A Systematic Review of AloT and Large Language Model Applications in Smart Urban Transport

Rapid urbanization and the limited ability of transportation infrastructure to expand have intensified traffic congestion, shortages in public transit services, and rising air pollution in cities around the world. These conditions increasingly reduce the efficiency, sustainability, and overall quality of urban life. This study presents a systematic review of how the combined use of the Artificial Intelligence of Things and Large Language Models can help address these challenges by supporting more adaptive and data driven approaches to urban mobility management. The integration of connected vehicles, IoT based sensing systems, and LLM supported reasoning enables continuous information exchange that can improve congestion prediction, reduce fuel use, and lower environmental impacts. Such systems can also identify abnormal patterns, monitor urban conditions in real time, and escalate unpredictable or high risk situations for human oversight when needed. The review identifies several key barriers, including the absence of alternative routes in highly congested corridors, which reflects structural limitations in many cities, as well as a shortage of specialists who can translate AI generated insights into practical operational strategies. By synthesizing findings from recent research, this paper provides a structured framework for AI supported smart urban transport, highlights major limitations and research gaps, offers insights for improving LLM performance in smart city environments, and outlines the emerging role of self learning LLM based networks in future mobility systems.

کلیدواژه ها:

AIOT ، Large Language Models ، Smart Transport ، Intelligent Transportation Systems ، Smart Cities ، Urban Mobility