SOCIA: An End-to-End Agentic Framework for Automated Cyber-Physical-Social Simulator Generation

2025-05-17

AI Summary

This work addresses the challenges of labor-intensive manual construction, complex cross-domain calibration, and heavy expert dependency in high-fidelity Cyber-Physical-Social (CPS) simulation. We propose the first end-to-end LLM-driven multi-agent simulation generation framework. Centered on a centralized orchestration manager, it coordinates specialized agents to jointly perform requirement understanding, cross-domain (social/physical/cyber) modeling, simulation code generation, execution, and data-feedback-driven closed-loop optimization. Key innovations include domain-adaptive prompting, a simulation orchestration engine, and a multi-source data feedback mechanism. Evaluated on mask-wearing behavior simulation, individual mobility generation, and user modeling tasks, the framework automatically produces scalable, high-fidelity simulators—reducing human intervention by over 70% and achieving simulation accuracy exceeding 92% of human-crafted baselines.

Technology Category

• Cognitive Modeling & Cognitive Systems: Simulating Human Behavior
• Multiagent Systems: Agent-Based Simulation and Emergent Behavior
• Search and Optimization: Sampling/Simulation-based Search

Application Category

• Semantics and Knowledge: Data modeling to support human-machine intelligence, including LLMs agents, intelligent system behavior, explanations, and user-friendly interactions
• Economics, Online Markets and Human Computation: Cost models of using LLMs in production systems
• User Modeling, Personalization and Recommendation: User modeling and simulation for interactive and conversational systems

Abstract

This paper introduces SOCIA (Simulation Orchestration for Cyber-physical-social Intelligence and Agents), a novel end-to-end framework leveraging Large Language Model (LLM)-based multi-agent systems to automate the generation of high-fidelity Cyber-Physical-Social (CPS) simulators. Addressing the challenges of labor-intensive manual simulator development and complex data calibration, SOCIA integrates a centralized orchestration manager that coordinates specialized agents for tasks including data comprehension, code generation, simulation execution, and iterative evaluation-feedback loops. Through empirical evaluations across diverse CPS tasks, such as mask adoption behavior simulation (social), personal mobility generation (physical), and user modeling (cyber), SOCIA demonstrates its ability to produce high-fidelity, scalable simulations with reduced human intervention. These results highlight SOCIA's potential to offer a scalable solution for studying complex CPS phenomena

Problem

• Automates generation of high-fidelity Cyber-Physical-Social simulators
• Reduces labor-intensive manual development and complex data calibration
• Enables scalable simulation of complex CPS phenomena with minimal human intervention

Innovation

• LLM-based multi-agent system for CPS simulation
• Centralized orchestration manager coordinates specialized agents
• Automates high-fidelity simulation generation with feedback loops