Artificial Intelligence in Smart Cities: Technologies, Applications, and Challenges

Abstract

Pressure on transportation system, energy system, health and safety system, environment system, governance system and unprecedented volume of urban data have found a way to increase due to the growing urbanization. Artificial Intelligence (AI) has firmly entrenched itself as an analytical tool for smart cities, and is increasingly used to convert multi-modal data from various sensors and administrative and geospatial data and citizens into predictions, decisions and services. The present review aims at adopting as a topic the use of Artificial Intelligence (AI) technologies and applications in smart cities, dealing with papers published from 2018 to 2025. In this paper, all these technologies are amalgamated in machine learning, deep learning, computer vision, NLP, reinforcement learning, federated learning and edge AI and generative AI-powered digital twins. Applications are considered in a wide range of mobility applications, energy, public safety, environmental monitoring, infrastructure maintenance, municipal governance, and urban planning. They include common implementation challenges, including exposure to privacy risks, cybersecurity risks, interoperability problems, data quality issues, model opacity, algorithmic bias, infrastructure costs, and inequitable citizen participation. The conceptual framework is suggested to be part of the governance process, where the design of the processes for data acquisition, intelligent processing, domain applications and accountability controls all fit within this process. The paper argues that the future of smart cities powered by AI technologies should focus on privacy-centric analytics, human-centric and energy-efficient artificial intelligence, digital twin’s interoperability, and collaboration and citizen participation in decision-making and social value.