Role of AI in Edge Computing

Abstract

Edge computing has developed into an essential approach for addressing the shortcomings of cloud-based architectures by facilitating data processing near the sources of that data. The integration of Artificial Intelligence (AI) greatly improves edge computing by allowing for real-time data analysis, smart decision-making, and the operation of autonomous systems in environments with limited latency and bandwidth. This review article examines the significance of AI within edge computing, focusing on its architectures, supporting technologies, challenges, and various applications. A practical case study from the industry is presented to showcase the efficacy of AI-driven edge systems. Edge computing offers a crucial solution to the limitations of traditional cloud architectures by processing data closer to its source. The power of Edge computing is significantly enhanced through the incorporation of Artificial Intelligence (AI), enabling real-time data analysis, intelligent decision-making, and the operation of autonomous systems in bandwidth- and latency-constrained environments. This review article investigates the vital role of AI in edge computing. Specifically, it explores the foundational architectures, enabling technologies, inherent challenges, and diverse applications of AI-driven edge systems. An industry-based case study is also included to demonstrate the practical effectiveness of these systems.

Introduction

The rapid growth of IoT devices has led to massive data generation at the network edge, creating challenges for traditional cloud computing, including high latency, bandwidth limitations, and privacy concerns. Edge computing addresses these by processing data near its source, enabling low-latency, bandwidth-efficient, and privacy-preserving computation.

Edge AI is the integration of Artificial Intelligence (AI) with edge computing, allowing real-time, autonomous decision-making without reliance on centralized clouds. It finds applications in:

• Smart manufacturing: predictive maintenance, quality control, robotic coordination.

• Healthcare monitoring: real-time biometric analysis for critical event detection.

• Autonomous vehicles: instant sensor data processing for navigation and safety.

• Smart cities: traffic optimization, energy management, public safety monitoring.

Key Roles of AI at the Edge:

1. Real-time processing and inference: Local analysis of sensor and video data reduces latency, enabling immediate responses.

2. Resource optimization: AI schedules computing, memory, and bandwidth dynamically to prevent bottlenecks and improve energy efficiency.

3. Enhanced security: Local anomaly detection and intrusion prevention mitigate threats quickly.

4. Context-aware services: Personalized insights and actions based on localized data (e.g., adaptive traffic lights, predictive maintenance).

Enabling Technologies for Edge AI:

• Model optimization: Pruning, quantization, and knowledge distillation reduce AI model size and computational needs for edge deployment.

• Lightweight architectures: MobileNet, ShuffleNet, EfficientNet, and SqueezeNet balance performance with efficiency.

• Hardware accelerators: NPUs, GPUs, TPUs, and FPGAs enable fast, energy-efficient inference on edge devices.

• Federated learning (FL): Allows distributed model training while preserving data privacy and minimizing communication overhead.

• Edge orchestration platforms: Manage AI model deployment, monitoring, scaling, and heterogeneous edge hardware effectively.

Impact: Edge AI transforms simple edge devices into intelligent, autonomous nodes capable of real-time analytics, secure operations, and personalized, context-aware services, complementing cloud computing rather than replacing it.

Conclusion

AI is vital for maximizing edge computing\'s potential, enabling intelligent, low-latency, and privacy-aware applications. This review examined AI\'s role in edge computing, enabling technologies, challenges, and an industrial case study. Edge AI\'s continued evolution will significantly impact next-generation intelligent systems.

References

REFERENCES [1] W. Shi et al., “Edge computing: Vision and challenges,” IEEE Internet of Things Journal, vol. 3, no. 5, pp. 637–646, 2016. [2] M. Satyanarayanan, “The emergence of edge computing,” IEEE Computer, vol. 50, no. 1, pp. 30–39, 2017. [3] Y. LeCun, Y. Bengio, and G. Hinton, “Deep learning,” Nature, vol. 521, pp. 436–444, 2015. [4] C. Zhang et al., “Edge intelligence: Paving the last mile of artificial intelligence,” Proc. IEEE, vol. 107, no. 8, pp. 1738–1762, 2019. [5] S. Han et al., “Deep compression,” ICLR, 2016. [6] A. Howard et al., “MobileNets: Efficient CNNs for mobile vision applications,” arXiv:1704.04861, 2017. [7] B. McMahan et al., “Communication-efficient learning of deep networks from decentralized data,” AISTATS, 2017. [8] X. Chen and X. Ran, “Deep learning with edge computing,” Proc. IEEE, vol. 107, no. 8, pp. 1655–1674, 2019. [9] M. Chiang et al., “Fog and IoT: An overview of research opportunities,” IEEE Internet of Things Journal, vol. 3, no. 6, pp. 854–864, 2017. [10] S. Teerapittayanon et al., “Distributed deep neural networks over the cloud, the edge and end devices,” IEEE ICDCS, 2017.

Copyright

Copyright © 2026 Ms. Mohini Suresh Fulzele, Prof. Roshni Sonsare, Mr. Sandeep Ramesh Sonaskar, Ms Neema Ukani. This 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.