Related Experiment Videos
Creating pedestrian crash scenarios in a driving simulator environment
Susan T Chrysler1, Omar Ahmad, Chris W Schwarz
1a Texas A&M Transportation Institute, College Station , Texas.
Traffic Injury Prevention
|June 2, 2015
Summary
Pedestrian safety is critical, as they represent a significant portion of traffic fatalities. This study used driving simulators to analyze driver responses in crash-imminent situations, informing the development of advanced driver assistance systems.
Area of Science:
- Traffic Safety Research
- Human Factors Engineering
- Automotive Safety
Background:
- Pedestrians constitute a disproportionately high percentage of traffic fatalities in the United States and globally.
- Existing research methods for studying driver-pedestrian interactions in critical situations have limitations.
- High-fidelity driving simulators offer a safe and controlled environment for evaluating driver behavior.
Purpose of the Study:
- To examine driver responses to crash-imminent pedestrian situations using a driving simulator.
- To develop and present a methodology for creating realistic pedestrian crash scenarios in a simulation environment.
- To identify key factors influencing driver behavior and decision-making in potential pedestrian crashes.
Main Methods:
- Developed 18 unique crash scenarios based on analysis of real-world pedestrian crash data (trajectories, speeds, behaviors).
- Utilized a high-fidelity, full-motion driving simulator with realistic virtual pedestrian models.
- Manipulated driver expectation and visual attention using intersection/mid-block crossings and environmental cues across three distinct driving environments (residential, rural, urban).
Main Results:
- The study focused on scenarios involving straight vehicle travel, representing the majority of fatalities.
- Iterative testing was employed to refine scenario parameters, particularly pedestrian action triggers.
- Consideration was given to numerous variables including vehicle/pedestrian dynamics, road/environmental conditions, and driver/pedestrian characteristics.
Conclusions:
- Crash analysis data can effectively inform the creation of virtual test scenarios for driver behavior evaluation.
- Driving simulators provide a safe platform for testing advanced driver assistance systems (ADAS) by replicating real-world crash conditions.
- Research findings can guide the development of pedestrian crash avoidance and mitigation systems by identifying driver error patterns and response characteristics (braking vs. steering).