Improving FPGA Based Impedance Spectroscopy Measurement Equipment by Means of HLS Described Neural Networks to Apply Edge AI
, Rafael Gadea-Gironés *,†, Jose M. Monzo †, Ángel Tebar-Ruiz † and Ricardo Colom-Palero †
Round 1
Reviewer 1 Report
1.A large number of abbreviations are used in the draft without giving the full name, which affects the reading comprehension of the paper. Many acronyms only appear once and should be given their full names directly. It is suggested to give the full name in the form of a table to help readers read.
2.As mentioned in the literature (doi.org/10.1016/j.automatica.2021.110087 and 10.1109/TAC.2019.2920837) , network embedding will affect the information security of the system. Please discuss how to protect the privacy security of the system.
3.The picture and its text description should be placed on the same page as much as possible.
4.The document format needs to be unified.
Author Response
1. A large number of abbreviations are used in the draft without giving the full name, which affects the reading comprehension of the paper. Many acronyms only appear once and should be given their full names directly. It is suggested to give the full name in the form of a table to help readers read.
Response: We have included the full names of all abbreviations in all three contexts: abstract, text, and figures/tables.
2. As mentioned in the literature (doi.org/10.1016/j.automatica.2021.110087 and 10.1109/TAC.2019.2920837) , network embedding will affect the information security of the system. Please discuss how to protect the privacy security of the system.
Response: We have added a mention of safety aspects in the introduction to the article. Fortunately, the industrial applications where we want to apply these impedance analyzers with AI are not as problematic as those mentioned in the above-mentioned articles; but we have seen fit, because of its importance, to make the brief review that we have included in the introduction.
" The developed system is a prototype that allows us to verify the feasibility of the technique in an industrial environment. In order to design a definitive system for an industrial environment, the security of the system should be improved to protect the system from potential threats. Using a zynq-based system allows us to perform these security features \cite{trimberger2014fpga}."
3. The picture and its text description should be placed on the same page as much as possible.
4. The document format needs to be unified.
Response: we have managed to improve the formatting to ensure that the explanatory text of each figure is as much as possible on the same page.
With respect to English, we have revised the whole article.
Attached pdf to highlight the differences with the first version we sent for review
Author Response File: 
Author Response.pdf
Reviewer 2 Report
1. Figure-x is not understandable, enough information about the figure is not given in the text. It would be nice to redraw it.
2. Sufficient and up-to-date reference is cited.
3. The objectives are clearly stated.
4. The method they used is not explained clearly and understandably enough.
Author Response
- Figure-1 is not understandable, enough information about the figure is not given in the text. It would be nice to redraw it
Response: We have divided the figure in two and have included more information in the text.
2. Sufficient and up-to-date reference is cited.
3. The objectives are clearly stated.
4. The method they used is not explained clearly and understandably enough.
We have expanded the explanation of the method in section 3 and subsection 2.3.
Attached pdf to highlight the differences with the first version we sent for review
Author Response File: Author Response.pdf
Reviewer 3 Report
This is a good work in the field that try to improve the equipment base on impedance spectroscopy measurement using AI implemented in FPGA.
The paper is organized on 5 chapters, starting with a extensive introduction. Then is presented the architecture of the system implemented in FPGA and followed by optimization and improves. Those are sustained by results and conclusions.
I think isn’t necessary to introduce the code in performance evaluation. Also, I think, the table 4 goes to the reference paragraph.
Author Response
Thank you very much for the review. We have tried that the figures are explained as much as possible in the same page where they are introduced and that the tables and algorithms are also close to the sections where they are mentioned and explained.
Round 2
Reviewer 2 Report
Thanks for the corrections.
