An Occupancy Information Grid Model for Path Planning of Intelligent Robots

Round 1

Reviewer 1 Report

In the manuscript, the authors present their research activity in mapping/path planning with a regular hexagonal space representation. The idea they followed is not new, as also stated in the "related work" section and ref to [22] - https://ieeexplore.ieee.org/abstract/document/4367728.

Therefore the most interesting part consists in the experimental section. However, there are some parts that are missing such as the computational performances of algorithms or technical details about the robotic platform used. For example, the experimental area seems huge, but no information about the range of the lidar uses is provided. 

From a "robotic" point of view and given the lack of novelty, the presented work is more an exercise than real research. Moreover, the experimental section is too short and does not provide enough information to be published. The experimental section includes just 5 experiments of path planning with only the standard A* algorithm and does not include any computational performance information. Some information about how to store the hex representation should be included: if a grid structure is clearly stored as an array and good optimization exists for these structures, does this hex representation affect the space or time performances? These are the details I would like to see in this type of manuscript. 

However, the topic is interesting and the work is well presented. The English level is good (proofreading by AJE service) and I did not notice any problem reading the manuscript. 

There are a lot of issues with this paper that I'd like to share.

1. The paper is really too long. The way that the authors wrote this manuscript is more appropriate to a book chapter or thesis. 
2. Apart from the introduction and related work sections that I would definitely reduce in size,  there is a big block that in my opinion is too similar to standard literature in the robotics field. For example on page 12 and section 4.2 "probability calculation", all these sections are too similar to the Probabilistic Robotics book by Thrun/Burgard/Fox, page 284 and following. For example, Equation 11 is the same as Equation 9.4., table 2 is the same as table 9.1, and no citation to that book is provided (instead there is an auto-ref to [37]). I will not go into this, but the problem is that all these sections do not provide any added value to the manuscript itself. You have a 24-page manuscript that can be drastically reduced by simply avoiding these description parts that are in my opinion unnecessary nowadays, at least for the robotics community. 
3. Some references in the related work sections seem to not have any "real" connection. I was not able for example to retrieve any information about [17] that was referred to with a sentence between quotation marks on page 4, "based on the concept of circular partitioning..." apart from these
   https://link.springer.com/book/10.1007/978-3-642-33024-7
   https://link.springer.com/content/pdf/10.1007%2F978-3-642-33024-7.pdf
   http://en.cnki.com.cn/Article_en/CJFDTotal-DLGT201201002.htm
   https://ieeexplore.ieee.org/abstract/document/8128393/references#references
4. With reference to the previous point, I believe that the authors mixed the research fields from robotics and "Geography and Geo-Information Science" fields, which have obviously different objectives.
5. On page 5, "The ratio of the nearest center distance to the farthest is called the grid center distance error". Are you sure it's not the other way around?
6. Tab.1, the 1.150 value is wrong. If you use all the decimals, it should be 1.154(7), but again, it reinforces the previous point: it is max/min, not mix/max. 
7. Minor: figure 3, where are (a) and (b)
8. Section 3.2: how is this part related to path planning? 
9. Minor: from time to time, the text size changes. See "The most important manifestation of the grid ATTRIBUTES is ...."
10. Section 3.2.1 is not related to the proposed algorithm
11. Section 4: "and the probability estimate of the grid where the laser point is located is calculated by applying THE HIDDEN MARKOV MODEL", which is not yet introduced. 
12. The Bresenham algorithm is usually defined for grids. It would be nice to have some info about how the authors adapted the algorithm to the hexagonal space decomposition.
13. In section 4.1, it is not clear what the "grid direct encoding method" is.
14. How the aforementioned method can "nearly save half of the computer storage space"? 
15. Figure 12: the authors overlap a cartesian/grid system to the hex-decomposition. The change in resolution is related to the measurement system of the world (I mean, centimeters vs the actual space decomposition?) 
16. Section 4.2 is not needed in my opinion, is just basic knowledge of the robotics field. Please refer to Probabilistic Robotics by Thrun/Burgard/Fox or artificial intelligence a modern approach textbook by Russell/Norvig. 
17. Some images are low quality/low res.

 

 

Author Response

Dear reviewer,

Thank you very much for your time involved in reviewing the manuscript and your very encouraging comments on the merits. We have carefully checked and improved the English writing in the revised manuscript. I use blue font to mark the major revisions in the manuscript.

