Artificial Intelligence, Machine Learning and Smart Technologies for Nondestructive Evaluation

Taheri, Hossein; Bocanegra, Maria Gonzalez; Taheri, Mahdi

doi:10.3390/s22114055

Cited by 36 publications

(11 citation statements)

References 53 publications

Supporting

Mentioning

Contrasting

Unclassified

Order By: Relevance

“…To reduce the false alarm in a monitoring methodology, in addition to the appropriate sensor integration, accurate and robust data analysis and signal processing are necessary. Machine learning and computer vision techniques can reduce the false alarm and enhance the accuracy of data analysis for this purpose [65].…”

Section: Discussionmentioning

confidence: 99%

Failures and Flaws in Fused Deposition Modeling (FDM) Additively Manufactured Polymers and Composites

Baechle-Clayton

Loos

Taheri³

et al. 2022

J. Compos. Sci.

Self Cite

View full text Add to dashboard Cite

In this review, the potential failures and flaws associated with fused deposition modeling (FDM) or fused filament fabrication (FFF) 3D printing technology are highlighted. The focus of this article is on presenting the failures and flaws that are caused by the operational standpoints and which are based on the many years of experience with current and emerging materials and equipment for the 3D printing of polymers and composites using the FDM/FFF method. FDM or FFF 3D printing, which is also known as an additive manufacturing (AM) technique, is a material processing and fabrication method where the raw material, usually in the form of filaments, is added layer-by-layer to create a three-dimensional part from a computer designed model. As expected, there are many advantages in terms of material usage, fabrication time, the complexity of the part, and the ease of use in FDM/FFF, which are extensively discussed in many articles. However, to upgrade the application of this technology from public general usage and prototyping to large-scale production use, as well as to be certain about the integrity of the parts even in a prototype, the quality and structural properties of the products become a big concern. This study provides discussions and insights into the potential factors that can cause the failure of 3D printers when producing a part and presents the type and characteristics of potential flaws that can happen in the produced parts. Common defects posed by FDM printing have been discussed, and common nondestructive detection methods to identify these flaws both in-process and after the process is completed are discussed. The discussions on the failures and flaws in machines provides useful information on troubleshooting the process if they happen, and the review on the failures and flaws in parts helps researchers and operators learn about the causes and effects of the flaws in a practical way.

show abstract

Section: Discussionmentioning

confidence: 99%

Failures and Flaws in Fused Deposition Modeling (FDM) Additively Manufactured Polymers and Composites

Baechle-Clayton

Loos

Taheri³

et al. 2022

J. Compos. Sci.

Self Cite

View full text Add to dashboard Cite

show abstract

“…A SVM vem sendo utilizada por pesquisadores para a identificação, classificação e quantificação de defeitos na área de ensaios não destrutivos [16], [17]. Adicionalmente, existem diversos estudos na área que demonstram uma alta taxa de sucesso quando combinados à técnica PEC [18].…”

Section: B Máquina De Vetores De Suporteunclassified

Classificação de Tensões em Chapas de Aço IF utilizando Aprendizado de Máquina Aplicado à Sinais de Correntes Parasitas Pulsadas

Larocca¹,

Britto²,

Filho³

et al. 2022

Anais Do XL Simpósio Brasileiro De Telecomunicações E Processamento De Sinais

View full text Add to dashboard Cite

Resumo-Aços IF são empregados na fabricação de peças na indústria automobilística. Ao serem transportados e armazenados em forma de bobinas, suas propriedades mecânicas estão sujeitas a alterações por ação das tensões residuais intensificadas pelo próprio peso, o que pode gerar falhas. Este trabalho pretende classificar amostras de aço IF a partir da posição onde elas foram extraídas de uma bobina de armazenamento. Para isso, foram adquiridos sinais de correntes parasitas pulsadas das amostras e técnicas de aprendizado de máquina foram utilizadas para classificação. Com o método proposto, foi obtida uma acurácia de até 86,10% na identificação das classes de interesse.

show abstract

“…Machine learning (ML) is a subfield of computer science that focuses on methods and techniques for automating solutions to complex problems that are difficult to address using standard programming methods. Figure 2 shows that machine learning is a subset of artificial intelligence [ 43 ].…”

Section: Smart Inspectionmentioning

confidence: 99%

On Smart Geometric Non-Destructive Evaluation: Inspection Methods, Overview, and Challenges

et al. 2022

View full text Add to dashboard Cite

Inspection methods, also known as non-destructive evaluation (NDE), is a process for inspecting materials, products, and facilities to identify flaws, imperfections, and malfunctions without destruction or changing the integrity of materials, structures, and mechanisms. However, detecting those defects requires test conducting and results inferring, which is highly demanding in terms of analysis, performance, and time. New technologies are therefore needed to increase the efficiency, probability of detection, and interpretability of NDE methods to establish smart inspection. In this context, Artificial intelligence (AI), as a fundamental component of the Industry 4.0, is a well-suited tool to address downsides associated with the current NDE methods for analysis and interpretation of inspection results, where methods integrating AI into their inspection process become automated and are known as smart inspection methods. This article sheds a light on the conventional methods and the smart techniques used in defects detection. Subsequently, a comparison between the two notions is presented. Furthermore, it investigates opportunities for the integration of non-destructive evaluation (NDE) methods and Industry 4.0 technologies. In addition, the challenges hindering the progress of the domain are mentioned as the potential solutions. To this end, along with Industry 4.0 technologies, a virtual inspection system has been proposed to deploy smart inspection.

show abstract

Artificial Intelligence, Machine Learning and Smart Technologies for Nondestructive Evaluation

Cited by 36 publications

References 53 publications

Failures and Flaws in Fused Deposition Modeling (FDM) Additively Manufactured Polymers and Composites

Failures and Flaws in Fused Deposition Modeling (FDM) Additively Manufactured Polymers and Composites

Classificação de Tensões em Chapas de Aço IF utilizando Aprendizado de Máquina Aplicado à Sinais de Correntes Parasitas Pulsadas

On Smart Geometric Non-Destructive Evaluation: Inspection Methods, Overview, and Challenges

Contact Info

Product

Resources

About