Automated Visual Inspection System for Validation of Company Vehicle Standards in Delivery Fleets, Teslimat Filolarinda Sirket Ara Standartlarinin Dogrulanmasi i in Otomatik G rsel Denetim Sistemi

Abstract

This paper presents a preliminary study on the automatic detection of compliance with predefined standards for vehicles in delivery fleets. The aim is to minimise the dependency of human labour in determining whether different brand/model vehicles used by Yurtiçi Kargo Servisi AŞ during distribution services meet the company's standards. For this purpose, images of the company's vehicles taken from four different angles are uploaded to the system and analysed by a YOLOv11-based model. The model has been trained with different versions of YOLOv11 and the test results obtained are reported. The model detects the presence of a vehicle in each image, determines its orientation (if present), and identifies the company logo or small logo on the vehicle. Studies on the detection of other requirements of vehicle standards (fading, deterioration, logo compliance, etc.) are ongoing. © 2025 Elsevier B.V., All rights reserved.