Sustainable and Adaptive Smart Homes with Personalised Edge AI: Technologies, Applications and Research Challenges

Abstract

The smart home concept has changed significantly since its inception, from primarily remote-controlled devices and timers to advanced systems that can sense, learn, and adapt to occupants' preferences and behaviour. This shift has been made possible by integrating personalised edge AI, which lets smart home devices operate locally without relying entirely on the cloud. By using personalised edge AI, smart homes can respond in real time, protect users’ data by processing it locally, learn their preferences over time, and remain resilient to changes in network availability. This chapter reviews the fundamental architecture, critical technologies, learning techniques, and practical implementation of smart homes that utilise personalised edge AI. It addresses the advantages and disadvantages of placing AI at the edge, compares cloud-based and edge-based intelligence, and explores methods for achieving a balance between computational efficiency and end-user customization. It then describes how personalised edge AI systems are changing everyday life, drawing on a number of example applications: intelligent climate control, adaptive lighting and ambience, security and safety monitoring, voice interaction, and health monitoring. This chapter also discusses ethical and legal considerations, as well as user-experience design issues that enable wider adoption. Finally, this chapter concludes with suggestions for potential next steps on this interdisciplinary subject. Smart homes powered by personalised edge AI will revolutionise how AI is integrated into our daily lives by learning and adapting to us while ensuring our privacy and autonomy are respected. The result is homes that are more user-friendly, efficient, secure, and grounded in human-centred design.