This research introduces a novel framework that leverages Large Language Models to create more realistic and efficient transportation system simulations.

• Proposes LLM-agent-based modeling that enhances behavioral realism in transportation simulations
• Outlines how LLM agents can simulate diverse human decision-making processes in traffic scenarios
• Addresses resource constraints of traditional agent-based transportation models
• Demonstrates potential for more accurate traffic forecasting and system optimization

This framework represents a significant engineering advancement by combining AI capabilities with transportation planning, potentially revolutionizing how urban mobility solutions are designed and evaluated.

Toward LLM-Agent-Based Modeling of Transportation Systems: A Conceptual Framework