Research Progress on Synergistic Technologies of Agricultural Multi-Robots

Round 1

Reviewer 1 Report

1. As for the first contribution, authors are trying to comment on synergistic technologies of agricultural multi-robots. In practical point of view, the paper provides with valuable data for researchers in multi-robot field.
2. However, the originality or the paper is very slim.
3. To enhance the quality of the paper, more works on new technology or algorithm based on the technologies mention in the paper.
4. In colocation methods, more comments on sensors and sensor fusion are recommended.

Author Response

Dear reviewer:

I am very grateful to your comments for the manuscript on  14 Jan 2021. According with your advice, we amended the relevant part in manuscript. Some your questions were answered below.

1)

The reviewer’s comment:

To enhance the quality of the paper, more works on new technology or algorithm based on the technologies mention in the paper.

The author’s answer:

From 3.2 Environment Awareness to 3.5 Formation Control of the manuscript, I have added information about how the new algorithms for agricultural multi-robots are applied in the corresponding technologies and reviewed the strengths and weaknesses of the work.

2)

The reviewer’s comment:

In colocation methods, more comments on sensors and sensor fusion are recommended.

The author’s answer:

I have added a description of the application of this method in 3.2.1 Co-localization, and since no research literature on interactive localization in cooperative localization of agricultural multi-robots has been found, I review what is missing in this area. Since co-localization is inseparable from the use of sensors and information fusion, for a closer illustration, I point out the main sensor types and application methods currently used for localization in agricultural multi-robots in 3.2.2 Information fusion and review them both in 3.2.2 Summary and 4 Discussion. Also in the discussion, I further describe the challenges of agricultural multi-robots in this area and point out future research directions in this area.

Reviewer 2 Report

This paper is a review of high quality. The concept is clear, the approach scientifically sound and the topic interesting. Two minor suggestions to be taken into account by the authors: a) in the introduction section to add some examples not only from China but also from other countries and b) add some words about the potential link between synergistic technologies-multi robots with decision support systems (DSS). Moreover, some minor corrections-revisions could be performed e.g. line 85 "...cooperative operations operated by human beings..." 

Author Response

Dear reviewer:

I am very grateful for your comments on the manuscript on 14 Jan 2021. According to your advice, we amended the relevant part of the manuscript. Some of your questions were answered below.

1)

The reviewer’s comment:

 in the introduction section to add some examples not only from China but also from other countries 

The author’s answer:

I added data on the relationship between agricultural labor costs and agricultural productivity for three economically developed countries, Germany, Japan, and the United States, for use in the opposite direction to show that to increase a country's competitiveness internationally, new types of technology are needed to replace human work.

2)

The reviewer’s comment:

add some words about the potential link between synergistic technologies-multi robots with decision support systems (DSS).

The author’s answer:

Since task assignment and coordination and collaboration mechanism cannot be separated, I added the relationship between DDS system and task assignment in the case of centralized assignment in 3.3 Task Assignment for illustrating the importance of DDS.

3)

The reviewer’s comment:

Moreover, some minor corrections-revisions could be performed e.g. line 85 "...cooperative operations operated by human beings..." 

The author’s answer:

I rechecked the grammar of the draft and also double-checked it with a grammar tool to make sure it was rigorous.

Reviewer 3 Report

Summary

The paper starts with an Introduction justifying the urgent need for automation of agriculture through broad exploitation of multi-robot systems. This problem is not novel – it is well known in the worldwide context and thus the paper motivation and its significance does not need such comprehensive elaboration. I recommend reducing this part of the paper at least twice by omitting China-specific social aspects motivating the paper orientation because these social aspects are practically the same for any country in the world.  

The next section considers the specific factors and characteristics peculiar to the use of multi-robot systems in the agricultural environment from such points of view as crops growth patterns developing in time and associated treatment needed, agricultural robots used and peculiarities of their exploitation, peculiarities of farmland and environment, crop treatment techniques and technology components, etc. It seems that this section could be written in a more structured way, e.g. through an explicit list of these factors and characteristics and specific problems they cause. In the present form, the reader does not get the overall structured picture of environment-caused problems.

