Development of a YOLO- and MQTT-Based Overtaking Warning System for Intelligent Driver Safety Education

Haidar, Muhammad Faris
Wibowo, Helmi

Information Technology Education Journal
Vol. 5, No. 2, May (2026)

Purpose – This study aims to develop an intelligent overtaking warning system based on YOLOv8 object detection and Vehicle-to-Vehicle (V2V) communication using MQTT, designed as a prototype with the potential to serve as an interactive learning medium for driver safety and vehicular communication concepts. The study is motivated by the limited availability of practical educational tools for understanding overtaking processes and real-time communication in the Internet of Vehicles (IoV) context. Design – The research adopts a research and development (R&D) approach, including system design, implementation, and testing stages. The system is built using Raspberry Pi and ESP32, integrating GPS and LiDAR sensors with OCR-based recognition, and is evaluated through technical performance testing and user perception analysis using a Likert-scale questionnaire with validity and reliability testing. Findings – The results show that the system achieves an average end-to-end detection and processing delay of 0.911 seconds, while MQTT communication latency averages approximately 0.1269 seconds under controlled network conditions, with stable bidirectional communication. The YOLOv8 model performs optimally at a confidence threshold of 0.4, and the GPS and LiDAR sensors produce average error rates of 3.99% and 2.86%, respectively, while MQTT communication achieves a 100% success rate under tested conditions. Questionnaire results indicate that respondents reported positive perceptions regarding system usefulness, with most questionnaire items meeting validity criteria (r > 0.31) and a Cronbach’s Alpha of 0.952, indicating high reliability. Research Implication – These findings suggest that the system is technically feasible and demonstrates perceived educational potential as an interactive learning medium. However, this study is limited by the number of respondents and simulation-based testing; therefore, future work should include real-world traffic testing and larger-scale evaluations to improve system robustness and applicability. Originality – This study integrates YOLOv8-based object detection with MQTT-based V2V communication to develop an intelligent overtaking warning system as an interactive learning medium in IoV contexts.