Enhancing Road Safety Education Using an AI-Driven Mobile Traffic Sign Recognition System

Abstract

Road safety education plays a crucial role in reducing traffic accidents, particularly among novice drivers who often struggle to recognize and recall traffic signs in real-world situations. Conventional learning methods such as manuals and classroom instruction lack interactivity and contextual visualization. To address this limitation, this study proposes an AI-powered mobile application for traffic sign recognition using a TensorFlow Lite model. The system enables real-time detection of Malaysian traffic signs through a smartphone camera or image selection, providing instant sign identification and explanatory information. In addition to detection, the application incorporates learning mode, scan history tracking, and quiz-based assessments to enhance user engagement and knowledge retention. The system was developed using Flutter for the mobile interface, a PHP-based backend, and a MySQL database for content management. Experimental evaluation demonstrates that the proposed system achieves accurate traffic sign recognition while offering significant educational value through its interactive features. The findings indicate that integrating lightweight AI models with mobile learning applications can effectively support traffic sign education and promote road safety awareness.