Section 3 presents the main content of the paper – an overview of technological progress and development characteristics, advantages and drawbacks of agricultural multi-robot systems according to the following agenda: (1) environment perception, (2) plant treatment task allocation, (3) multi-robot path planning, (4) structures of multiple robot “organizations” (multi-robot formations) and associated control, and (5) robot communication environment and tasks. Section 4 discusses and summarizes problems and challenges according to the aforementioned five issues of agricultural multi-robot systems. The conclusion is of a formal text; it provides no new information and actually can be omitted or enriched with information on missed parts of agricultural multi-robot scientific and technical agenda omitted in the paper and motivation.

 

Broad comment to the paper

1. The explicitly claimed objective of the paper is the summarization of “the research progress on synergetic technologies of cooperative operations of the operation of the agricultural multi-robots and characterizes the expected research of related technologies.” According to my opinion, this research progress should include not only what is either modeled, or simulated, or implemented in purely agriculture-oriented collective robotic. This research progress also should include general domain-independent theory and practice of teamwork and collective behavior coordination that applies to agricultural multi-robot scope. However, these very rich and fundamental results, specifically for autonomous mobile multi-robot systems, are fully missed in the paper. I know that these results need some special overview, but it is impossible to talk about perspectives of agricultural collective robotics with full ignorance of these results.
2. The mapping problem (section 3.3.3) looks a bit outdated. First of all, SLAM technology is currently intensively being developed. However, the paper does not pay the necessary attention to this approach to map robot position in environment map. Moreover, a technology developed by Japanese company Topcon sells a positioning system for unmanned agricultural vehicles that provides the positioning precision of 2 cm, which is enough to support any positioning task of agricultural multi-robot systems. However, the paper does not even mention this system. The Topcon system is based on the integration of GPS and local positioning sources of information. Such systems look very perspective and competitive for the positioning tasks in question.
3. Multi-robot task allocation is very popular in the scientific research community and there exist a lot of approaches and algorithms to solve. The most computationally efficient and scalable algorithms applicable to the planning for agricultural multi-robot systems are developed using artificial intelligence and self-organization models, e.g. based on meta-heuristics and local policies discovered through Machine Learning techniques and market- based policies like Contract Net and other auction-like protocols. Reviving integer programming is too surprising.
4. It seems that planning task statements reviewed in the paper correspond to practically independent operation of robots assigned particular zones of operation and the essence of coordination, in these task statements, is reduced to obstacle and collision avoidance. However, the practical plant treatment tasks are not independent. The simplest and most often case is that they are partially ordered. In such cases, any of the reviewed approaches are not applicable at all. At the same time, a synergy of multi-robot systems emerges through collaborative situations, when the robots need to coordinate their operations due to the dependences of the task to be solved through collective efforts. In such cases, robots need not only to coordinate their behavior but also to synchronize the individual actions along a timeline. I think that the case of a partially ordered task should be analyzed.
5. The main requirement for the robots’ information exchange to coordinate and synchronize their individual behaviors in multi-robot teamwork is to support so-called situational awareness. However, the paper does not analyze how the situational awareness problem is solved in the case of agricultural multi-robot systems. Indirectly, the paper mentions situational awareness in lines 573 – 575 where it is written: «Each robot can know the intent, goal, motion, and current status of the environment of other robots through communication to conduct effective consultation and coordinate tasks [28].» This is an incorrect statement, because what a particular robot has to know about other robots and about itself significantly depends on many factors. It depends on multi-robot system architecture, on coordination and synchronization strategies used, among others system properties (see, e.g.

(a) Situational Awareness. by Aaron S. Dietz, Eduardo Salas ISBN 9780754629733, Published September 23, 2011, by Routledge 544 Pages. or

(b) Designing for Situation Awareness. An Approach to User-Centered Design, Second Edition By  Mica R. Endsley. Second Edition. Published January 30, 2012, by CRC Press, 396 Pages).

