Review Reports - Research on Dynamic Center-of-Mass Reconfiguration for Enhancement of UAV Performances Based on Simulations and Experiment

Round 1

Reviewer 1 Report

Comments and Suggestions for Authors

This paper presents a proposal to dynamic center of mass reconfiguration for UAV stabilization. This is a challenge for critical safety operations in multirotor design. The authors successfully integrate mechanical actuation, control modeling, and experimental validation.

The manuscript is clearly written, technically sound, and supported by strong simulation and experimental results. Overall, this is a high-quality contribution. The paper is ready for publication after only very minor revisions that do not affect the technical content or scientific quality of the work.

Minor remarks:

Figure 11: “roll angel” should be “roll angle.”
Figures 17 and 18: “25-meter Hight” and “15-meter Hight” , should be “Height.”
Line 465. Table 2, must be table 1.Table line 519. IDEM.
Line 154, please check: “𝑚𝑠 ≈ 0.20ms≈0.20 kg”

Please, check the maximum torque values

0.039 N·m
1.2375 N·m
0.078 N·m

Section 4.3.2: “1800mAh (1.8A per cell)” clarify units.Reference 17 missing publication year and source.
Reference 24 lists only author names (“Wiley, J. Mark W. Spong…”)

Author Response

Comment 1: “Figure 11: ‘roll angel’ should be ‘roll angle.’
Figures 17 and 18: ‘25-meter Hight’ and ‘15-meter Hight’ should be ‘Height.’
Line 465: Table 2 must be Table 1. Table line 519: idem.
Line 154: please check: ‘mₛ ≈ 0.20ms ≈ 0.20 kg’.”

Response 1: Thank you very much for these observations. We agree with the reviewer’s comments. All typographical mistakes have been corrected in the revised manuscript. Specifically, “roll angel” was changed to “roll angle” in Figure 11, the word “Hight” was corrected to “Height” in Figures 17 and 18, the incorrect table numbering was fixed on lines 465 and 519, and the formatting error in line 154 (“mₛ ≈ 0.20 kg”) was corrected to its proper form.

These corrections have been implemented on page 19, 20, 21, 27, 29, 30, 34

lines 621, 624, 639, 640, 650, 673, 831, 908, 910, 915, 917, 927, 929, 1070-1086 of the revised manuscript.

Comment 2: “Please check the maximum torque values: 0.039 N·m, 1.2375 N·m, 0.078 N·m.

Section 4.3.2: ‘1800mAh (1.8A per cell)’ — clarify units.

Reference 17 missing publication year and source.

Reference 24 lists only author names (‘Wiley, J. Mark W. Spong…’).”

Response 2: We appreciate the reviewer’s detailed feedback. All issues listed have been corrected in the revised version. The maximum torque value was re-verified and updated for consistency, with the value 1.2375 N·m in line 556,561,772,905. The battery specification in Section 4.3.2 was clarified (replacing “1.8A per cell” with the correct unit description), Reference 17 was completed by adding its publication year and source, and Reference 24 was corrected to the full citation for Robot Modeling and Control by Spong, Hutchinson, and Vidyasagar.

Author Response File: Author Response.pdf

Reviewer 2 Report

Comments and Suggestions for Authors

Dear Authors,

my compliments for your paper: it is interesting and shows a little of originality. The modeling and simulations are well conducted, and the methodologies are used correctly. The experimental part adds completeness to the work.

My criticism is about the excessive length of the document: some concepts are repeated too many times, I think it is appropriate to summarise some parts, avoiding repetitions. The methodologies used are almost identical in the various cases (roll stabilization, free fall, etc.). I suggest avoiding repetition and showing the significant results.

Author Response

Comment:“My criticism is about the excessive length of the document: some concepts are repeated too many times. I think it is appropriate to summaries some parts, avoiding repetitions. The methodologies used are almost identical in the various cases (roll stabilization, free fall, etc.). I suggest avoiding repetition and showing the significant results.”

Response:Thank you very much for this constructive and thoughtful comment. We fully agree with the reviewer’s observation regarding unnecessary repetitions and overly detailed explanations of similar methodologies. In the revised manuscript, we have carefully shortened multiple sections, removed redundant descriptions, and consolidated repeated explanations of the methodology. The presentation of results has been streamlined to focus only on the essential findings, ensuring a clearer and more concise narrative.

