This work presents the design and implementation of an autonomous classifier that scores the quality of wire rod coil formation based on the irregularity of the coil formation on the roller table. Digital processing techniques were applied to grab images of the coil former/roller table. In the classifier design were consider regions of interest, in which attributes were extracted and labels were created. The scores estimated by the designed classifier are in consonance with the scores assigned by an expert in wire rod coils formation process. Thus, the classifier can be used to assist operational decision making related to changes in process parameters, such as formation speed, in order to increase the coils uniformity on the roller table, which can minimizes possible material losses and increases the production and the quality of the final product.