The paper mainly revised the following aspects:

1. Described the main idea of this manuscript through Figure 1.
2. Modified Section 5, adding the principle analysis that the influence of terrain features, grid element attribute, and grid edge attributes on the path planning algorithm.
3. Added experimental content, namely Section 6.4, describing the influence of terrain characteristic, grid element attribute, and grid edge attributes on the path planning algorithm, this is experimental content.
4. Added Bresenham algorithm description for regular hexagonal grid.
5. Modified the content of "Related Work", and used 17 references from the past five years.
6. Major revisions have been made to Section 4.2, only the basic content is retained, and a reference to "probabilistic robots" is added.
7. According to the content, the title of this manuscript was revised to " An Occupancy Information Grid Model for Path Planning of Intelligent Robots ", i.e. the new model is an occupancy information grid model. In this way, the intent of the fusion of robots and geographic information grids is taken into account.
8. Modified Section 4.1 to remove misleading content.
9. Re-written the Abstract of this manuscript.
10. Modified all graphics in this manuscript.
11. Invited AJE to polish this manuscript again.

We also appreciate your clear and detailed feedback and hope that the explanation has fully addressed all of your concerns. In the remainder of this letter, we discuss each of your comments individually along with our corresponding responses.

To facilitate this discussion, we first retype your comments in italic font and then present our responses to the comments. 

Please see the attachment.

Thank you and best regards!

 

Yours sincerely, Jinming Zhang

Author Response File: Author Response.pdf

Reviewer 2 Report

I find the articles to be a comprehensive, very sound, and interesting investigation, which I can't find any fault other than a very minor English language check required. Good job! I recommend the paper acceptance.

Author Response

Dear reviewer,

Thank you very much for your time involved in reviewing the manuscript and your very encouraging comments on the merits. We have carefully checked and improved the English writing in the revised manuscript. I use blue font to mark the major revisions in the manuscript.

 

The paper mainly revised the following aspects:

1. Described the main idea of this manuscript through Figure 1.
2. Modified Section 5, adding the principle analysis that the influence of terrain features, grid element attribute, and grid edge attributes on the path planning algorithm.
3. Added experimental content, namely Section 6.4, describing the influence of terrain characteristic, grid element attribute, and grid edge attributes on the path planning algorithm, this is experimental content.
4. Added Bresenham algorithm description for regular hexagonal grid.
5. Modified the content of "Related Work", and used 17 references from the past five years.
6. Major revisions have been made to Section 4.2, only the basic content is retained, and a reference to "probabilistic robots" is added.
7. According to the content, the title of this manuscript was revised to " An Occupancy Information Grid Model for Path Planning of Intelligent Robots ", i.e. the new model is an occupancy information grid model. In this way, the intent of the fusion of robots and geographic information grids is taken into account.
8. Modified Section 4.1 to remove misleading content.
9. Re-written the Abstract of this manuscript.
10. Modified all graphics in this manuscript.
11. Invited AJE to polish this manuscript again.

 

We also appreciate your clear and detailed feedback and hope that the explanation has fully addressed all of your concerns. In the remainder of this letter, we discuss each of your comments individually along with our corresponding responses.

 

To facilitate this discussion, we first retype your comments in italic font and then present our responses to the comments.

Please see the attachment.

Thank you and best regards!

 

Yours sincerely, Jinming Zhang

Author Response File: Author Response.pdf

Reviewer 3 Report

The paper deals with the path planning problem for intelligent robots. 
The authors present an interesting approach using a hexagonal grid model 
and an extended version of the A* algorithm to work in this topology. 
The paper presents a thorough literature review, along with the details 
of the proposed algorithm. The authors discuss the advantages of their 
algorithm and present the conclusions.

The overall structure and presentation of the paper are good. There are, 
however, several issues that need to be addressed. More specifically:

1) The authors propose an extension of the A* algorithm for path planning 
in environments using a hexagon-based grid, instead of the traditional 
square-based grid. In section 3, the authors present the benefits of 
this grid model, when compared to the triangle and square models. 
However, their explanation in sections 3.1 and 3.2 is not enough. Please provide better reasoning as to why hexagon-based grids are superior to path planning.

2) There are two main (path planning) algorithm categories, offline and 
online. Offline algorithms know from the start all the information about 
the topology and the position of the robot(s). In contrast, online 
algorithms only require information to start running, because they gain 
information on each step. Reading the manuscript does not make it clear 
early enough what problem exactly the proposed algorithm solves and in 
which category it belongs.

3) The paper is lacking state of the art literature review. The "Related 
Work" section is full of old outdated papers, as old as 1998! I would 
strongly advise you to improve the Bibliography section using modern 
state-of-the-art research works.