The revised version now presents the experimental and simulation results without duplication, highlights the significant outcomes more effectively, and improves the overall readability and flow of the paper. These changes have been implemented throughout the manuscript, particularly in the Introduction, Methods, Results, and Discussion sections

Author Response File: Author Response.pdf

Reviewer 3 Report

Comments and Suggestions for Authors

The idea, which is already explored extensively in satellites, is great and could be a very interesting UAV tech in the future. However, the paper needs some work.

Fig 13 part (b) shows higher amplitudes of oscillations when the device is enabled. Authors, please explain what you mean by "The enhanced dampening effect" in Line 579. I see more oscillations in part (b) compared to part (a: when the device is not enabled).

In Figs 18 and 17, why are the yellow lines are called z-axis? Do they mean the yaw motion? If yes, why can't the yaw be controlled well, like pitch and roll?

The discussion section is confusing. Why are the authors citing other papers when talking about their own results, for example in line 763, when citing [2]. This happens multiple times. Are these their own results or someone else’s?! (Also line 428)

Other comments:

What are the units of Figures 18 and 17, especially the vertical axis? Are they in degrees?

Conclusion is just repetition of Discussion. Remove one of them.

The title should emphasize the simulation; this was not built and tested, rather, it was simulated in MATLAB. This should be reflected in the title.

Fig 12, and Fig 7 need labels that identify different parts. Where is the proposed system on that?

Figures and tables numbers need to be checked. Some figures are hard to see (low quality and resolution, like Fig 5)

Figs 9 and 10 not legible. Move to appendix, make bigger.

Why the red background in Fig 13? It’s not easy on the eyes. Also, hard to read the axis labels.

Instead of Fig 13 and Figs 18 and 17, I suggest the authors plot the attitude errors (actual minus desired) for both of the cases (with and without CoM shifting device). This way, they can see if/how their system is enforcing errors to zero.

I’m interested to see a video of their test setup (section 5.4) in action.

The writing needs some work. I suggest the authors significantly revise the language from A to Z. In a paragraph, we see a different font size. We see sentences with no caps in the beginning. We see incorrect use of caps (like in the word ATTITUDE; doesn’t need to be in caps.) CoM is written in different ways throughout the text. Sometimes acronyms are defined multiple times (like UAV), sometimes they aren’t defined at all (like FC MSP). Section titles sometimes not in caps (Like Sec. 2). Redundant figures (like 1(a) and 3). Where is tau_theta defined? Leave space before citing something like this [2], not this[2]. What is Line 470?! What are Lines 516, 647??! Labels cutoff in Fig 11. Fig 15 too small, low quality to read. References not in order of appearance. Why is L 787 has cited [4] but not line 525, stating the same information?

Comments on the Quality of English Language

Writing needs to be improved. See comments.

Author Response

Comment 1:
“Fig 13 part (b) shows higher amplitudes of oscillations when the device is enabled. Authors, please explain what you mean by "The enhanced dampening effect" in Line 579. I see more oscillations in part (b) compared to part (a: when the device is not enabled).

Response 1:
Thank you for pointing this out. We agree with the reviewer’s observation. The original wording (“enhanced dampening effect”) was misleading because Figure 13b indeed exhibits higher oscillation amplitudes when the CoM-shifting device is active. Therefore, we have revised the text to clarify that the improvement lies not in reducing oscillation amplitude, but in shortening the settling time and increasing the recovery rate following disturbances.

This clarification has been incorporated in Section 5.3, page 21, lines 663–681 of the revised manuscript.
The updated text is as follows (marked in turquoise in the manuscript)

Comment 2:
“In Figs 18 and 17, why are the yellow lines are called z-axis? Do they mean the yaw motion? If yes, why can't the yaw be controlled well, like pitch and roll?”

Response 2:
Thank you for this comment. We confirm that the yellow curve represents the yaw angle. In this experiment, yaw was not actively controlled because the CoM-shifting mechanism physically affects only roll and pitch. Since lateral mass translation cannot generate any restoring moment around the yaw axis, yaw stabilization was outside the scope of this test. The yellow trace is included only for completeness to show the full 3-axis IMU output; however, the performance evaluation in this section focuses exclusively on roll and pitch, which are the axes directly influenced by the CoM-shifting device. This clarification has been added to the revised manuscript .

