A Secure, Scalable, and Intelligent IoT-Based Smart Home Automation Framework Using Edge Computing for Real-Time Monitoring and Energy Optimization

Abstract: The widespread adoption of IoT technology has quietly reshaped how we think about the homes we live in. What was once a collection of isolated appliances is now becoming a network of communicating, sensing, and responding systems. Yet for all this progress, most smart home architectures still lean heavily on cloud servers to do the heavy lifting — and that creates real problems. When every sensor reading has to travel to a distant data centre before a decision is made, the delays add up, privacy is harder to guarantee, and the system becomes brittle whenever the internet connection is unreliable.

This paper presents SESI-EH — a Secure, Edge-based, Scalable, and Intelligent smart home framework built around edge computing. Rather than routing everything to the cloud, the system moves intelligence to the edge: a Raspberry Pi 4 gateway handles real-time anomaly detection, energy scheduling, and actuation locally, while ESP32 microcontrollers form the sensing and control layer. Lightweight machine learning models — Isolation Forest, an LSTM Autoencoder, and an SVM classifier — run directly on the edge device, achieving an ensemble intrusion detection accuracy of 96.3% with a total end-to-end response time under 46 ms. Energy experiments over 30 days showed a 52.8% reduction in daily consumption and an 81.2% cut in cloud-bound network traffic. These results suggest that bringing intelligence closer to home is not just technically feasible — it is the more practical path forward.

Keywords: Internet of Things (IoT), Smart Home Automation, Edge Computing, Real-Time Monitoring, Energy Optimization, Machine Learning, Anomaly Detection, Home Security, Raspberry Pi, ESP32, Intelligent Automation.