Automated inspection in robotic additive manufacturing using deep learning for layer deformation detection
- PDF / 9,423,640 Bytes
- 14 Pages / 595.276 x 790.866 pts Page_size
- 15 Downloads / 300 Views
Automated inspection in robotic additive manufacturing using deep learning for layer deformation detection Omid Davtalab1,2
· Ali Kazemian3 · Xiao Yuan2,4 · Behrokh Khoshnevis2,4
Received: 30 January 2020 / Accepted: 28 September 2020 © Springer Science+Business Media, LLC, part of Springer Nature 2020
Abstract In this paper, an automated layer defect detection system for construction 3D printing is proposed. Initially, a step-by-step procedure is implemented to develop a deep convolutional neural network that receives images as input and is able to distinguish concrete layers from other surrounding objects through semantic pixel-wise segmentation. Using data augmentation techniques, 1M labeled images are generated and used to train and test the CNN model. Then, a defect detection module is developed which is able to detect deformations in the printed concrete layers extracted from the images using the CNN model. The evaluation results based on metrics such as accuracy, F1 score, and miss rate verify the acceptable performance of the developed system. Keywords Deep learning · Semantic segmentation · Automated inspection · Material extrusion
Introduction Construction 3D printing is deemed to be an ongoing revolution in the construction industry and this new technology has received considerable attention in the past few years (Wu et al. 2016; Ghaffar et al. 2018; Craveiroa et al. 2019; Hojati et al. 2018; Viktor et al. 2019). While many applications have been envisioned for construction 3D printing, currently it is mainly utilized for automated building shell construction using cementitious materials. Contour Crafting (CC) is the first construction-scale additive manufacturing technology, which was developed at the University of Southern California and enables automated onsite construction of customized structures. CC technology is able to reduce the construction time and cost, and offers various technical advantages such as freedom in the architectural design, a wide selection of printing materials, and superior
B
Omid Davtalab [email protected]
1
Department of Computer Science, University of Southern California, Los Angeles, CA, USA
2
Contour Crafting Corporation, El Segundo, CA, USA
3
Department of Construction Management, Louisiana State University, Baton Rouge, LA, USA
4
Department of Industrial and Systems Engineering, University of Southern California, Los Angeles, CA, USA
surface finish (Davtalab et al. 2018). CC is also regarded as a viable approach for building extraterrestrial settlements and infrastructure development (Kazemian et al. 2017; Khoshnevis et al. 2015; Khoshnevis et al. 2016; Khoshnevis and Kazemian 2020). During the past few years, many studies have been carried out in order to investigate the design principles and performance requirements for materials which are used in construction 3D printing. These studies have been mainly focused on the cementitious printing materials (Kazemian et al. 2017; Wolfs et al. 2019; Perrot et al. 2016; Wolfs et al. 2018; Xia et al. 2019; Zhu et al
Data Loading...
