Jawad, Abdul (University of California Santa Cruz), Whitehead, Jim (UC Santa Cruz)

Accident Scenario Generation Using Driver Behavior Model

Scheduled for presentation during the Regular Session "Generating driving scenarios II" (FrBT10), Friday, September 27, 2024, 14:50−15:10, Salon 18

2024 IEEE 27th International Conference on Intelligent Transportation Systems (ITSC), September 24- 27, 2024, Edmonton, Canada

This information is tentative and subject to change. Compiled on October 7, 2024

Abstract

Situational awareness is essential for safe driving, yet human drivers are often limited by their perceptual and cognitive capacities, leading to significant road accidents. Autonomous vehicles promise to eliminate humans from driving responsibility and ensure safer travel. Effective testing of autonomous vehicles, especially through scenario-based testing, is critical to their development. Accident scenarios are of particular importance in different phases of scenario-based testing. In this paper, we propose an accident scenario generation method for scenario-based testing using a human driver behavior model, CogMod, that explicitly models human perceptual and cognitive limitations. Through experimentation, we investigate the effects of perceptual and cognitive limitations on accident causation and demonstrate how this can be used in generating diverse accident scenarios.