Indonesia Prototype Traffic Light Control System Based on Ambulance Objects Using YOLO Method

Dimas Tubagus Hanggar Kesuma, Mila Kusumawardani, Azam Muzakhim ImammuddinNon-Scopus30 Juni 2025Afiliasi: ID

Abstrak

Ambulances are emergency service vehicles, but they are often caught in traffic jams that can potentially cause death to patients due to the need for quick action. Then, there was a case of an ambulance accident at a traffic intersection when requesting priority access to the road because it was carrying a corpse. With a traffic light control system using the YOLOv8 image processing method as an ambulance car detector and an audio warning signal output, it can provide warnings to prevent accidents at traffic intersections and prioritize ambulances to arrive faster to their destination. In the system, the ambulance is detected in front and rear view with a minimum confidence value of 70% to send a command to the ESP32 to activate or turn off the output. The ambulance detection system based on distance obtained a confidence value of 95% at a distance of 30cm, where the system is more accurate because the object is closer to the camera. Then, the detection system with light and speed obtained an average confidence value of 78.3% which is getting lower at night due to poor lighting conditions, in the object speed factor obtained a confidence value of 89% at a speed of 40km / hour. The confidence value is also influenced by the angle of the camera position when detecting objects, at an angle of 45° obtained a value of 89% and an angle of 135° with a value of 82%. The system is able to distinguish ambulances with similar objects, other objects detected obtained a confidence value of 43% - 56%. There is a delay in object recognition, the average delay value is 18.9 m/s which is influenced by computer performance. In the overall system delay there is an average delay value of 3 seconds. In order for the system to control traffic lights and output on time, it is determined that the minimum detection distance is 66.6m which is influenced by a delay of 3 seconds and an ambulance speed of 22.2 m/s.

Identifier

Scroll to Top