A Generalized Simulation Framework for Tethered Remotely Operated Vehicles in Realistic Underwater Environments
Round 1
Reviewer 1 Report
Paper presents an interesting realistic simulation environment for underwater environment. In order to improve the paper, I have the following comments:
As authors said in page 3, their simulation will be reconsidered in a huge approach in they increase the interaction with the environment, like including a robot arm. How big will be the reconsideration required? How much limit the uses of the simulation that point?
In page 5, authors use a box in order to compute the inertial matrix, how much will be different the results if they compute the real one?
In equation 9 authors do not define ‘v’.
Start describing the ROV simulation then the Cable simulation. For the ROV they provide the example of one specific, how can be extrapolate to other underwater robots with different parameters? It is necessary to rewrite all the equations for a new robot or only with the main parameters are enough?
In Section 4.3 authors said:
- “The experiments were done in the editor environment…” that means that it cannot be used independent of the editor environment? Using in editor or release
environment will affect the computation time of the simulation, in which factor?
- “A live video demo can be seen in… Figure 9”. In Figure 9 only one image of the simulator is show. That’s interesting but authors can include a sequence of several frames of the video to visualize the movement effect.
To sum up, authors present a simulation environment for an specific task, it is not clear how to generalize or utilize it by others researchers in order to obtain a simulation of underwater environment. They provide more a application example of their underwater simulation than a research, which is exactly the research novelty of this system? Is only a specific simulation? Or it can be easily parametrized in order to perform other simulations.
Finally, in Introduction section, authors present several underwater simulators. They do not provide a comparison among them with their propose, in which points is better or worsts that these other simulations?
Author Response
Our comments in blue
Paper presents an interesting realistic simulation environment for underwater environment. In order to improve the paper, I have the following comments:
As authors said in page 3, their simulation will be reconsidered in a huge approach in they increase the interaction with the environment, like including a robot arm. How big will be the reconsideration required? How much limit the uses of the simulation that point?
Our model uses assumptions of a basic rigid body dynamics model and additional properties can be added if required. We have now mentioned this aspect of our model in Section 1.4 (lines 75-77)
In page 5, authors use a box in order to compute the inertial matrix, how much will be different the results if they compute the real one?
The results of the simulation will be different when the inertia terms are changed. If accurate measurements of inertia terms are known, then can be added to the simulation model in Eq.(8)
In equation 9 authors do not define ‘v’.
We have added the definition of 'v' in the text.
Start describing the ROV simulation then the Cable simulation. For the ROV they provide the example of one specific, how can be extrapolate to other underwater robots with different parameters? It is necessary to rewrite all the equations for a new robot or only with the main parameters are enough?
It is not necessary to rewrite all the equations, as explained in Section 3.2. Our model is highly configurable. For example, changing the position of the thruster will involve only changing the L point referred by eq 11. Adding another thruster will involve only adding another component to eq 11 which is basically the position of the thruster and a vector which defines the orientation of that thruster. Similarly, any change in inertia model assumptions could be incorporated in Eqs. (6)-(8).
In Section 4.3 authors said:
- “The experiments were done in the editor environment…” that means that it cannot be used independent of the editor environment? Using in editor or release
environment will affect the computation time of the simulation, in which factor?
The editor environment was used only for convenience. In advanced game engines such as Unreal, you can test visual concepts without the need to compile a packed game for every change. In our case, in order to demonstrate the behaviour of the cable we used the editor environment which allows us to move the ROV with the mouse and see how the cable behaves regardless the ROV dynamics. In terms of performance since the cable is mainly linear object (not like cloth) not much GPU or CPU power is needed in the editor environment nor did the packed game.
- “A live video demo can be seen in… Figure 9”. In Figure 9 only one image of the simulator is show. That’s interesting but authors can include a sequence of several frames of the video to visualize the movement effect.
We changed the wording to ".. a snapshot is given in Figure 9" (line 350). The purpose of this image was to explain the experimental setup presented in the videos.
To sum up, authors present a simulation environment for an specific task, it is not clear how to generalize or utilize it by others researchers in order to obtain a simulation of underwater environment. They provide more a application example of their underwater simulation than a research, which is exactly the research novelty of this system? Is only a specific simulation? Or it can be easily parametrized in order to perform other simulations.
To address the issues you mentioned we did the following modifications:
To explain the novelty of our research, we expanded Section 1.3 "related work" subsection to explain the current methods used in the computer vision underwater research which involves mainly highly controlled environments such as swimming pools and we explained the advantages of our approach e.g realistic simulation.
In subsection 1.4 "Our Contribution" we emphasized the fact that we compared computer vision experiments (such as feature tracking) in our realistic simulation compared to the real environment. To the best of our knowledge, that has not been done yet for the underwater environment.
Finally, in Introduction section, authors present several underwater simulators. They do not provide a comparison among them with their propose, in which points is better or worsts that these other simulations?
In section 1.3 "related work" we presented the main problems with the existing family of solutions which make them not usable for computer vision research. The open source family which is not realistic enough for computer vision research and the closed source simulators which are created for the purpose of personnel training for usually a specific task with a specific ROV due to the high costs of operating ROVs on top of that these are closed source which makes them inaccessible for purpose of open research as explained in the subsection. We think that the explanation of these disadvantages for the two types of families is enough to disqualify them from being suitable for the computer vision underwater research.
Reviewer 2 Report
Minor grammar changes would definitely make this a stellar report. A few more commas are needed in your expanded thoughts. Use or lack of plurality for singular items and incorrect pronouns ("who" used for a software package) tend to sound unprofessional.
