Next Article in Journal
Design of Multi-Band Bandstop Filters Based on Mixed Electric and Magnetic Coupling Resonators
Next Article in Special Issue
Rolling Bearing Residual Useful Life Prediction Model Based on the Particle Swarm Optimization-Optimized Fusion of Convolutional Neural Network and Bidirectional Long–Short-Term Memory–Multihead Self-Attention
Previous Article in Journal
Leveraging Neighbor Attention Initialization (NAI) for Efficient Training of Pretrained LLMs
Previous Article in Special Issue
Rolling Bearing Fault Diagnosis Based on SVD-GST Combined with Vision Transformer
 
 
Article
Peer-Review Record

Real-Time Defect Detection in Electronic Components during Assembly through Deep Learning

Electronics 2024, 13(8), 1551; https://doi.org/10.3390/electronics13081551
by Eyal Weiss *, Shir Caplan, Kobi Horn and Moshe Sharabi
Reviewer 1: Anonymous
Reviewer 2: Anonymous
Electronics 2024, 13(8), 1551; https://doi.org/10.3390/electronics13081551
Submission received: 29 February 2024 / Revised: 12 April 2024 / Accepted: 17 April 2024 / Published: 19 April 2024
(This article belongs to the Special Issue Fault Detection Technology Based on Deep Learning)

Round 1

Reviewer 1 Report

Comments and Suggestions for Authors The topic about screening of defective electronic components is very interesting and the authors have not properly delivered the details of the proposed work. Paper structure and organization is very poor. The authors should carefully revise the complete paper and submit again. 
Do you consider the topic original or relevant in the field? Does it address a specific gap in the field?The proposed methodology has not been reflected in the abstract.
Unnecessary keywords should be avoided such as quality etc.
In several headings the authors stated sections without any relevance.
Change the title of the paper and make it specific and relevant to the proposed research e.g. image processing or deep learning should be mentioned. How the authors are incorporating AI and Big Data in the proposed work.
Lastly, the authors should do some comparisons with others to validate their proposed work.

 

Comments on the Quality of English Language

1The English language of the paper is understandable, however, there are areas where specificity could be improved to enhance the overall quality and impact of the paper.   

Author Response

Response to Reviewer Comments:

We appreciate the detailed feedback you provided on our manuscript. Your suggestions helped improve our work. We've carefully looked into each point you raised and made the necessary changes to make our paper better.

Firstly, we've worked on the structure and organization of the paper to make it easier to follow. We've reorganized the content so that it flows more logically, with each section contributing to a better understanding of our research.

We understand the importance of highlighting the significance of our research topic. In the revised abstract, we've explained our proposed methodology more clearly, making sure it reflects how we're using AI and Big Data techniques for real-time defect detection in electronic components.

We've also gone through the keywords and headings used in the paper and made sure they're relevant to our focus. Additionally, we've updated the title of the paper to be more specific and reflective of what we're studying, including keywords like image processing, deep learning, AI, and Big Data.

The new title, "Real-Time Defect Detection in Electronic Components During Assembly Through Deep Learning," better represents what our research is about.

In response to your suggestion about comparing our work with others, we've looked into related studies and added some comparisons where it made sense. For instance, we've included a comparison with Robust Principal Component Analysis (RPCA) to give more context to our findings and show how our method stands out.

We appreciate your feedback, and we believe these changes have made our paper much better. Thank you for helping us improve.

Author Response File: Author Response.pdf

Reviewer 2 Report

Comments and Suggestions for Authors

The paper titled "In-Situ Screening of Defective Electronic Components through Edge Real-Time Big Data AI Analysis" presents a strategy for enhancing quality control in electronic component assembly through real-time image processing. By capturing images from pick-and-place machines during the component pickup and mounting process, the approach enables the identification of defects in the line, significantly minimizing the likelihood of defective products.

Despite the potential of this paper, there are several critical issues that require the authors' attention to improve the manuscript.

First, it does not adequately reflect the original contributions of the authors to the current state of knowledge. To address this, the authors must clearly state their original contributions, highlighting the novel aspects of their research from a specific point onwards. Additionally, they must provide a detailed description of their original methods, results, and conclusions because the current version of the paper is insufficiently detailed in these areas.

Second, the manuscript's current form resembles a commercial product description for quality control in electronic component assemblies. However, the study lacks information about the methodology used, mathematical details, and comparisons with other approaches presented in the literature. To improve the manuscript, the authors should provide more details about their methodology, including mathematical aspects, and offer comparisons with other studies in the field.

