This paper presents the development and evaluation of an artificial intelligence-based system for automated dimensional inspection of extruded tubes in the Brazilian automotive industry. The primary objective is to replace traditional aluminum jigs used for geometric validation with a flexible and intelligent system based on the YOLOv8 object detection algorithm, explicitly designed for deployment on mobile devices such as smartphones. This mobile-centered approach aims to reduce infrastructure costs, improve portability, and enable on-site, real-time inspection with limited computational resources. Two model variants—YOLOv8n (nano) and YOLOv8s (small)—were trained on a dataset of 1905 annotated images over 100, 200, and 300 epochs. The evaluation considered key performance metrics: precision, recall, mAP@0.5, and mAP@0.5:0.95. The YOLOv8s model trained for 300 epochs achieved the best mAP@0.5:0.95 (0.866) and mAP@0.5 (0.954), while YOLOv8n with 100 epochs achieved the highest precision (0.970) and mAP@0.5 of 0.948—ideal for mobile applications due to its lightweight architecture (3.2 million parameters and 8.7 GFLOPs). Average inference time remained below 25 ms per image, with confidence levels of 0.96 for conforming parts and 0.91 for non-conforming detections. The results confirm the technical viability and industrial applicability of using YOLOv8 on mobile platforms, supporting the transition from manual to intelligent, automated inspection in alignment with Industry 4.0. Future work will focus on dataset expansion, multi-class detection, and integration of real-time feedback mechanisms.