Embedded AI

AI Comes to the Operational Network's Edge

- By Curtis Franklin, Editor-In-Chief

New developments in single-board computers and the chips that power them are bringing the dream of operational artificial intelligence in devices at the network's edge closer to reality than ever before. We're looking at how close we are to widely distributed, on-device intelligence and which trends and developments seem most likely to push us across the goal line.

The integration of Artificial Intelligence (AI) into Internet of Things (IoT) edge devices marks a fundamental shift from centralized cloud computing to distributed intelligence, bringing AI processing closer to data as it's generated and used. This shift, known broadly as Edge AI, is rapidly transforming various industries by enabling devices to process data and make control decisions locally without relying on distant cloud servers or models controlled by third parties.

A SMARTER EDGE

Edge AI is deploying artificial intelligence models directly on edge devices, including sensors, microcontrollers (MCUs), or gateways, enabling them to process data locally. This local processing capability is crucial for enhancing the functionality and efficiency of embedded systems.

The core concepts of Edge AI center on bringing computational power to the network's periphery. This approach offers several significant advantages over traditional client-server or cloud computing. Many of the advantages build on the basic benefits of not shipping data back and forth over long distances, but those benefits show themselves in a number of different ways.