[공지사항]을 빙자한 안부와 근황 
Show more

Daily Arxiv

This is a page that curates AI-related papers published worldwide.
All content here is summarized using Google Gemini and operated on a non-profit basis.
Copyright for each paper belongs to the authors and their institutions; please make sure to credit the source when sharing.

GraphTrafficGPT: Enhancing Traffic Management Through Graph-Based AI Agent Coordination

Created by
  • Haebom

Author

Nabil Abdelaziz Ferhat Taleb, Abdolazim Rezaei, Raj Atulkumar Patel, Mehdi Sookhak

Outline

This paper proposes GraphTrafficGPT, a graph-based architecture, to improve the efficiency of intelligent traffic management systems based on large-scale language models (LLMs). Existing chain-based systems (e.g., TrafficGPT) have limitations in applying them to complex real environments due to sequential task execution, high token usage, and low scalability. GraphTrafficGPT uses a graph that represents tasks and their dependencies as nodes and edges to enable parallel execution and dynamic resource allocation. The core is a Brain Agent that decomposes user queries, constructs an optimized dependency graph, and coordinates a network of expert agents for data retrieval, analysis, visualization, and simulation. It efficiently processes interdependent tasks through context-aware token management and concurrent multi-query processing support. Experimental results show that GraphTrafficGPT reduces token consumption by 50.2%, average response delay by 19.0%, and improves concurrent multi-query execution efficiency by up to 23.0% compared to TrafficGPT.

Takeaways, Limitations

Takeaways:
We present a novel graph-based architecture that can significantly improve the efficiency of LLM-based intelligent transportation management systems.
Parallel processing and dynamic resource allocation increase applicability in complex real-world environments.
System performance was improved by reducing token consumption and shortening response delay time.
Improved responsiveness to real-time traffic management through simultaneous multi-query processing.
Limitations:
Additional experiments and verification are needed for the application of the proposed model to real large-scale transportation systems.
There is a lack of detailed description of the design and optimization of Brain Agent.
Evaluation of generalization performance for various types of traffic situations and queries is required.
Consideration must be given to communication overhead and fault handling between agents.
👍