Analysis of AI and IoT-based Innovative Urban Predictive Intelligence Applications in Smart Cities

Abstract

Predictive intelligence in smart cities is one of the prevalent research domain for artificial intelligence and internet of things. However, there is no sufficient study on the mapping of its scholarly growth like thematic evolution and future trajectories. This study utilized bibliometric analysis on Scopus indexed documents from 2020 to 2025, applying PRISMA workflow, multi domain Boolean search strategy and analytical tools and science mapping techniques included in bibliometrix R-package and VOSviewer. Publication trends, citation structures, keyword co-occurrence mapping, co-authorship networks, thematic evolution and conceptual structure mapping were investigated. More or less 17,557 scopus-indexed publications were analyzed showing an annual growth rate of 16.42% where India, China and the United States lead global contributions. Predominant journals were found in the IEEE Internet of Things and IEEE Transactions on Intelligent Transportation Systems. Keyword co-occurrence revealed machine learning, deep learning, IoT, edge computing and smart city as the core thematic anchors and well-developed research areas. This study confirms that Artificial Intelligence and Internet of Things based forecasting in smart cities is a rapidly advancing research field. A roadmap for scholarly inquiry and policy-relevant applications can be explored where critical gaps and opportunities in areas such as federated learning, digital twins and cross-domain predictive analytics were highlighted.