Main Results:

• The framework successfully models situation-dependent response timing.
• It resolves the challenge of unambiguously defining stimuli in traffic conflicts.
• Demonstrated applicability to real-world crash and near-crash data.

Conclusions:

• The novel framework provides a robust method for analyzing human response timing in naturalistic driving.
• This approach is applicable to automated driving systems and traffic safety domains.
• The belief update and surprise-driven model offers a more accurate representation of human reactions.