6. The comments have given above are mainly about what is missed in the paper. They are my recommendations on how to improve the paper. However, I know that it would be more correct to comment on what is written in the paper. In this respect, I am sure that the main paper results are of value, information given in the tables and discussion presented in Section 4 corresponds to the paper's contribution supporting its originality, novelty, and impact.

 

Specific comments 

Some mistakes needed to correct:

Line 550: the reference to Table 8 has to be substituted by reference to Table 6.

line 619: the reference to Table 1 has to be substituted by reference to Table 7.

Author Response

Dear reviewer:

I am very grateful for your comments on the manuscript on 20 Jan 2021. According to your advice, we amended the relevant part of the manuscript. Some of your questions were answered below.

1) 

The reviewer’s comment:

I recommend reducing this part of the paper at least twice by omitting China-specific social aspects motivating the paper orientation because these social aspects are practically the same for any country in the world.  

The author’s answer:

I added data on the relationship between agricultural labor costs and agricultural productivity for three economically developed countries, Germany, Japan, and the United States, for use in the opposite direction to show that to increase a country's competitiveness internationally, new types of technology are needed to replace human work.

2)

The reviewer’s comment:

The next section considers the specific factors and characteristics peculiar to the use of multi-robot systems in the agricultural environment from such points of view as crops growth patterns developing in time and associated treatment needed, agricultural robots used and peculiarities of their exploitation, peculiarities of farmland and environment, crop treatment techniques and technology components, etc. It seems that this section could be written in a more structured way, e.g. through an explicit list of these factors and characteristics and specific problems they cause. In the present form, the reader does not get the overall structured picture of environment-caused problems.

The author’s answer:

I modified the end of the second part to illustrate four specific differences in the use of multiple robots in agricultural environments through the case of harvesters harvesting in large fields, and to propose corresponding collaborative technologies for multiple robots in agriculture to address these differences.

3)

The reviewer’s comment:

 Section 4 discusses and summarizes problems and challenges according to the aforementioned five issues of agricultural multi-robot systems. The conclusion is of a formal text; it provides no new information and actually can be omitted or enriched with information on missed parts of agricultural multi-robot scientific and technical agenda omitted in the paper and motivation.

The author’s answer:

I have rewritten the discussion in Part 4, which introduces the current state of research in each technology in Part 3, respectively, and summarizes the shortcomings in development, and presents the challenges in the development of multiple robots in agriculture. And in the conclusion of Part 5, the corresponding future research directions are proposed to address the challenges in Part 4.

4)

The reviewer’s comment:

The explicitly claimed objective of the paper is the summarization of “the research progress on synergetic technologies of cooperative operations of the operation of the agricultural multi-robots and characterizes the expected research of related technologies.” According to my opinion, this research progress should include not only what is either modeled, or simulated, or implemented in purely agriculture-oriented collective robotic. This research progress also should include general domain-independent theory and practice of teamwork and collective behavior coordination that applies to agricultural multi-robot scope. However, these very rich and fundamental results, specifically for autonomous mobile multi-robot systems, are fully missed in the paper. I know that these results need some special overview, but it is impossible to talk about perspectives of agricultural collective robotics with full ignorance of these results.

The author’s answer:

The suggestions of the reviewers are greatly appreciated. Considering that the focus of this paper is to present the research progress in collaborative multi-robot operation technologies in agriculture, some review articles have already provided detailed descriptions of multi-robot coordination and practices, and in order to focus more on this aspect of technology and to allow the reader to better understand the technologies presented in the article, the article provides a conceptual introduction to the technologies used.

5)

The reviewer’s comment:

The mapping problem (section 3.3.3) looks a bit outdated. First of all, SLAM technology is currently intensively being developed. However, the paper does not pay the necessary attention to this approach to map robot position in environment map. Moreover, a technology developed by Japanese company Topcon sells a positioning system for unmanned agricultural vehicles that provides the positioning precision of 2 cm, which is enough to support any positioning task of agricultural multi-robot systems. However, the paper does not even mention this system. The Topcon system is based on the integration of GPS and local positioning sources of information. Such systems look very perspective and competitive for the positioning tasks in question.

