With the success of the applicability of Artificial Intelligence and Machine Learning (AI/ML), approaches in the field of computer vision has been highlighted, such as the recognition of objects and their intrinsic details. Within the manufacturing sector, especially in the automotive industry, there has been a motivation for urgent innovations and implementations of computer vision systems to control the assembly of the parts and their visual quality lately, in order to replace old equipment, such as cameras and scanners that have carried out this very limited verification until today, as well as the software license costs. Although there are multiple suppliers of this type of software on the market, it’s important to explore alternatives to reduce this expensive hardware and software license cost that comes with this innovation predicted by the Industry 4.0 global scope. In this sense, using AI/ML approaches for computer vision, in this work, we propose a cheaper method to explore and implement object detection applications in the assembly lines, as well as their particularities and characteristics, calculating distances between components and managing other visual quality elements, built on top of the YOLO architecture, one of the biggest and leading algorithms if the computer vision field. Promising results have been achieved in real scenarios, namely in the detection of the bag ear, plastic and metal clip, tapes and oeticker detection and measurements and in smart area inference simulations.