top of page

VAICeramics - Visual and Acoustics Inspection of Ceramics

PROJECT FORM

Project name: VAICeramics - Visual and Acoustics Inspection of Ceramics

Universal Code: 69987

Main Objective: OT1 - Strengthen research, technological development and innovation

Intervention Region: North, Center and Lisbon

Beneficiary entity: AGIX, LDA

Objectives, activities and expected / achieved results:

 

The project VAICeramics - Visual and Acoustics Inspection of Ceramics aims to develop a technology that allows to automation of the inspection process of ceramic items, using computer vision and audition with machine learning. This is a pre-industrial prototype that will be composed of a system that will allow the automatic recognition of the type of piece (eg: plate, bowl, cup), including a robot, designed tothe pieces to be inspected, which will allow the evaluation of structural integrity, functional aspects and surface defects. This inspection will also integrate an acoustic test that will allow the collection of a touch sound made by the parts manipulating robot . The system automatically adapts to the different part models registered, without human intervention, separating the parts according to the analysis result. The promoting consortium is composed by 2 companies, MATCERÂMICA - FABRICO DE LOUÇA SA and AGIX, LDA, and 2 NE Entities of SI&I, the ASSOCIAÇÃO FRAUNHOFER PORTUGAL RESEARCH and the CENTRO TECNOLOGICO DA CERAMICA E DO VIDRO (CTCV).

Approval date: 01/04/2021

Start date: 01/04/2021

Conclusion date: 06/30/2023

Total eligible cost: €980,512.90

European Union Financial Support: ERDF - €622,818.76

Modelos-Barras-FUNDOS-v04_3logos-FCoesao-descritivo.png

LATEST DEVELOPMENTS

November 2022

Excerpt from the CTCV publication: Technical Journal of Ceramics and Glass

(NOVEMBER/DECEMBER 2022).

Authors

Margarida Felício [1], Rui Carreira [1] Idalina Eusébio [1], Filipe Soares [2], João Gonçalves [2], Luís Rosado [2], Alberto Gil [3], André Santos [3], Victor Francisco [4], Rui Neves [4]


[1] MATCERÂMICA - Fabrico de louça SA
[2] Associação Fraunhofer Portugal Research (FhP)
[3] AGIX, Lda
[4] CTCV – Centro Tecnológico da Cerâmica e do Vidro

VAICeramics - Visual and acoustic inspection of ceramics

Ongoing Developments

The project started with a study, selection and testing of existing technologies in the market for automatic detection of surface defects and structural integrity. This work included a preliminary analysis of applicable technologies to the project in order to validate which solutions already exists in the market, more feasible to acquire, and confirm which solutions develop in the Project.

This was followed by a study of acoustics inspection on samples, comprising the performing tests in a controlled environment, using samples of ceramic products, to identify specific and differentiating properties of each type of piece from specific lines. As part of the task, a search for studies related to the inspection of acoustics in ceramic pieces was preform. The tests were selected according to their ability to obtain acoustic signals emitted by parts that can be used in quality control. The response to the impact applied to different types of ceramic pieces was evaluated in the temporal, spectral and statistical domains. This task allowed to reinfuse the team's knowledge regarding the problem approched, thus allowing a better assessment and mitigation of risks in prototyping tasks.

The need for a software license and the high price of the software with the greatest potential, the difficulty of customizing software libraries for computer vision (for example, not compatible with the need to explore more methods of segmentation, object detection and classification) , the limited use of annotation tools specific to the use case, and the difficulty in acquiring specific hardware not available for delivery (for example, smart cameras), reinforced the need for the VAICeramics consortium to develop its own software with acquisition of cameras with image sensors and communication protocols commonly used in the artificial vision industry.

A report about the characterization of defects occurring in ceramic products was also prepared, with the aim of a characterizing and selecting of the main and most common defects that occurs in utilitarian ceramic tableware production, in these case in the production process of MATCERÂMICA company.

This characterization of defects was carried out by processing the information available according to the occurrence of each type of defect:

  • Pores

  • Cracked

  • Dirt in the glaze

  • Black spots

  • Lack of glaze

  • Air pockets

  • Trash resulting from firing

  • Broken base

Regarding the aims of the project, to create a system for the automatic detection and identification of defects in ceramic pieces, using artificial intelligence methods, it was necessary to create a platform for manual annotation of defects and areas of the pieces, so that the system learns how to identify correctly. Usability tests were preformed, and users interacted with the annotation platform, to assess its usability, and the need for improvement was identified. The objectives of the tests were to evaluate how users interact with the interface and to understand the general usability of the platform.

 

The overall system architecture and back-end requirements and integration services, the data acquisition and machine learning specification, and the front-ends specification have also been implemented.

 

We share the full article [PT] from CTCV in: https://issuu.com/ctcv/docs/revista-tecnica17-novembro-dezembro/28  

bottom of page