A Systematic Review on Smart Insole Prototypes: Development and Optimization Pathways

Round 1

Reviewer 1 Report

Comments and Suggestions for Authors

This research work is unfortunately not significant and sound enough to justify its acceptance for publication. Reasoning:

1. The title emphasizes "optimizing the path", but the entire text does not systematically summarize the optimization strategies. It is suggested to add a special chapter for analyzing the optimization path.

2. The time range of the review was not specified (for example, the search was conducted until June 2025). The starting and ending years of the included studies need to be provided.

3. The font size of the VOSviewer graph in Figure 4-6 is too small, and the clustering logic is not explained.

4. By using only 3 databases (PubMed/Scopus/WoS), and omitting key databases in the engineering field (such as IEEE Xplore, Engineering Village), important literature may be missed.

5. Only 10 pieces of literature integrated AI (Table 2), but the application scenarios and advantages/disadvantages of mainstream algorithms (such as RF/SVM/LSTM) were not summarized.

6. The lightweighting and real-time performance bottlenecks of ML models on edge devices (such as insoles) have not been thoroughly explored.

7. Valvez et al. conducted research on optimizing PETG + CF, but did not explain the influence of key parameters such as layer thickness and filling rate on the performance of the sensor.

8. The core objective of "identifying optimization factors" in the introduction has not been clearly addressed. It is recommended to summarize the key optimizations.

9. The conclusion states that "AI significantly improves performance", but some studies only conducted offline analysis and did not verify the real-time effects.

Comments for author File: Comments.pdf

Author Response

Please see the attachment.

Author Response File: Author Response.pdf

Reviewer 2 Report

Comments and Suggestions for Authors

The development and optimisation of smart insole prototypes are the main topics of this systematic review, with a focus on wireless capabilities and AI integration. The authors offer a thorough overview of the body of research, point out developments in hardware, sensors, materials, and artificial intelligence methods, and suggest future paths. The review is methodologically sound and follows PRISMA guidelines. However, the writing's clarity, structure, and critical analysis's depth—particularly with regard to limitations and research gaps—need to be improved.

An excessive number of dimensions are attempted to be covered in the review, which can be overwhelming and superficial. It is advised to provide more targeted sub-sections under a common theme or to restrict the scope. Comparative evaluation tables for prototypes, trade-offs, and an explicit identification of research gaps could all help improve the review's critical analysis and comparative synthesis, which are also lacking.

A comparative table summarising ML models, dataset sizes/types, features used, accuracy, etc., discussing issues like overfitting, generalisation, and clinical validation, would improve the review's lack of a structured comparison of models used, datasets, and performance metrics.

The manuscript needs extensive language editing by a native speaker or professional service because it contains a lot of grammatical errors, awkward phrasing, and inconsistent terminology. Minor criticisms include the abstract's excessive length, the figures and tables' lack of succinct captions outlining their significance, and the terminology's need for uniformity between terms like "machine learning" and "ML," "smart insole" and "insole system," and "wireless communication" and "wireless capability."

A clear "Limitations" section to acknowledge potential biases, a graphical summary or conceptual diagram summarising the ideal smart insole architecture based on reviewed works, and an expansion of future directions to include user-centric design, ethics/privacy concerns in data collection, and commercialisation challenges are some suggestions for improvement.

Author Response

Please see the attachment.

Author Response File: Author Response.pdf

Reviewer 3 Report

Comments and Suggestions for Authors

This manuscript presents a systematic review of smart insole prototypes that incorporate artificial intelligence (AI) and wireless communication features. The authors provide a structured overview of methodologies, materials, sensors, data types, and hardware systems used in these prototypes. The review is thorough, including 39 studies sourced from three major databases and supplemented with manual selections. The authors highlight developments in AI integration, sensor technology, and material design, and discuss performance metrics and applications of smart insoles in clinical and non-clinical contexts. Future directions are proposed for optimizing these wearable devices in terms of cost, comfort, data diversity, and battery life.

