Inspection of mechanical assemblies based on 3D Deep Learning approaches

Assya Boughrara (1, 2) , Igor Jovančević (2) , Hamdi Ben Abdallah (1) , Benoît Dolives (2) , Mathieu Belloc (2) , Jean-José Orteu (1)

1 ICA - Institut Clément Ader
2 DIOTA

Our research work is being carried out within the framework of the joint research laboratory ”Inspection 4.0” between IMT Mines Albi/ICA and the company DIOTA specialized in the development of numerical tools for Industry 4.0. In this work, we are focused on conformity control of complex aeronautical mechanical assemblies, typically an aircraft engine at the end or in the middle of the assembly process. A 3D scanner carried by a robot arm provides acquisitions of 3D point clouds which are further processed by deep classification networks. Computer Aided Design (CAD) model of the mechanical assembly to be inspected is available, which is an important asset of our approach. Our deep learning models are trained on synthetic and simulated data, generated from the CAD models. Several networks are trained and evaluated and results on real clouds are presented.

Mots clés: 3D modeling; Data modeling; Inspection; 3D scanning; Computer aided design; Solid modeling; 3D acquisition; Clouds; Laser scanners; Optical inspection

Domaines: Sciences de l'ingénieur [physics]; Intelligence artificielle [cs.AI]; Traitement des images [eess.IV]

Assya Boughrara, Igor Jovančević, Hamdi Ben Abdallah, Benoît Dolives, Mathieu Belloc, et al.. Inspection of mechanical assemblies based on 3D Deep Learning approaches. QCAV’2021 - 15th International Conference on Quality Control by Artificial Vision, May 2021, Tokushima (online), Japan. 8 p., ⟨10.1117/12.2588986⟩. ⟨hal-03230297⟩