The author’s answer:

I added newer research in 3.2.3 and mentioned the development of SLAM for multi-robot applications in the conclusion. And GPS localization is also mentioned in 3.2.2 Sensor fusion, and the Japanese Topcon set of systems is also added.

6)

The reviewer’s comment:

Multi-robot task allocation is very popular in the scientific research community and there exist a lot of approaches and algorithms to solve. The most computationally efficient and scalable algorithms applicable to the planning for agricultural multi-robot systems are developed using artificial intelligence and self-organization models, e.g. based on meta-heuristics and local policies discovered through Machine Learning techniques and market- based policies like Contract Net and other auction-like protocols. Reviving integer programming is too surprising.

The author’s answer:

I added cases on integer planning methods to the assignment and mentioned development challenges and future directions in this regard in the discussion and conclusion.

7)

The reviewer’s comment:

It seems that planning task statements reviewed in the paper correspond to practically independent operation of robots assigned particular zones of operation and the essence of coordination, in these task statements, is reduced to obstacle and collision avoidance. However, the practical plant treatment tasks are not independent. The simplest and most often case is that they are partially ordered. In such cases, any of the reviewed approaches are not applicable at all. At the same time, a synergy of multi-robot systems emerges through collaborative situations, when the robots need to coordinate their operations due to the dependences of the task to be solved through collective efforts. In such cases, robots need not only to coordinate their behavior but also to synchronize the individual actions along a timeline. I think that the case of a partially ordered task should be analyzed.

The author’s answer:

I have additional content about the task assignment time series in Task Assignment. As for zoning assignment and obstacle avoidance, which are the most basic in task assignment and have been implemented in the case, I have also added comments in this area in the task assignment.

8)

The reviewer’s comment:

    The main requirement for the robots’ information exchange to coordinate and synchronize their individual behaviors in multi-robot teamwork is to support so-called situational awareness. However, the paper does not analyze how the situational awareness problem is solved in the case of agricultural multi-robot systems. Indirectly, the paper mentions situational awareness in lines 573 – 575 where it is written: «Each robot can know the intent, goal, motion, and current status of the environment of other robots through communication to conduct effective consultation and coordinate tasks [28].» This is an incorrect statement, because what a particular robot has to know about other robots and about itself significantly depends on many factors. It depends on multi-robot system architecture, on coordination and synchronization strategies used, among others system properties (see, e.g.

(a) Situational Awareness. by Aaron S. Dietz, Eduardo Salas ISBN 9780754629733, Published September 23, 2011, by Routledge 544 Pages. or

(b) Designing for Situation Awareness. An Approach to User-Centered Design, Second Edition By  Mica R. Endsley. Second Edition. Published January 30, 2012, by CRC Press, 396 Pages).

The author’s answer:

The reason for quoting this passage in communication is to show that the robot can obtain information about other robots through communication. Of course, there is currently some research on multiple robots in the absence of communication. And implicit communication allows robots to obtain environmental information through sensors to accomplish collaboration. Here I have also given a case study for illustration.

To illustrate the importance of communication technology for multiple robots and to avoid ambiguity, I modified it to"In agricultural production, many factors affect the fine operation of agricultural robots, and to maintain coordination and cooperation among multiple robots and to gain a more comprehensive understanding of the environment in which multiple robots perform tasks, robots need to interact with each other through information to better perform a given task "

9)

Some mistakes needed to correct:

Line 550: the reference to Table 8 has to be substituted by reference to Table 6.

line 619: the reference to Table 1 has to be substituted by reference to Table 7.

The reviewer’s comment:

I have fixed these minor errors in the article.

Reviewer 4 Report

The paper describes research progress of the multi-robot systems in agriculture. The problem is clearly motivated and definitely worth investigation. However, in the reviewer's opinion, the paper is very far from being ready for publication. In some places it is even not understandable, to the extent that makes it really hard to provide a comprehensive review, i.e. judge the contributions and scientific value of this work. Below the exact problems are described.