Comment 3:
“The discussion section is confusing. Why are the authors citing other papers when talking about their own results, for example, in line 763 when citing [2], This happens multiple times. Are these their own results or someone else’s?! (Also line 428)”

Response 3:
Thank you for pointing this out. We agree with the reviewer’s observation. The cited references in these lines were inserted unintentionally and incorrectly, which made the discussion appear as if our experimental results were taken from external sources. These results are entirely our own.

Therefore, we have removed the misplaced citations and revised the text to clearly indicate that the described findings originate from our experiments. The corrections have been implemented in the Discussion section and the Conclusion section in the revised manuscript (page 28, lines 859–954).

Comment 4:
“What are the units of Figures 17 and 18, especially the vertical axis? Are they in degrees?”

Response 4:
Thank you for the comment. We apologize for the missing unit annotation. The vertical axis in Figures 14 and 15 represents angular displacement in degrees (°) for the roll, pitch, and yaw axes. We have updated the figure captions and axis labels accordingly to clearly indicate this information. The revised figures and descriptions are now included in the manuscript (page 27, lines 849,856).

Comment 5:
“Conclusion is just a repetition of Discussion. Remove one of them.”

Response 6:
Thank you for this valuable observation. We agree that in the original submission, the Discussion and Conclusion sections contained overlapping content. In the revised manuscript, both sections have been substantially rewritten to eliminate redundancy. The discussion has been condensed to focus solely on the interpretation and implications of the results, while the conclusion has been shortened and reformulated to present only the key final statements and forward-looking perspectives.

Both sections are now significantly more concise, clearly differentiated in purpose, and free of repetition. These improvements have been implemented on pages 28–29 of the revised manuscript.

Comment 6:

“The title should emphasize the simulation; this was not built and tested, rather, it was simulated in MATLAB. This should be reflected in the title.”

Response 6:

Thank you for the comment. We agree that the original title did not clearly indicate that the results were obtained through MATLAB simulations rather than physical flight tests. Therefore, we have updated the title to explicitly reflect the simulation-based nature of the study. The revised title now reads:

Research on Dynamic Center-of-Mass Reconfiguration for Enhancement of UAV performances based on simulations and Experiment

Comment 7:
“Figs. 9 and 10 not legible. Move to appendix, make bigger.”

Response 8:
Thank you for pointing this out. We agree with the reviewer that the original Figures 9 and 10 were not sufficiently legible in their initial size. In the revised manuscript, both figures have been enlarged for clarity and relocated to the Appendix as requested. Their captions have also been updated accordingly. These changes improve the readability of the figures without interrupting the flow of the main text.

The updated figures can now be found in Appendix A, Figures A1, A2, and A3 (pages 31–32) of the revised manuscript.

Comment:
“Why the red background in Fig 13? It’s not easy on the eyes. Also, hard to read the axis labels.”

Response:
Thank you for the comment. Figure 13 has been redesigned with a neutral blue background, and all axis labels have been enlarged for clarity. The updated figure is now easier to read and visually consistent with the rest of the manuscript.

Comment:

The writing needs some work. I suggest the authors significantly revise the language from A to Z. In a paragraph, we see a different font size. We see sentences with no caps in the beginning. We see incorrect use of caps (like in the word ATTITUDE; doesn’t need to be in caps.) CoM is written in different ways throughout the text. Sometimes acronyms are defined multiple times (like UAV), sometimes they aren’t defined at all (like FC MSP). Section titles sometimes not in caps (Like Sec. 2). Redundant figures (like 1(a) and 3). Where is tau_theta defined? Leave space before citing something like this [2], not this[2]. What is Line 470?! What are Lines 516, 647??! - in Fig 11. Fig 15 too small, low quality to read. References not in order of appearance. Why is L 787 has cited [4] but not line 525, stating the same information?

Response:

Thank you for these detailed remarks. We have completely revised the manuscript to address all the issues highlighted. The language has been edited thoroughly from beginning to end, ensuring consistency, clarity, and proper academic style. All font-size inconsistencies, capitalization errors, and spacing issues have been corrected. The notation “CoM” is now written uniformly throughout the manuscript using Cambria Math formatting. All acronyms are defined once at first appearance and used consistently thereafter, including UAV, FC, MSP, etc.

