Smart Anti-Theft Bag: Design and Implementation of a Dual-Sensor Latching Security System Using Arduino

by Gian Carlo G. Onia, Jarold Jan Silverio, Kevin Timothy Doniza, Minerva C. Zoleta, Regine Libongcogon, Ronnie Bas

Published: January 31, 2026 • DOI: 10.47772/IJRISS.2026.10100210

Abstract

The successful safeguarding of individual property against not only unobtrusive intrusion, but also aggressive dispossession, is a root requirement to commuter security in the swiftly developing environment of the modern transit systems. Conventional security systems like padlocks are the common ones; however, this system is passive and does not give an immediate notification to the owner at the most crucial stages of the theft attempt. As a result, intelligent and automated personal security systems development is highly demanded. The current paper describes the design, production, and implementation of a low-cost and automated anti-theft bag, which is specifically designed to help in reducing the cases of pick pocketing and snatching. The project will attempt to emulate commercial capabilities of security solutions and combine the ideas of sensor fusion, kinetic monitoring, and embedded system control into a wearable form factor.
The Arduino Nano microcontroller is the core element of the system architecture as it is the main processing unit. It uses a dual-sensor system with an interlocking reed switch that detects zipper violations, and an SW-420 vibration sensor that reacts to violent movement that can serve as an indicator of snatching; the latter has a variable sensitivity potentiometer to reduce false alarms due to normal walking. The security system is actuated by using a dual-feedback system with a high-decibel piezo buzzer as an acoustic deterrent and high-intensity LED to indicate the status. The control software is a finite state machine (FSM) with the logic of a Smarts Latch, Wiggle Filtering to eliminate spurious activations, and manual reset hardware interrupt routines. Practical confirmation The experimentally determined detection accuracy is 100 per cent with sensor thresholds suitably adjusted.
To conclude, this project illustrates that a practical implementation of robotics and computer programming can be used to develop wearable technology and it is conclusive that highly challenging security problems can be effectively modeled with the help of affordable, easily accessible hardware.