A Survey on Generative AI in Transportation Systems Management and Operation

The integration of Generative AI (Gen-AI) into Transportation Systems Management and Operations (TSMO) offers transformative potential to address persistent knowledge management challenges. This survey study explores how Gen-AI can revolutionize TSMO through enhanced data integration, automated knowledge extraction, and predictive modeling. By synthesizing diverse datasets, Gen-AI facilitates the creation of unified knowledge repositories, improves real-time decision-making, and supports proactive scenario planning. The study introduces a comprehensive framework for embedding Gen-AI into TSMO workflows, enabling streamlined operational efficiency, cross-agency collaboration, and scalable data-driven strategies. Key applications of Gen-AI in TSMO include adaptive traffic control, crash response planning, and predictive modeling for future traffic scenarios. The study highlights innovative Gen-AI models and techniques, such as generative adversarial networks (GANs), large language models (LLMs), and hybrid augmented intelligence frameworks, which collectively enhance TSMO's capacity for resilience and responsiveness. Despite its transformative potential, Gen-AI adoption faces critical challenges, including ethical considerations, computational constraints, and the need for stakeholder trust. The paper emphasizes the importance of responsible AI development, fairness, and explainability to ensure sustainable adoption. By addressing these barriers, the proposed framework sets the stage for a new era of intelligent, adaptive, and community-centered transportation system management,