Vision-Based UAV Landing with Guaranteed Reliability in Adverse Environment

Round 1

Reviewer 1 Report

I send the review in the attachment.

Comments for author File: Comments.pdf

Author Response

Please see the attachment

Author Response File: Author Response.pdf

Reviewer 2 Report

The subject of this paper is interesting and may attract researchers' attention. This paper is easy to read. I have some suggestions for authors to improve the presentation of this paper as follows. 

(1) After reading this paper, I found that the research question of this paper is challenging and very important in practice.. However, I could not find any description about contributions of this paper. A paragraph about contributions of this paper should be added to pin point the challenging issues.  

(2) Development of a method to respond to wind disturbance is an important issue in the control of UAV landing. In control theory, a wide variety of robust control methods had been proposed and used in real application scenarios. In the context of robust control theory, a robust controller is designed based on an uncertainty model and the  robust controller must be able to stabilize the controlled system to achieve the goal under the uncertainty model. Although the results presented in 4.2.2. (Case 2: Landing with slight wind disturbance) and 4.2.3. (Case 3: Landing with strong wind disturbances) indicate that the proposed PID controller seem to be effective for dealing with disturbances, it is vague for readers to understand "slight wind disturbance" and "strong wind disturbance".  I suggest the authors to describe "slight wind disturbance" and "strong wind disturbance" quantatively and formally just like the .the uncertainty model in robust control theory. That is, please define a "slight wind uncertainty model" for  "slight wind disturbance" and a "strong wind  uncertainty model" for  "strong wind disturbance".

Based on "slight wind uncertainty model" and "strong wind  uncertainty model", the results presented in 4.2.2. (Case 2: Landing with slight wind disturbance) and 4.2.3. (Case 3: Landing with strong wind disturbances) should be stated as follows:

Case 2 is an instance of "slight wind uncertainty model". The results indicate that proposed method works for "slight wind uncertainty model". 

Case 3 is an instance of "strong wind  uncertainty model". The results indicate that proposed method works for ""strong wind  uncertainty model". 

(3)In Section 5 (Conclusion and discussion), the authors said "According to the experimental results, the landing was smooth and accurate with a 100% landing success rate". This is because the wind disturbance encountered is not strong enough.

I believe that no controller can work effectively and deal with every types of uncertainty or disturbances. No one can claim there exists a controller that works for every situation. Each control design has its limitations. I believe there must be some strong wind disturbance that lead to failures in UAV landing using any types of controllers including the one proposed in this paper. 

If possible, I suggest the authors to use a negative example (failures in landing using the proposed controller) to illistrate the above point. This will provide a good future research direction for more robust UAV landing.

 

 

Author Response

Please see the attachment

Author Response File: Author Response.pdf

Reviewer 3 Report

Reviewer’s report on a paper submitted to the journal of MDPI Electronics, with the title of “Vision-based UAV landing with guaranteed reliability in adverse environment”

 

The paper proposed a reliable vision-based landing approach for landing of UAVs on a multi-level platform mounted on a UGV. Real-world flight tests were conducted to evaluate the effectiveness of the proposed algorithm. The paper’s subject is interesting and it is written in overall well. The paper provides good theoretical background and sufficient practical applications. The reviewer would like to provide some points for authors to improve their work before it can be considered further:

-        Abstract can be expanded by one or two sentences about the contribution of the research. The research gaps are stated (i.e., a challenging task when the altitude of landing target is different from the ground or when the UAV is working in adverse environments) and it’ll be good to add one sentence about the contribution of the work too.

-        Section 1 (introduction) is written well; however, the literature review is not comprehensive. The literature papers are not reviewed thoroughly. The authors can review briefly some of the recently published important papers/theses and outline the advantages and limitations of the methods they have adopted. Some more papers from the years 2023, 2022 and 2021 can be added to the reference list. Some examples are:

Xin, L.; Tang, Z.; Gai, W.; Liu, H. Vision-Based Autonomous Landing for the UAV: A Review. Aerospace 2022, 9, 634. https://doi.org/10.3390/aerospace9110634

 

Bektash, O., Naundrup, J.J., and la Cour-Harbo, A. Analyzing visual imagery for emergency drone landing on unknown environments. International Journal of Micro Air Vehicles, 2022, 14, https://doi.org/10.1177/17568293221106492

Piponidis, M.,, Aristodemou, P., and Theocharides, T. Towards a Fully Autonomous UAV Controller for Moving Platform Detection and Landing, 2022. https://arxiv.org/pdf/2210.08120.pdf

Alam, Md S., Oluoch J. A survey of safe landing zone detection techniques for autonomous unmanned aerial vehicles (UAVs). Expert Systems with Applications 2021, 179, 115091.

Pluckter, K. Precision UAV Landing in Unstructured Environments. MSc thesis, Carnegie Mellon University, July 2019.

-        Section 2 introduces the developed UAV vision-based target detection and safe landing system. However, more details are required about the protocols followed, different components being designed/developed, and hardware and software specifications of the components.

-        Section 4 is perhaps the most interesting part of the paper. However, it lacks discussion. Many readers would love to see a good discussion out of such interesting results. Making some comparisons (and discussing the results of comparisons) for the performance of the proposed algorithm could be insightful.

-        Some future directions for research can be included in Section 5.  

Author Response

Please see the attachment

Author Response File: Author Response.pdf

Round 2

Reviewer 3 Report

The authors have successfully addressed the comments and implemented the required changes to the revised version. Hence, the reviewer recommends the paper to be accepted.