A Hybrid Microstructure Piezoresistive Sensor with Machine Learning Approach for Gesture Recognition
Round 1
Reviewer 1 Report
In this manuscript, the authors presented a comprehensive work that combines a cost-effective, fabricated tactile sensor with a machine learning approach for gesture recognition in the context of robotics surgery. There is valuable work behind the paper with the novel scientific principle. It is suitable for most of the work focused on the sensor electrical and piezoresistive characterization, sensing performance, and practical application demonstrations.
- The paper is technically sound and clearly explained.
- The structure of the paper is organized very well
- Few minor paragraphs should be improved, such as typesetting of the manuscripts,
- Figures should be controlled and unified Captions in the body of the paper.
It is novel work and interesting, so I highly recommend this article for publication, considering my suggestions to improve the article.
Author Response
Thank you for your invaluable comments and suggestions that inspired us to improve our article. We highly appreciate all the comments you have raised. We have improved the typesetting and used consistent formatting for all figures in the manuscript. Further, we have improved figure layouts and captions to ensure that they do not contract with text in the revised manuscript.
Author Response File: Author Response.docx
Reviewer 2 Report
In this manuscript, Al-Handarish et al. reported a carbon black loaded PDMS sponge used as a piezoresistive sensor.
Its prototype sensing system was developed with machine learning method for gesture recognition in clinical applications.
This work is a follow-up of the study Nanomaterials 10, 1941, and I would support publication if the authors can address the following comments.
1. Could the authors also show SEM images of neat PDMS sponge and CB? Fig. 1b cannot provide sufficient information on material morphology.
2. Why thickness of 400 um was chosen? Is there a relationship between sensor performance and thickness?
3. And also, is there a relationship between sensor performance and environmental temperature?
4. In Fig.3, please explain during the same period of time, why the number of peaks in 10 kHz is not ten times that of 1 kHz?
5. In Table 1, I cannot find significant advantage of LSTM + Dense, compared to MLP, or even KNN. Please explain more on this issue.
6. If possible, please prepare a schematic drawing to show your setup for the robotic sensing system in Fig. 6.
7. I wonder there are some problems in Fig.7, as plots shown in a and b (also c and d) are totally overlapped. Please have a check on it.
8. Please be careful the typos through the manuscript, such as page 3, line 105; page 7, line 220; page.
Author Response
We thank the reviewer for summarizing the paper’s contribution and this constructive comment. Your suggestion is highly appreciated and we updated the manuscript accordingly.
Author Response File: Author Response.docx