4) Even though the experimental results are very detailed, the authors 
do not provide any details about the experimental setup. The authors 
state that they used a robot with a Kinect camera and an RPLidar A2 
module, but they do not give any more details. I would like to see at 
least one (1) page giving a thorough explanation of the hardware of the 
robot and the reasoning for the selection of said hardware. Furthermore, they do not compare extensively their path planning algorithm with other authors. Also, it is important to state what algorithms are running, if the vehicle has complete knowledge of the grid, how the grid and sensor inputs are fused together, and so on. Many important bits are missing.

5) In a few parts of the paper the authors write something that is not connected to the paper, or they do not say why it is important to write. It seems like multiple authors have contributed and nobody checked the consistency and the coherency. For example, in the Related work sections, the authors write about grid computing which is 100% irrelevant to this paper. 

6) Even though the paper consists of many pages, there is not enough information that can help another research team to reimplement this technique. There are bits of information here and there but they are not complete. Nobody can repeat these experiments by reading this paper. Proper publishing means disclosure of all information so the community can benefit from it. As it is, it is not complete.

7) The authors should provide a minimal working example, starting from the map and working the mathematics towards the provided path solution to illustrate how everything fits together. Start for example with a map of 20  hexagonal of various types, annotate start and end hexagons and demonstrate how the equations are utilized.

 

Author Response

Dear reviewer,

Thank you very much for your time involved in reviewing the manuscript and your very encouraging comments on the merits. We have carefully checked and improved the English writing in the revised manuscript. I use blue font to mark the major revisions in the manuscript.

 

The paper mainly revised the following aspects:

1. Described the main idea of this manuscript through Figure 1.
2. Modified Section 5, adding the principle analysis that the influence of terrain features, grid element attribute, and grid edge attributes on the path planning algorithm.
3. Added experimental content, namely Section 6.4, describing the influence of terrain characteristic, grid element attribute, and grid edge attributes on the path planning algorithm, this is experimental content.
4. Added Bresenham algorithm description for regular hexagonal grid.
5. Modified the content of "Related Work", and used 17 references from the past five years.
6. Major revisions have been made to Section 4.2, only the basic content is retained, and a reference to "probabilistic robots" is added.
7. According to the content, the title of this manuscript was revised to " An Occupancy Information Grid Model for Path Planning of Intelligent Robots ", i.e. the new model is an occupancy information grid model. In this way, the intent of the fusion of robots and geographic information grids is taken into account.
8. Modified Section 4.1 to remove misleading content.
9. Re-written the Abstract of this manuscript.
10. Modified all graphics in this manuscript.
11. Invited AJE to polish this manuscript again.

 

We also appreciate your clear and detailed feedback and hope that the explanation has fully addressed all of your concerns. In the remainder of this letter, we discuss each of your comments individually along with our corresponding responses.

 

To facilitate this discussion, we first retype your comments in italic font and then present our responses to the comments.

Please see the attachment.

Thank you and best regards!

 

Yours sincerely, Jinming Zhang

Author Response File: Author Response.pdf

Round 2

Reviewer 1 Report

Dear authors,

I want to start this review by thanking you for your great efforts in revising the manuscript. I carefully read your rebuttal letter, and I believe that most of my concerns were adequately addressed. 

I still believe that the most interesting part consists of the experimental section and the paper is still too long, but I do appreciate the amount of work you've done. 

I noticed you added one reference to the code in GitHub, but as of today this page contains no code. https://github.com/VGEZhangJM/RoboticGrid Please fix this issue.

Author Response

Dear reviewer,

Thank you very much for your time involved in reviewing the manuscript and your very encouraging comments on the merits. We have carefully checked and improved the English writing in the revised manuscript..

 

The paper mainly revised the following aspects:

1. Invited AJE to polish this manuscript again, and attached the editing certificate. This is the latest edited draft returned on the 25th.
2. In view of the long paper, some figures and tables that have little impact on the paper have been deleted, for example, the original Table 1 and the original Figure 9 have been deleted. In this way, the length of the paper is limited to 22 pages.
3. Adjusted the size of the figure to fit the layout effect.

 

We also appreciate your clear and detailed feedback and hope that the explanation has fully addressed all of your concerns. In the remainder of this letter, we discuss each of your comments individually along with our corresponding responses.

 

To facilitate this discussion, we first retype your comments in italic font and then present our responses to the comments.

 

Thank you and best regards!

Yours sincerely, Jinming Zhang

Corresponding Dr. Jinming Zhang, E-mail: [email protected]

Author Response File: Author Response.pdf