Paper Contents
Abstract
The convergence of artificial intelligence (AI) with edge computing is fundamentally reshaping the landscape of mobile computing. This powerful synergy facilitates real-time decision-making processes, significantly curtails latency, and bolsters user privacy by decentralizing data processing. Such an approach effectively addresses critical bottlenecks inherent in purely centralized cloud computing models, including constrained network bandwidth, delays in communication, and potential privacy breaches. Notable advancements propelling this field include the emergence of Tiny Machine Learning (TinyML), which specializes in adapting complex deep learning models for deployment on resource-constrained microcontrollers, and federated learning, a technique that permits collaborative model training across multiple devices without necessitating the sharing of raw, sensitive data. These methodologies are particularly well-suited for environments where computational and power resources are limited, finding extensive applications in domains such as advanced healthcare monitoring, the development of intelligent urban infrastructures, and the optimization of industrial Internet of Things (IIoT) systems. Despite the immense promise, persistent challenges related to the heterogeneity of edge devices, the current scarcity of robust development and testing frameworks, and evolving security vulnerabilities require ongoing attention. This paper offers a comprehensive examination of the foundational principles, prevalent techniques, architectural designs, diverse applications, and prospective future trajectories of AI operating at the network edge, underscoring its profound and transformative impact across a multitude of sectors.
Copyright
Copyright © 2025 Aprameya C V. This is an open access article distributed under the Creative Commons Attribution License.
