Deteksi Otomatis Pejalan Kaki di Zebra Cross menggunakan Metode Sensor Fusion berdasarkan Internet of Things

Abstract

Pedestrian safety at zebra crossings remains a significant challenge due to the lack of driver awareness regarding pedestrian rights and priority. To minimize the risk of accidents, this study proposes an automatic pedestrian detection system at zebra crossings using a sensor fusion method based on the Internet of Things (IoT), employing mmWave and HC-SR04 sensors. The system is designed to detect the presence of pedestrians in the zebra crossing area and provide warnings in the form of sirens and indicator lights, which are controlled by an ESP32 microcontroller. The indicator lights can be activated by users through the Telegram application. The sensor fusion method is utilized to improve detection accuracy by combining data from both sensors, enabling the system to deliver more precise and consistent alerts. The designed system successfully detects the presence of pedestrians using mmWave sensors, and detects vehicles using HC-SR04 sensors in real-time. The test was conducted 3 (three) times, on May 17, 2025 with the buzzer and light sensor active, May 20, 2025 with only the buzzer active, and May 26, 2025 with the buzzer and light sensor active. The test results show that the system works stably, quickly (response < 2 seconds), and can operate continuously without interruption, with successful data communication sent to the server via Wi-Fi network.