1. The contribution/aim of the paper is not clearly stated. From the last sentence of introduction it seems that there are two contributions:
- summary of the research progress on synergetic technologies
- predicted problems to address in research for the multi-robot systems to exist in practice.
Whereas, the summary of state-of-the-art is needed, it does not bring much of
the scientific value. The second part --- predicted research topics --- seems to be much more valuable, but at the same time it is a bit neglected in the text. I
assume that section 4 (Discussion) was meant to answer this question, but there are only some statements, without introduction and clear logical connection with the rest of the paper.

2. There are many language problems in the text, which in some places might
hinder understanding. The whole text needs to be carefully proof-read and
corrected. Just to name a few examples:
- line 109: "uneven and uneven",
- line 119: a dot is missing at the end of the sentence, part "identify multi-robot in the ground cross" is not understandable,
- line 121: "multiple robots cross on the ground", should it be "paths of multiple robots cross"?
- line 207: I believe authors meant "extended Kalman filter",
- line 402: "to avoid obstacle avoidance".

3. In some places the paper lacks logical flow or is inconsistent. It looks like it was not revised after being written. To give some of many examples:
- In section 3.3.3 Mapping (part of the section 3.2 --- mind the wrong numbering!). The paragraphs seem to be completely independent and are unclear. Moreover, the statement in lines 301-305 seems to be misleading: do the authors mean a specific example or is having at least two robots to need a whole map is a general rule?
- Figure 17: Explicit communication appears in 3 different places, perhaps those three categories should be called in various ways. Explicit communication is divided  in "Explicit communication" and "Implicit communication", which does not have much sense. For the "upper" "Implicit communication" transmission modes are mentioned ("unicast", "multicast", "broadcast"), whereas for "Explicit: upper level collaboration" an example is "wireless network communication". Are those modes not realized with a wireless network? If they are, why are they in different categories?

4. References should be checked. While reading, I spotted some errors, but there might be more:
- In [8] and [9] DOI starts with "doi:doi:"
- In [23] conference name is missing.

5. The paper contains some figures from various sources, the authors should make  sure that the copyrights are not violated.

Author Response

Dear reviewer:

I am very grateful for your comments on the manuscript on 12 Jan 2021. According to your advice, we amended the relevant part of the manuscript. Some of your questions were answered below.

1)

The reviewer’s comment:

 I assume that section 4 (Discussion) was meant to answer this question, but there are only some statements, without introduction and clear logical connection with the rest of the paper.

The author’s answer:

I have rewritten the discussion in Part 4, which introduces the current state of research in each technology in Part 3, respectively, and summarizes the shortcomings in development, and presents the challenges in the development of multiple robots in agriculture. And in the conclusion of Part 5, the corresponding future research directions are proposed to address the challenges in Part 4.

2)

The reviewer’s comment:

There are many language problems in the text, which in some places might
hinder understanding.

The whole text needs to be carefully proof-read and
corrected. Just to name a few examples:
- line 109: "uneven and uneven",
- line 119: a dot is missing at the end of the sentence, part "identify multi-robot in the ground cross" is not understandable,
- line 121: "multiple robots cross on the ground", should it be "paths of multiple robots cross"?
- line 207: I believe authors meant "extended Kalman filter",
- line 402: "to avoid obstacle avoidance".

The author’s answer:

I rechecked the grammar of the draft and also double-checked it with a grammar tool to make sure it was rigorous.

3)

The reviewer’s comment:

In some places the paper lacks logical flow or is inconsistent. 

 To give some of many examples:
- In section 3.3.3 Mapping (part of the section 3.2 --- mind the wrong numbering!). The paragraphs seem to be completely independent and are unclear. Moreover, the statement in lines 301-305 seems to be misleading: do the authors mean a specific example or is having at least two robots to need a whole map is a general rule?
- Figure 17: Explicit communication appears in 3 different places, perhaps those three categories should be called in various ways. Explicit communication is divided  in "Explicit communication" and "Implicit communication", which does not have much sense. For the "upper" "Implicit communication" transmission modes are mentioned ("unicast", "multicast", "broadcast"), whereas for "Explicit: upper level collaboration" an example is "wireless network communication". Are those modes not realized with a wireless network? If they are, why are they in different categories?

