Evaluating the Use of Artificial Neural Networks and Graph Complexity to Predict Automotive Assembly Quality DefectsSource: Journal of Computing and Information Science in Engineering:;2017:;volume( 017 ):;issue: 003::page 31017Author:Patel, Apurva
,
Andrews, Patrick
,
Summers, Joshua D.
,
Harrison, Erin
,
Schulte, Joerg
,
Laine Mears, M.
DOI: 10.1115/1.4037179Publisher: The American Society of Mechanical Engineers (ASME)
Abstract: This paper presents the use of subassembly models instead of the entire assembly model to predict assembly quality defects at an automotive original equipment manufacturer (OEM). Specifically, artificial neural networks (ANNs) were used to predict assembly time and market value from assembly models. These models were converted into bipartite graphs from which 29 graph complexity metrics were extracted to train 18,900 ANN prediction models. 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 with a second-order graph seeding ensured that 70% of all 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 subassemblies' complexity data.
|
Show full item record
contributor author | Patel, Apurva | |
contributor author | Andrews, Patrick | |
contributor author | Summers, Joshua D. | |
contributor author | Harrison, Erin | |
contributor author | Schulte, Joerg | |
contributor author | Laine Mears, M. | |
date accessioned | 2017-11-25T07:20:33Z | |
date available | 2017-11-25T07:20:33Z | |
date copyright | 2017/26/7 | |
date issued | 2017 | |
identifier issn | 1530-9827 | |
identifier other | jcise_017_03_031017.pdf | |
identifier uri | http://138.201.223.254:8080/yetl1/handle/yetl/4236534 | |
description abstract | This paper presents the use of subassembly models instead of the entire assembly model to predict assembly quality defects at an automotive original equipment manufacturer (OEM). Specifically, artificial neural networks (ANNs) were used to predict assembly time and market value from assembly models. These models were converted into bipartite graphs from which 29 graph complexity metrics were extracted to train 18,900 ANN prediction models. 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 with a second-order graph seeding ensured that 70% of all 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 subassemblies' complexity data. | |
publisher | The American Society of Mechanical Engineers (ASME) | |
title | Evaluating the Use of Artificial Neural Networks and Graph Complexity to Predict Automotive Assembly Quality Defects | |
type | Journal Paper | |
journal volume | 17 | |
journal issue | 3 | |
journal title | Journal of Computing and Information Science in Engineering | |
identifier doi | 10.1115/1.4037179 | |
journal fristpage | 31017 | |
journal lastpage | 031017-10 | |
tree | Journal of Computing and Information Science in Engineering:;2017:;volume( 017 ):;issue: 003 | |
contenttype | Fulltext |