Third, the following aspects must be emphasized in a more profound manner: the context in which the authors should place the issue addressed by the manuscript in a broader context and highlight the purpose of the study, the methods used to solve the identified issue, the main findings of the article, and the main conclusions or interpretations. In other words, the authors must declare and justify the originality of their work.

For instance, a common method to carry out in-situ analysis of image data for quality inspection could be based on Robust Principal Component Analysis (RPCA), which separates sparsely corrupted anomalous components from a low-rank background. RPCA can be applied to images for the simultaneous detection and isolation of anomalies. Additionally, there are numerous post-processing procedures for image denoising and segmentation that can be applied in case studies. A comprehensive literature review was conducted to highlight the main contributions of this study.

Fourth, it would be advantageous for the manuscript if the authors included a proper "Conclusions" section in which they state the most crucial outcome of their work. The authors should not merely restate (as they did in the current version of the manuscript) what they have done or what the article does, but rather focus on what they have discovered and, most importantly, what their findings mean.

Finally, there are several minor typographical errors that make the manuscript difficult to read. The figures illustrating computational stress are not useful to the reader. It would be better to report numerical measurements that can be replicated experimentally. Furthermore, some figures are not cited in the text, or there are errors in citations (see line 574).

Author Response

Response to Reviewer Comment 1:

We are grateful for the reviewer's valuable feedback regarding the need for a clearer articulation of our original contributions and a more detailed description of our methods, results, and conclusions. In response to these insightful comments, we have diligently revised the introduction section of the paper to better elucidate the original contributions of our research and underscore the novel aspects of our approach.

 

Specifically, we have furnished a clear statement delineating our original contributions from a specific point onwards, elucidating the distinctive facets of our research within the current state of knowledge. Additionally, we have expanded upon the description of our original methods, results, and conclusions, thereby ensuring a more comprehensive understanding of the intricacies of our research methodology and the significance of our findings.

 

 

Response to Reviewer Comment 2:

We appreciate the insightful feedback provided by the reviewer, highlighting the importance of a more detailed methodology description and comparisons with existing studies in the field. While our manuscript primarily focuses on advancing the algorithm from cloud computing to real-time processing, we acknowledge the necessity of addressing these aspects to strengthen the overall scholarly contribution.

In response to the reviewer's constructive comments, we have implemented significant revisions to the manuscript. Specifically, we have expanded the methodology section to provide more comprehensive insights into our approach, thereby enhancing clarity and transparency in our methods. Furthermore, we have incorporated comparisons with other relevant studies in the field to offer a broader context for our research.

These revisions are aimed at elevating the research rigor and completeness of our manuscript, ensuring that it meets the scholarly standards and expectations of our readers and peers. We believe that these enhancements will enrich the scholarly discourse surrounding our study and contribute to the advancement of knowledge in the field of image processing for quality inspection.

 

Response to Reviewer Comment 3:

In response to these points, we have undertaken significant revisions to enhance the clarity and depth of our manuscript. First and foremost, we have expanded the discussion to contextualize our research within the broader landscape of image data analysis for quality inspection. By providing this contextual framework, we aim to underscore the significance of our study within the field, emphasizing its relevance and contributions.

Furthermore, we have refined the articulation of the objectives driving our research efforts. Our goal is to convey the purpose of our study more effectively, enabling readers to grasp the motivations behind our work with greater clarity.

In terms of methodology, we have provided a comprehensive overview of the methods employed in our research, ensuring transparency and reproducibility. While avoiding redundancy with previous work, we have focused on delineating the advancements made from cloud to edge real-time processing, detailing algorithm adjustments and the supporting computational infrastructure. This includes a thorough description of our experimental setup, data collection methods, and analytical procedures.

Moreover, we have accentuated the main findings of our study, presenting key results and observations derived from our analyses. Additionally, we have engaged in a discussion of these findings vis-à-vis existing literature, thus demonstrating the originality and novelty of our contributions. Notably, we have included a quantitative comparison with the suggested RPCA method for anomaly detection, thereby showcasing the superiority of our proposed approach through empirical evidence and rigorous analysis.

 

Response to Reviewer Comment 4:

We recognize the importance of moving beyond mere repetition and instead focusing on the discoveries made and the implications of our findings. In response to this valuable feedback, we have revised and expanded the "Conclusions" section of our manuscript to offer a more insightful and meaningful summary of our study.

In this updated section, we not only provide a summary of the key findings but also explore the significance of these findings for the broader field of electronic component assembly and quality control. We highlight the implications of our research in terms of advancing real-time image processing techniques for quality inspection in manufacturing environments. Furthermore, we elaborate on how our approach contributes to enhancing efficiency, reliability, and compliance with industry standards.

