ADAPTIVE EDGE INTELLIGENCE FOR SMART CITIES: A MACHINE LEARNING FRAMEWORK FOR REAL-TIME URBAN DATA PROCESSING

Smart cities generate enormous amounts of data through Internet of Things (IoT) devices, surveillance cameras, environmental sensors, and connected transportation systems. Processing this continuously growing data using traditional cloud computing introduces communication delays, bandwidth limitations, and higher operational costs. Edge intelligence combines edge computing with artificial intelligence to process data closer to its source, enabling faster decision-making while reducing network congestion. This paper presents an adaptive machine learning framework that performs intelligent data processing at edge nodes within smart city infrastructures. The proposed framework dynamically allocates computational resources according to workload characteristics while employing lightweight machine learning models for real-time inference. Experimental analysis demonstrates improvements in latency, bandwidth utilization, and energy efficiency compared with conventional cloud-centric approaches. The proposed architecture supports scalable deployment for traffic management, environmental monitoring, public safety, and healthcare applications. Results indicate that integrating adaptive edge intelligence significantly enhances smart city performance while maintaining data privacy and system reliability.