Major Comments and Suggestions

1. Although the PRISMA diagram is included, the rationale behind adding 30 manually selected papers needs clearer explanation.
2. Suggestion: Provide a transparent justification for manual inclusion and ensure reproducibility by elaborating on this method.
3. The manuscript lists various machine learning methods but lacks deeper technical comparison or justification of their selection.
4. Suggestion: Include a comparative table summarizing performance metrics (e.g., accuracy, RMSE) for the ML models used in the reviewed studies.
5. There is no critical evaluation of the potential bias in included studies or limitations of the review process.
6. Suggestion: Add a dedicated limitations section discussing publication bias, sample diversity, and database limitations.
7. The readability of the manuscript is sometimes reduced due to passive constructions and lengthy descriptions.
8. Suggestion: Revise selected paragraphs for brevity and active voice, particularly in the methodology and results sections.
9. Some terms like "smart insole with AI and wireless system" are used repetitively.
10. Suggestion: Consider defining acronyms early (e.g., SIAWS) and using them consistently.
11. English Language Quality:
    The manuscript is readable, but minor grammatical errors and awkward phrasing are present throughout. Examples include:
• “This increasing popularity of smart insoles was gained due to...” → “The increasing popularity of smart insoles can be attributed to...”
• “So, performing a study with a diverse sample can be an asset...” → “Therefore, conducting studies with diverse samples may be beneficial...”
• Recommendation: Light copy-editing is advised for clarity and fluency.
12. Figures and Captions:
    Some figures (e.g., keyword density maps) lack interpretive captions.
• Suggestion: Expand captions to explain the relevance of the visualized data.
13. Citations:
• Ensure all citations are properly formatted, especially those in Section 4 referencing external technologies and materials.
14. Were any attempts made to evaluate the long-term wearability and comfort of smart insoles across different demographics (e.g., elderly, children, athletes)?
15. Do the authors envision a universal smart insole design, or will application-specific customization (e.g., for rehabilitation vs. sports) dominate future development?
16. Have you considered integrating federated learning or privacy-preserving techniques to address data sharing limitations across users or institutions?
17. How do you assess the readiness level of current AI-powered insoles for commercial deployment?
18. Can the authors elaborate on the potential ethical or data privacy concerns in the real-time monitoring of user gait and pressure data?

Comments on the Quality of English Language

The overall English is acceptable but not polished. Several areas would benefit from grammar correction, stylistic refinement, and simplification. The structure is sound, but the tone is occasionally too informal for a scientific review.
Recommendation: Moderate language editing is recommended.

 

Comments on the Quality of English Language

The overall English is acceptable but not polished. Several areas would benefit from grammar correction, stylistic refinement, and simplification. The structure is sound, but the tone is occasionally too informal for a scientific review.
Recommendation: Moderate language editing is recommended.

Author Response

Please see the attachment.

Author Response File: Author Response.pdf

Round 2

Reviewer 1 Report

Comments and Suggestions for Authors

The paper provides a comprehensive review of the prototype design of smart insoles integrating AI and wireless communication, covering materials, sensors, hardware, algorithms, and optimization paths, with clear academic and practical value. However, but there are issues such as insufficient depth in result analysis and problems with chart presentation, which require significant revision before publication.

 

1. Only the accuracy rates of some algorithms are listed, but no systematic summary of the applicable scenarios of the algorithms is provided. It is suggested to add a table classifying the algorithms and discuss the root causes of performance differences.
2. It is claimed that "3D printed capacitive insoles and PDMS capacitive sensors have the best performance" (pages 7 and 16), but no quantitative comparison indicators are provided. It is recommended to supplement quantitative evidence.
3. The vertical axis of Figure 3 (year distribution) has no label.
4. Some reference information is incomplete (e.g., Eldridge, S.; Poluan, R.; Chen, Y.-A. Using Smart Insoles and RGB Camera for Identifying Stationary Human Targets; 2019;) 。

Comments for author File: Comments.pdf

Author Response

Reply to reviewers’ comments and queries

Manuscript No. actuators-3769645

Title: A Systematic Review on Smart Insole Prototypes: Development and 
Optimization Pathways

The authors would like to express their sincere gratitude to the editor and reviewers for their careful and constructive feedback, which has contributed significantly to enhancing the quality of the manuscript. We have addressed each comment individually and made the corresponding revisions, with a strong commitment to ensuring that our responses and the updated content are as clear and concise as requested. In addition, we thoroughly revised the entire document. Where appropriate, we have also integrated our responses directly into the manuscript, as we believe this may assist future readers who might have similar questions.

