DEEP LEARNING FOR EDGE COMPUTING IN INDIAN SMART CITIES : REAL TIME ANLYTICS ON LOW POWER DEVICES

Authors

  • 1ANJURI AJAY, 2 S.K.ALISHA Author

DOI:

https://doi.org/10.64751/

Abstract

The rapid growth of smart city technologies has enabled virtual monitoring systems for applications such as traffic management, healthcare, agriculture, and home automation. These systems rely on low-power sensors that continuously collect environmental data and transmit it to cloud servers for processing. However, traditional cloud-based architectures face significant challenges, including high latency, increased bandwidth usage, and potential data privacy risks. To address these issues, this project proposes a deep learningbased edge computing framework that performs real-time analytics on low-power devices within Indian smart city environments. In the proposed system, edge servers are deployed closer to data sources and are equipped with sufficient computational resources to process sensor data locally using deep learning models such as YOLO (You Only Look Once) for object detection. The system is designed and implemented as a simulation where traffic videos are processed at the edge server to detect and count vehicles. Instead of transmitting raw video data, only processed results are sent to the cloud, significantly reducing latency and bandwidth consumption while enhancing data privacy. Experimental results demonstrate that transmitting processed data from the edge reduces latency compared to sending complete data to the cloud. The system achieves high prediction accuracy and efficient real-time performance. Overall, the proposed approach highlights the effectiveness of combining edge computing with deep learning to enable faster, secure, and scalable smart city applications.

Downloads

Published

2026-04-08

How to Cite

1ANJURI AJAY, 2 S.K.ALISHA. (2026). DEEP LEARNING FOR EDGE COMPUTING IN INDIAN SMART CITIES : REAL TIME ANLYTICS ON LOW POWER DEVICES. International Journal of Data Science and IoT Management System, 5(2), 1988-1996. https://doi.org/10.64751/