Some of the recreated images are amazing! Replication of dirt mask is great!
Something about the way your figures are aligned, or sized, or setup on some of the pages seems odd. Maybe it is just a lot of wasted space in the draft which I know will be cleaned up, or perhaps they could be arranged differently. I feel like we lose some of the flow of the paper as the figures appear.
I am definitely not a software expert, but the approach to simulation for the drone world is of definite interest!
Author Response
Our comments in blue
Minor grammar changes would definitely make this a stellar report. A few more commas are needed in your expanded thoughts. Use or lack of plurality for singular items and incorrect pronouns ("who" used for a software package) tend to sound unprofessional.
We fixed some of the grammar errors we found and added a few more commas where needed
Some of the recreated images are amazing! Replication of dirt mask is great!
Thank you, just for clarification, our final simulation is a combination of assets we bought from the marketplace (ref 39) together with assets we added such as the plankton particles and mist.
Something about the way your figures are aligned, or sized, or setup on some of the pages seems odd. Maybe it is just a lot of wasted space in the draft which I know will be cleaned up, or perhaps they could be arranged differently. I feel like we lose some of the flow of the paper as the figures appear.
We rearranged the images besides the ones that take the entire page. We also assume that the positions will change in the final version.
I am definitely not a software expert, but the approach to simulation for the drone world is of definite interest!
Reviewer 3 Report
Please find my comments in the document attached.
Comments for author File: 
Comments.pdf
Author Response
Our comments in blue
In this paper, the authors developed a framework for simulating the motion of remotely operated
underwater vehicles in realistic environments. The overall quality of the manuscript is good and the topic addressed in this investigation is of interest for this important journal. However, before being considered for publication, I recommend solving the issues described below (Minor Revisions).
1. The introduction of the paper should be expanded and improved. In order to explain more clearly the problem addressed in this work, the introduction should be modified as follows. Section 1 (Introduction) should be subdivided into five brief Subsections: 1.1) Background, 1.2) Formulation of the Problem of Interest for this Investigation, 1.3) Literature Survey, 1.4) Scope and Contribution of this Study, 1.5) Organization of the Paper. New material should be concisely added in all these subsections. In Subsection 1.1, general information about the problem addressed in this paper should be presented. In Subsection 1.2, the authors should illustrate the challenges of the research that they are discussing in the paper. In Subsection 1.3, a detailed literature survey on the problem addressed in the paper should be provided for the benefit of the reader that may be not familiar with the problem investigated in the paper. In Subsection 1.4, the differences between the method proposed in the paper with respect to the results already published in the literature should be briefly explained. In Subsection 1.5, the organization of the paper should be concisely reported.
- We divided the section into subsections.
- We added to sections 1.3 (related work) an example of tests being made in swimming pools to test underwater concepts and emphasized the advantages of our approach.
2. The paper lack a proper conclusion section. This section should be added to the revised paper.
For example, the title “Summary and Conclusions” can be used. In this new section, the authors
should report a summary of the contributions of the paper, the conclusion drawn in this
investigations, and some suggestions for future directions of research.
We modified the last section to provide a summary and concluding remarks on our work, and also to outline future directions. The discussion will focus about the usefulness of the environment and motivation behind it, and will summarize the results.
3. Please improve the quality of Figures 2 and 17.
We have replaced the figures with higher resolution images.
4. Figures 3, 11, 12, and 13 should be modified including the measurement units on the x and y
axes.
Thanks for pointing out the omission in the graphs. We have added the measurement units along the axes.
5. The bibliography of the paper should be expanded. For example, below please find some
potential additional references:
- Chamayou, G. (2015). A Theory of the Drone. New Press, The.
- Fahlstrom, P., & Gleason, T. (2012). Introduction to UAV systems. John Wiley & Sons.
- Pappalardo, C. M., and Guida, D., 2018, “Use of the Adjoint Method in the Optimal Control
Problem for the Mechanical Vibrations of Nonlinear Systems”, Machines, 6(2), 19.
- Zhu, H. Y., Niu, Y. F., Shen, L. C., & Zhang, G. Z. (2010). State of the art and trends of
autonomous control of UAV systems. Journal of National University of Defense Technology, 32(3),
115-120.
- McLain, T. W., Chandler, P. R., Rasmussen, S., & Pachter, M. (2001). Cooperative control of
UAV rendezvous. In American Control Conference, 2001. Proceedings of the 2001 (Vol. 3, pp.
2309-2314). IEEE.
We found the book "A Theory of the Drone" touches more ethical aspects of drones with special attention for remote warfare but the "Introduction to UAV systems" seems to serve as a good reference to get the necessary background to understand the general principles of Remote Vehicles Operations so we add it as part of the references in the introduction.
Two more references added to emphasize the current problem domain:
Myint, M.; Yonemori, K.; Lwin, K.N.; Yanou, A.; Minami, M. Dual-eyes vision-based docking system for autonomous underwater vehicle: an approach and experiments. Journal of Intelligent & Robotic Systems 2018, 92, 159–186.
Myint, M.; Yonemori, K.; Yanou, A.; Ishiyama, S.; Minami, M. Robustness of visual-servo against air bubble disturbance of underwater vehicle system using three-dimensional marker and dual-eye cameras. OCEANS’15 MTS/IEEE Washington. IEEE, 2015, pp. 1–8.
In my opinion, addressing all the points mentioned above will help in improving the quality of the
paper.
