Copyright © 2026 Authors retain the copyright of this article. This article is an open access article distributed under the Creative Commons Attribution License which permits unrestricted use, distribution, and reproduction in any medium, provided the original work is properly cited.

@article{195305,
        author = {Gollapaka Harshitha and Harkala Chandhana and Thalagampala Akshaya and Baikadi Archana and Y.Greeshma and S.Shiva Prasad},
        title = {Edge-Intelligent Deep Learning Framework for Real-Time Urban Analytics on Resource-Constrained Devices},
        journal = {International Journal of Innovative Research in Technology},
        year = {2026},
        volume = {12},
        number = {10},
        pages = {7854-7865},
        issn = {2349-6002},
        url = {https://ijirt.org/article?manuscript=195305},
        abstract = {Modern urban infrastructures continuously generate large volumes of visual and sensory data that must be analyzed without delay to support effective operational decision-making. Centralized cloud-based analytics approaches often introduce communication latency, consume excessive network bandwidth, and raise data privacy concerns. This paper proposes an edge-intelligent deep learning framework that enables real-time analytics directly on low-power computing devices. Lightweight convolutional neural networks are optimized using parameter reduction and numerical precision scaling techniques to ensure efficient execution on resource-constrained hardware. The proposed system is validated using traffic monitoring scenarios, where localized inference demonstrates reduced processing latency, lower energy consumption, and stable performance compared with cloud-dependent execution. Experimental results confirm that edge-based intelligence is a practical, scalable, and privacy-aware solution for latency-sensitive urban analytics. In addition, the framework emphasizes decentralized decision-making by minimizing reliance on continuous cloud connectivity. The experimental evaluation highlights consistent inference performance under varying network conditions, demonstrating robustness in real-world deployments.},
        keywords = {Edge Intelligence, On-Device Deep Learning, Urban Data Analytics, Low-Latency Inference, Resource-Constrained AI, Edge AI Optimization, Decentralized Analytics, Real-Time Inference Systems, Lightweight Neural Networks.},
        month = {March},
        }