Concept of an enhanced V2X pedestrian collision avoidance system with a cost function-based pedestrian model

Area of Science:

• * Automotive Safety Engineering
• * Intelligent Transportation Systems
• * Human-Behavior Modeling

Background:

• * Current collision avoidance systems rely on simplistic pedestrian behavior models (constant acceleration/velocity).
• * Limited predictive accuracy hinders effective collision mitigation.
• * Emerging sensor data from smart devices offers potential for enhanced pedestrian modeling.

Purpose of the Study:

• * To develop and implement a Vehicle-to-Everything (V2X) pedestrian collision avoidance system.
• * To integrate novel information sources for improved pedestrian behavior prediction.
• * To address challenges associated with new data integration in V2X systems.

Main Methods:

• * Conducted a literature review of existing collision avoidance systems, pedestrian behavior models, and traffic simulations.
• * Analyzed typical pedestrian patterns to identify key behavior prediction parameters.
• * Developed and implemented a V2X system concept based on derived requirements.

Main Results:

• * Analysis revealed the complexity of predicting pedestrian behavior in traffic.
• * A V2X collision avoidance system concept was developed, utilizing a cost function for near-future pedestrian presence prediction.
• * The implemented concept addresses privacy, localization, and prediction inaccuracies.

Conclusions:

• * An enhanced V2X pedestrian collision avoidance system concept was successfully developed.
• * The system effectively uses smart device data to improve near-future pedestrian presence prediction.
• * Key challenges related to data privacy and system accuracy were considered.