The symbol of tau_theta has now been properly introduced and defined in the Methods section. Citation spacing and formatting have been corrected, and the previously unclear sections have been fixed. Figures have been fully redesigned at higher resolution and improved readability. Finally, the reference list has been reordered and updated to follow the correct order of appearance and complete citation format. These revisions appear throughout the updated manuscript.

Comment:

"I’m interested to see a video of their test setup (section 5.4) in action."

Response:
The video has been uploaded to Google Drive through this link:
https://drive.google.com/file/d/1Z4GQW28Q_ZMzLdbZVfK5Kmnnu03KdTuJ/view?usp=drive_link

Author Response File: Author Response.pdf

Reviewer 4 Report

Comments and Suggestions for Authors

In this paper, the authors investigates the problem of dynamic center-of-mass reconfiguration for enhanced UAV free-fall stability, structural protection, and energy efficiency. Although the paper has conducted good simulations and experiments, it is not well written in terms of innovation and structure. In addition, the structure of the paper needs to be adjusted and the text needs to be concise. The following comments suggest areas for improvement.

1.The structure figure corresponding to the actual system may be given, which will be facilitate for the readers.

2.What are the main difficulties and theoretical innovations in the article? What are the advantages compared to existing ones?

3.Some key content is not specific, such as how to select and design controller parameters in the control method?

4.How to perform CoM reconfiguration should be written in detail.

5.I suggest the authors to give a detailed method and innovation for this system, rather than just providing simulation and experimental results, and revising the paper sufficiently before acceptance.

Author Response

Comment 1: “The structure figure corresponding to the actual system may be given, which will facilitate the readers.”

Response 1: Thank you for the suggestion. In response, we have substantially expanded the structural description of the system. In sections 2.1， we have add Figure 1(a) to show the overall 3D structure of our actual system. And in section 3.10, the figure 7 also shows the actual system we used for experimentation

Comment 2: “What are the main difficulties and theoretical innovations in the article? What are the advantages compared to existing ones?”

Response 2: We appreciate the reviewer’s question and have added explicit explanations of both the technical challenges and the theoretical innovations of the work. These are now clearly addressed in Section 1.3. We discuss the challenges of stabilizing a multirotor under zero-thrust conditions, the need to generate gravity-based corrective torques, configuration-dependent inertia modeling, actuator nonlinearities, and feedback stability. The theoretical innovations—including the integration of gravity-induced moments, electromechanical stage dynamics, dual-axis mass shifting, propulsion-independent attitude control, and high-fidelity modeling—are now clearly explained. We also added a new subsection outlining the advantages over existing methods, including lower energy consumption, full functionality during thrust loss, and faster CoM displacement.

Comment 3: “Some key content is not specific, such as how to select and design controller parameters in the control method.”

Response 3: We thank the reviewer for this observation. A new subsection, 2.5 (Controller Parameter Selection and Tuning Methodology), has been added. It provides a step-by-step explanation of how PID gains were selected from the linearized hover model, how damping ratios were chosen, how actuators were saturated based on physical travel limits, how low-pass filters were tuned, and how the deadband was calibrated experimentally. This detailed methodology clarifies the complete parameter-selection process for reproducibility.

Comment 4: “How to perform CoM reconfiguration should be written in detail.”

Response 4:
Thank you for the insightful comment. We have added an entirely new section, 2.6 (Detailed Procedure for CoM Reconfiguration). This subsection explains the real-time CoM-shifting process in four consecutive stages: sensing, filtering and error computation, conversion to displacement commands, and execution via the XY prismatic stage. It also includes details about actuator limits, stepper-driver operation, FreeRTOS task structure, and telemetry monitoring, fully addressing the reviewer’s request for a detailed operational description.

Comment 5: “I suggest the authors give a detailed method and innovation for this system, rather than only results.”

Response 5: We appreciate this recommendation. In response, we have significantly expanded the methodological explanation of the system. Sections 1.3 and 2.6 now provide a complete description of the system’s design methodology, mechanical structure, control algorithms, controller tuning, actuator implementation, and system-level innovations. These additions clearly outline the novelty, detailed workflow, theoretical basis, and engineering contributions of the proposed CoM-shifting stabilization device, ensuring that the paper presents not only simulation and experimental results but also the full underlying methods and innovations.