Violation of motorcyclists in special road areas such as pedestrian paths, bus lanes, and toll roads still frequently occurs, posing the risk of accidents and disrupting other road users. Factors contributing to these violations include avoiding traffic jams to reach destinations more quickly. Therefore, a motorcycle detection system has been developed to facilitate the handling of violations by providing information to relevant security personnel. The detection system utilizes the YOLOv5 method, trained with 492 images with varying parameters such as epoch value ( 200) and batch size (24). Using the YOLOv5 model, it achieves a Mean Average Precision (MAP) value of 89.8%, indicating good detection quality. The detection system can be implemented in real-time using a webcam. Dummy violation data is processed using a Raspberry Pi 4B microcontroller. Testing the detection system in light intensity ranging from 6171 Lux to 140516 Lux demonstrates its quick response capability in sending information via the Telegram application, taking around 171 milliseconds for each data packet. The system effectively detects objects and promptly provides information to relevant security personnel, enhancing performance in addressing violations within the monitored area.