We have implemented the suggestions as following:

The corresponding changes have been highlighted in yellow for clarity.

Detailed response to Reviewer #1

This research work is unfortunately not significant and sound enough to justify its acceptance for publication. Reasoning:

Only the accuracy rates of some algorithms are listed, but no systematic summary of the applicable scenarios of the algorithms is provided. It is suggested to add a table classifying the algorithms and discuss the root causes of performance differences.

We sincerely thank the reviewer for this thoughtful observation. We acknowledge the importance of providing a more systematic classification of the algorithms and a deeper discussion of the factors influencing their performance.

However, at the time of writing, we did not have access to sufficient information across all relevant dimensions, such as data structure, model complexity, or specific usage contexts, to confidently present such a classification. Additionally, we opted not to include an additional table summarising algorithm applicability to avoid making the article more extensive and potentially unbalanced.

We also considered that conducting a detailed breakdown of algorithmic performance would require a parallel analysis of other equally relevant variables, such as the materials and structural configurations used in the manufacturing of the insole prototypes. Given the scope of the present study, which aimed to provide a general overview rather than a comprehensive benchmarking analysis, we chose to focus on reporting the most representative accuracy results available.

We fully agree that this is a valuable direction for future work, and we appreciate the reviewer’s insight and believe this comment has helped us clarify the scope and limits of our current contribution.

It is claimed that "3D printed capacitive insoles and PDMS capacitive sensors have the best performance" (pages 7 and 16), but no quantitative comparison indicators are provided. It is recommended to supplement quantitative evidence.

We thank the reviewer for the valuable observation. The statement regarding the superior performance of 3D-printed capacitive insoles and PDMS-based capacitive sensors is supported by quantitative data reported in benchmark studies, notably by Samarentsis et al. (2022) and Luna-Perejón et al. (2023). We made additional changes as suggested by the reviewer and included data shared by the author in their studies (183-190 lines).

The vertical axis of Figure 3 (year distribution) has no label.

Thank you for your observation. As it was mentioned, we made the suggested changes and include also the label in the horizontal axis (between 114 and 115 lines).

Some reference information is incomplete (e.g., Eldridge, S.; Poluan, R.; Chen, Y.-A. Using Smart Insoles and RGB Camera for Identifying Stationary Human Targets; 2019;).

Thank you again for your scrutiny. We add the missing information in the document (327 and 328 lines).

Author Response File: Author Response.pdf

Reviewer 2 Report

Comments and Suggestions for Authors

The manuscript was accurately revised.

Author Response

The authors would like to express their sincere gratitude to the editor and reviewers for their careful and constructive feedback, which has contributed significantly to enhancing the quality of the manuscript. We have addressed each comment individually and made the corresponding revisions, with a strong commitment to ensuring that our responses and the updated content are as clear and concise as requested. In addition, we thoroughly revised the entire document. Where appropriate, we have also integrated our responses directly into the manuscript, as we believe this may assist future readers who might have similar questions.

Reviewer 3 Report

Comments and Suggestions for Authors

Thank you for addressing the comments.

Author Response

The authors would like to express their sincere gratitude to the editor and reviewers for their careful and constructive feedback, which has contributed significantly to enhancing the quality of the manuscript. We have addressed each comment individually and made the corresponding revisions, with a strong commitment to ensuring that our responses and the updated content are as clear and concise as requested. In addition, we thoroughly revised the entire document. Where appropriate, we have also integrated our responses directly into the manuscript, as we believe this may assist future readers who might have similar questions.

Round 3

Reviewer 1 Report

Comments and Suggestions for Authors

None