Edge and Cloud Computing in Wearable AI Systems

Abstract

The swift evolution of wearable artificial intelligence (AI) systems has significantly impacted continuous health monitoring, personalized healthcare, and human–machine interaction. At the heart of this revolution is the complementary convergence of edge and cloud computing models, which together facilitate scalable, real-time, and intelligent processing of the massive amounts of physiological and contextual data generated by wearable devices. This chapter presents a thorough review of edge-cloud architectures for wearable AI systems, focusing on their significance in real-time health analytics, adaptive intelligence, and immersive healthcare applications in both physical and virtual spaces. Wearable devices like smartwatches, biosensor-embedded clothing, neural interfaces, and implantable sensors continuously collect multi-modal data such as vital signs, activity patterns, biochemical signals, and environmental context. Cloud-based processing of this data alone is associated with latency, privacy concerns, and bandwidth issues, which are not amenable to time-critical and safety-critical applications in healthcare. Edge computing addresses these challenges by facilitating on-device or near-device intelligence, which allows real-time inference, anomaly detection, and adaptive feedback with negligible latency. In contrast, cloud computing enables massive storage, long-term analytics, model training, population-level insights, and platform portability across healthcare systems. This chapter reviews hybrid edge–cloud architectures, AI model splitting, federated learning, resource-constrained inference, security, privacy preservation, and regulatory support. New applications in digital twins, immersive telemedicine, and metaverse-based healthcare are also covered. The chapter concludes with a discussion of potential research avenues for scalable, secure, and intelligent wearable AI systems.