The author’s answer:

The manuscript was reworked by adding examples of new works to avoid illogical or vague descriptions. For example, in 3.2.3, I added new technical descriptions to get a logical summary by commenting on these applications. Figure 17 was replaced and described with examples of actual applications.

4)  

The reviewer’s comment:

References should be checked. While reading, I spotted some errors, but there might be more:
- In [8] and [9] DOI starts with "doi:doi:"
- In [23] conference name is missing.

The author’s answer:

I rechecked the formatting of the literature to make sure it was consistent with the journal requirements.

5)

The reviewer’s comment:

The paper contains some figures from various sources, the authors should make  sure that the copyrights are not violated.

The author’s answer:

To avoid not raising copyright issues, I have cited the literature below the figure number of the original figure cited. For those that modified the original figure, I replaced the modified figure with a sketch I drew in order not to affect the description. (As shown in Figure 2(b))

Reviewer 5 Report

This paper reviews the latest research trends about the application of cooperative multi-robot technologies in the agricultural production domain. In particular, the paper focuses on five fundamental technologies of multi-robot systems: environment perception, task allocation, path planning, formation control, and communication. The paper concludes that despite the complex challenge of combining these technologies in a synergistic fashion, past research is promising and it is possible to expect such technologies being deployed in real agricultural multi-robot systems in the future.

The paper is easy to follow and understand, however, there are a lot of minor grammatical errors, some of which I report in the end of my review. A more comprehensive read of the paper has to be done to check and correct these errors.

I would change the wording about the three types of agricultural multi-robots to clarify that these are the three types that the review focuses on. There are however other types, such as for example USVs (Unmanned Surface Vehicles).

Since this is a review paper, it needs a section explaining what kind of methodology was used. I assume this was a systematic review of the literature, but this should be made explicit. It would also help to know which indexing research websites were used, what were the terms, how many papers were found, how many papers were discarded for being out of scope, etc.

While the format of describing a multi-robot concept (such as environmental perception), citing agricultural research that has used it (in general), and then specifying what most of the approaches do is interesting at first glance, it looses the information of which particular approach used which technique. For example, the following paragraph:
"In the processing of environmental perception information, agricultural multi-robots mainly adopt the feature layer information fusion method."
Tells us an important information, however, it is missing citations that indicate which ones use this method, something that would be valuable for readers of this review to know.

Similarly, it would be nice if there was more detail about at least some of these approaches (e.g. the more influential). Explaining how they used these multi-robot concepts, the scenario of application, and what were the results obtained. For example, Table 5 has various applications of centralised and distributed task allocation, however these applications are not further explained in the text.

Paragraphs from lines 438 to 459 in page 15 have a contradiction. First it is said that distributed path planning can only find suboptimal solutions (438-439), and then when describing the method given as example in agricultural multi-robots it is said that "The practice shows that the method can find the optimal solution" (456).

I was expecting to find a bit more in the Discussion section. It seems quite short for the amount of research that was reported. Future directions are also missing and would be a good addition to that section.

 

