Evaluating the Use of Artificial Neural Networks, Graph Theory, and Complexity Theory to Predict Automotive Assembly Defects

Artificial Neural Networks (ANNs) have been used to predict assembly time and market value from assembly models. This was done by converting the assembly models into bipartite graphs and extracting 29 graph complexity metrics which were used to train the ANN prediction models. This paper presents the use of sub-assembly models instead of the entire assembly model to predict assembly quality defects at an automotive OEM. The size of the training set, order of the bipartite graph, selection of training set, and defect type were experimentally studied. With a training size of 28 parts, an interpolation focused training set selection, and second order graph seeding, over 70% of the predictions were within 100% of the target value. The study shows that with an increase in training size and careful selection of training sets, assembly defects can be predicted reliably from sub-assemblies complexity data.