Many companies are already collecting data on the geometry of the components that they manufacture for use in the assembly of their products. This data can be used for better control of quality assurance by predicting the cumulative effect of geometric deviations that can have an impact on assembly. However, this data is not standardized, it is collected at different points in the process and stored in different systems. Therefore, a strategy is needed to facilitate using these geometric data by applying algorithms that allow using them to improve product assembly. To address this problem, in this article, we propose a technological framework to predict the quality defects that may occur in the production process, based on geometric deviations, using inspection data collected on the production line. This allows a virtual simulation of the assembly of the vehicles to prevent quality defects that may affect the assembly of parts on the vehicle.