Smart home security system using IoT for resilient remote tracing and mitigation of thefts

Abstract

Today people increasingly rely on technology as a fast and cost-effective means of improving their standard of living. Remote home-theft protection is no longer a naive concept, given how readily current microcontroller-based systems can be operated through Android control. Existing systems, however, still lack several features expected of a truly “smart” home theft-security system, including real-time notifications, face identification, reliability of the underlying technologies and high usability. The main objective of this research is therefore to identify, analyze and implement a solution to these drawbacks. The proposed system uses a Passive Infrared (PIR) sensor, a two-way relay switch, an Arduino Uno board and an HC-05 Bluetooth module, which link the home and the homeowner through a real-time Firebase online database; an accompanying Android application controls the system remotely and serves as the platform on which users are notified when a theft occurs. Face detection is performed on the phone using Google's Vision API. An end-to-end functional test confirmed that the proposed system fixes issues of current systems: intrusion events were correctly recorded ensuring resilience and that the corresponding notifications are delivered to registered users in near real-time.