Abstract:
The growth of smart home technology has made it possible to create intelligent, userconscious systems that add comfort, energy efficiency, and remote control. This work presents HomeMatrix, a sophisticated smart home automation system using Raspberry Pi, which utilizes real-time object detection through the YOLOv8 model and incorporates behavior learning by using machine learning algorithms like Random Forest.It detects human presence automatically to manage appliances such as lights and minimize unnecessary energy consumption. It learns user habits and movement patterns over time through timestamped information stored locally in SQLite to create predictive automation based on behavior. Moreover, Firebase Cloud Messaging (FCM) allows real-time mobile notifications upon detection of activity to keep users up to date.With its web dashboard and mobile support, HomeMatrix provides homeowners with complete visibility and control over their space. The initiative highlights the potential of merging computer vision, edge computing, and user behavior modeling to create a more intelligent, context-aware, and energy-efficient home.
