Design and Development of An Over-Dimension Estimation Device Using the Yolo Algorithm and Opencv

Operational violations of motor vehicles, particularly goods vehicles, have become an increasingly prevalent issue in Indonesia. Based on available data, 64.11% of violations are attributed to ODOL (Over Dimensions and Over Load), while the remaining 35.89% are related to violations of technical requirements, documentation, and loading procedures. Currently, the supervision of vehicle loads relies on Weigh in Motion (WIM) technology, which automatically weighs passing vehicles. However, the monitoring of vehicle dimensions is still conducted manually using measuring tools such as meters, which can lead to fatigue among officers and a reduction in measurement accuracy. To address this issue, this study aims to design an Over-Dimensions Estimator device using the YOLO Algorithm approach combined with OpenCV. The device is intended to estimate vehicle dimensions and identify dimensional violations based on variables such as Wheel Base (WB), Front Overhang (FOH), and Rear Overhang (ROH). The criteria for identifying dimensional violations are determined as follows: FOH_OD = FOH > (0.475 × WB) and ROH_OD = ROH > (0.625 × WB). This study is an applied research effort that produces a prototype device capable of detecting dimensional violations. With this tool, field officers will only need to conduct manual measurements when the device identifies a dimensional violation, thereby making the monitoring process more efficient. The device can be installed at Vehicle Weigh Stations (UPPKB) or other strategic locations, ensuring compliance with technical requirements, roadworthiness standards, and proper supervision of motor vehicle dimensions on the road.