Intelligent IoT Frameworks for Smart Cities: Challenges and Opportunities

Abstract

Smart cities represent one of the most complex and consequential application domains for intelligent Internet of Things (IoT) frameworks, requiring
the integration of heterogeneous sensor networks, real-time data processing pipelines, AI-driven decision-support systems, and citizen-facing
services across urban infrastructure at unprecedented scale. Despite growing global investment in smart city initiatives, systematic frameworks for
evaluating the architectural, technical, and governance challenges of intelligent IoT deployment in urban environments remain fragmented. This
review presents a comprehensive analysis of intelligent IoT frameworks for smart cities, synthesising 187 peer-reviewed studies published between
2015 and 2024 across five smart city domains: intelligent transportation, smart energy grids, environmental monitoring, public safety, and e-health. A
structured taxonomy of 14 architectural patterns, 8 interoperability challenges, and 6 AI integration opportunities is developed. Key findings indicate
that data heterogeneity (cited in 78.6% of studies), real-time processing latency (71.4%), and privacy-security trade-offs (84.2%) constitute the three
dominant challenges, while federated edge AI (opportunity score 4.7/5), digital twin integration (4.5/5), and semantic interoperability (4.3/5) emerge
as the highest-impact opportunities. A roadmap for next-generation intelligent IoT smart city frameworks is proposed, identifying critical research
gaps and standardisation priorities for 2025–2030.