Modeling realistic human behavior using generative agents in a multimodal transport system: Software architecture and Application to Toulouse

2025-10-22

AI Summary

This study addresses the challenges of modeling human mobility behavior and generating personalized travel solutions in multimodal transportation systems. We propose the first framework integrating large language model (LLM)-driven generative agents into agent-based traffic simulation. Methodologically, the framework synergizes the GAMA simulation platform, GTFS transit data, and the OpenTripPlanner routing engine, enabling agents to autonomously learn and evolve travel decisions within dynamic urban environments. Unlike static rule-based or statistical models, it supports the emergence of context-aware, stable individual travel habits over time. Evaluated in a one-month simulation for Toulouse, France, the framework demonstrates high behavioral fidelity and robust personalization capability. Results confirm enhanced ecological validity in mobility behavior modeling and improved utility for policy analysis and scenario forecasting.

Technology Category

• Multiagent Systems: Agent-Based Simulation and Emergent Behavior
• Planning, Routing, and Scheduling: Planning with Language Models
• Search and Optimization: Sampling/Simulation-based Search

Application Category

• User Modeling, Personalization and Recommendation: Large Language Models (LLM) for user modeling and recommendation
• Semantics and Knowledge: Data modeling to support human-machine intelligence, including LLMs agents, intelligent system behavior, explanations, and user-friendly interactions
• Social Networks and Social Media: Generative AI / large language models and their impact on social systems

Abstract

Modeling realistic human behaviour to understand people's mode choices in order to propose personalised mobility solutions remains challenging. This paper presents an architecture for modeling realistic human mobility behavior in complex multimodal transport systems, demonstrated through a case study in Toulouse, France. We apply Large Language Models (LLMs) within an agent-based simulation to capture decision-making in a real urban setting. The framework integrates the GAMA simulation platform with an LLM-based generative agent, along with General Transit Feed Specification (GTFS) data for public transport, and OpenTripPlanner for multimodal routing. GAMA platform models the interactive transport environment, providing visualization and dynamic agent interactions while eliminating the need to construct the simulation environment from scratch. This design enables a stronger focus on developing generative agents and evaluating their performance in transport decision-making processes. Over a simulated month, results show that agents not only make context-aware transport decisions but also form habits over time. We conclude that combining LLMs with agent-based simulation offers a promising direction for advancing intelligent transportation systems and personalised multimodal mobility solutions. We also discuss some limitations of this approach and outline future work on scaling to larger regions, integrating real-time data, and refining memory models.

Problem

• Model realistic human mobility behavior in multimodal transport systems
• Apply LLMs to capture decision-making in urban settings
• Develop generative agents for personalized mobility solutions

Innovation

• LLMs integrated with agent-based simulation for transport modeling
• GAMA platform combined with GTFS data and OpenTripPlanner
• Generative agents form context-aware habits over simulated month