Response to Reviewer Comment 5:

To address these concerns, we have thoroughly reviewed the manuscript to correct typographical errors and ensure the accuracy of our citations. Additionally, we have replaced some of the figures illustrating computational stress with numerical measurements.

Furthermore, we have cross-referenced all figures with their corresponding citations in the text to ensure consistency and accuracy.

Thank you for bringing these issues to our attention, and we are committed to making the necessary revisions to enhance the overall quality of the manuscript.

Author Response File: Author Response.pdf

Round 2

Reviewer 1 Report

Comments and Suggestions for Authors

Accept

Author Response

Thank you for accepting the manuscript

Reviewer 2 Report

Comments and Suggestions for Authors

 

 

 

This reviewer did not perceive any notable enhancements in the manuscript.

To ensure that the revisions are effectively assessed, it is essential to emphasize any modifications to the manuscript.

If the methodology for image processing, which focuses on identifying counterfeits, defects, solderability problems, and corrosion, has already been documented, then it appears that the primary contribution of this study is limited to the software architecture that has been developed.

Regarding minor issues, typos still exist, such as inconsistent numbering of paragraphs. For instance, there are instances where Section 1 is followed by Subsection 2.1, Section 2, and Subsections 2.1, 2.2, 2.3, and 2.4 are repeated twice, and Section 5 is repeated.

Author Response

We appreciate the detailed feedback you provided on our manuscript. Your suggestions helped improve our work. We've carefully looked into each point you raised and made the necessary changes to make our paper better.   

Comment 1: 

This reviewer did not perceive any notable enhancements in the manuscript. 

To ensure that the revisions are effectively assessed, it is essential to emphasize any modifications to the manuscript. 

Response to comment 1: 

We genuinely appreciate the reviewer's time and effort in providing feedback on our manuscript. We have rewritten many sections in the manuscript in response to the reviewer's comments and suggestions. We'd like to highlight the extensive revisions that were undertaken in response to their comments. 

 One significant addition we made was the inclusion of a new section comparing our results with the RPCA method, as suggested by the reviewer. This addition not only strengthens the validity of our findings but also provides valuable context for the reader.  

We hope that upon reevaluation, the reviewer will recognize the efforts made to improve the manuscript and find the revised version to be more satisfactory. We genuinely value the reviewer's input and thank them for their continued engagement with our work. 

 Comment 2: 

If the methodology for image processing, which focuses on identifying counterfeits, defects, solderability problems, and corrosion, has already been documented, then it appears that the primary contribution of this study is limited to the software architecture that has been developed. 

 Response to comment 2: 

We appreciate the reviewer's insightful observation regarding the potential perception of our study's primary contribution. While it may appear that the methodology for image processing has been previously documented, we'd like to highlight the broader impact and significance of our work.  

Indeed, while the core methodology may seem incremental, the true contribution of our study lies in the innovative software architecture and ai model changes that were developed to facilitate real-time defect detection during electronic component assembly. This architecture represents a significant advancement in the field, addressing critical technological challenges and achieving notable performance improvements.  

Moreover, our approach's applicability and contribution to electronic manufacturing extend beyond the methodology itself. The developed software architecture has garnered significant interest and recognition from leading researchers and developers in the electronic SMT industry. We have received collaboration requests from prominent pick-and-place manufacturers, including ASMPT, Fuji, Yamaha, and Mycronic, following the initial preprint publication of our findings. This industry engagement underscores the groundbreaking nature of our work and its potential to revolutionize quality control processes in electronic manufacturing.  

In summary, while the methodology may build upon existing techniques, our study's primary contribution lies in the novel software architecture and its implications for advancing real-time defect detection in electronic component assembly. We believe that our work has significant implications for the industry and represents a noteworthy advancement in this domain. 

 Comment 3: 

Regarding minor issues, typos still exist, such as inconsistent numbering of paragraphs. For instance, there are instances where Section 1 is followed by Subsection 2.1, Section 2, and Subsections 2.1, 2.2, 2.3, and 2.4 are repeated twice, and Section 5 is repeated. 

 Response to Comment 3: 

Thank you for bringing these minor issues to our attention. We apologize for any inconsistencies in the numbering of paragraphs and sections. We will ensure that these typos are corrected in the final version of the manuscript.  

Round 3

Reviewer 2 Report

Comments and Suggestions for Authors

Despite the existence of several limitations that make the manuscript difficult to read, such as the unreadable Table 3, the authors have provided ample explanations for several issues. In light of these findings, I recommend a thorough revision of the paper to rectify minor typos, including the aforementioned issue in Table 3, in order to improve its overall quality.

Back to TopTop