Deep learning for big data applications in CAD and PLM – Research review, opportunities and case study - Université de technologie de Compiègne Access content directly
Journal Articles Computers in Industry Year : 2018

Deep learning for big data applications in CAD and PLM – Research review, opportunities and case study

Abstract

With the increasing amount of available data, computing power and network speed for a decreasing cost, the manufacturing industry is facing an unprecedented amount of data to process, understand and exploit. Phenomena such as Big Data, the Internet-of-Things, Closed-Loop Product Lifecycle Management, and the advances of Smart Factories tend to produce humanly unmanageable quantities of data. The paper approaches the aforesaid context by assuming that any data processing automation is not only desirable but rather necessary in order to prevent prohibitive data analytics costs. This study focuses on highlighting the major specificities of engineering data and the data-processing difficulties which are inherent to data coming from the manufacturing industry. The artificial intelligence field of research is able to provide methods and tools to address some of the identified issues. A special attention was paid to provide a literature review of the most recent (in 2017) applications, that could present a high potential for the manufacturing industry, in the fields of machine learning and deep learning. In order to illustrate the proposed work, a case study was conducted on the challenging research question of object recognition in heterogeneous formats (3D models, photos and videos) with deep learning techniques. The DICE project – DMU Imagery Comparison Engine – is presented and has been completely open-sourced in order to encourage reuse and improvements of the proposed case-study. This project also leads to the development of an open-source research dataset of 2000 CAD Models, called DMU-Net available at: https://www.dmu-net.org.
No file

Dates and versions

hal-01996852 , version 1 (28-01-2019)

Identifiers

Cite

Jonathan Dekhtiar, Alexandre Durupt, Matthieu Bricogne, Benoit Eynard, Harvey Rowson, et al.. Deep learning for big data applications in CAD and PLM – Research review, opportunities and case study. Computers in Industry, 2018, 100, pp.227-243. ⟨10.1016/j.compind.2018.04.005⟩. ⟨hal-01996852⟩
155 View
0 Download

Altmetric

Share

Gmail Facebook X LinkedIn More