Minor comments and suggestions:
All mentions of "et.al" should be replaced by "et al.".
There are several instances where there is no space between a word and the citation, for example "multiple robots possible[21]". I point some below but a full check has to be made whenever there is a citation.
Some figures have very small text (e.g. Figure 9).
The tables look a bit strange to me, specially the ones that break pages. This is more a matter of style, but nonetheless tables over page breaks should definitely be avoided (perhaps
Page 1, line 40: the physical of the elderly -> the physical health of the elderly
Page 2, line 61: of operators[10] -> of operators [10]
Page 2, line 70: it is very necessary -> it is necessary
Page 2, lines 72-74: I would change the sentence: "Due to space limitations, this work summarizes only the research progress of cooperative operation, one of the key technologies of agricultural multiple robot systems.". Journal papers do not really have a limit for space; I think it is better to just say that the work focuses on cooperative operation.
Page 3, line 99: Multiple rotary tillers were plowed -> Multiple rotary tillers plowing
Page 3, line 100: Multiple seeders were sowed -> Multiple seeders sowing
Page 3, line 100: Multiple drones were sprayed -> Multiple drones spraying
Page 3, line 101: Multiple harvesters harvested -> Multiple harvesters harvesting
Page 3, line 109: are undulating, uneven and uneven -> are undulating and uneven,
Page 3, line 117: with high quality[20] -> with high quality [20]
Page 3, line 119: in the ground cross The -> in the ground cross. The
Page 3, line 125: To solve above problems -> To solve the above problems
Page 5, line 174: (These sensors -> these sensors
Page 12, line 344: Distributed distribution -> Distributed allocation
Page 16: has a large empty space that should be filled

Author Response

Dear reviewer:

I am very grateful for your comments on the manuscript on 15 Jan 2021. According to your advice, we amended the relevant part of the manuscript. Some of your questions were answered below.

1)

The reviewer’s comment:

I would change the wording about the three types of agricultural multi-robots to clarify that these are the three types that the review focuses on. There are however other types, such as for example USVs (Unmanned Surface Vehicles).

The author’s answer:

The section on robot classification has been revised at the end of the first part of the article.

"There are many types of agricultural multi-robots, and this article mainly focuses on unmanned aerial vehicles (UAVs), unmanned ground vehicles (UGVs), and unmanned underwater vehicles (UUVs). "

2)

The reviewer’s comment:

Since this is a review paper, it needs a section explaining what kind of methodology was used. I assume this was a systematic review of the literature, but this should be made explicit. It would also help to know which indexing research websites were used, what were the terms, how many papers were found, how many papers were discarded for being out of scope, etc.

The author’s answer:

The article is described at the beginning of Section 3.

"Since each technology does not solve the same agricultural problems, this section first classifies the types of development of these technologies, then describes and reviews each of these cooperative technologies in terms of research methods or problem solving, and finally summarizes their research development status and characteristics."

Also  the table is presented in Section 3.2 onwards to facilitate a clear understanding and search

3)

The reviewer’s comment:

While the format of describing a multi-robot concept (such as environmental perception), citing agricultural research that has used it (in general), and then specifying what most of the approaches do is interesting at first glance, it looses the information of which particular approach used which technique. For example, the following paragraph:
"In the processing of environmental perception information, agricultural multi-robots mainly adopt the feature layer information fusion method."
Tells us an important information, however, it is missing citations that indicate which ones use this method, something that would be valuable for readers of this review to know.

The author’s answer:

To represent the development of environmental perception more objectively and without causing disagreement. I have modified this section. By describing the current applications of agricultural multi-robots in this area, the current status and problems of the development of this technology are illustrated.

4)

The reviewer’s comment:

Similarly, it would be nice if there was more detail about at least some of these approaches (e.g. the more influential). Explaining how they used these multi-robot concepts, the scenario of application, and what were the results obtained. For example, Table 5 has various applications of centralised and distributed task allocation, however these applications are not further explained in the text.

The author’s answer:

I add more application of the technique of task assignment to agricultural multi-robots and explain how these concepts were applied, what problems were solved, and what the results of the solutions were. Since the original text was modified more, I will not describe it in detail here and hope that the changes are consistent with your suggestions.

5)

The reviewer’s comment:

Paragraphs from lines 438 to 459 in page 15 have a contradiction. First it is said that distributed path planning can only find suboptimal solutions (438-439), and then when describing the method given as example in agricultural multi-robots it is said that "The practice shows that the method can find the optimal solution" (456).

The author’s answer:

This was an incorrect description on my part and has been revised to a suboptimal solution.

5)

The reviewer’s comment:

I was expecting to find a bit more in the Discussion section. It seems quite short for the amount of research that was reported. Future directions are also missing and would be a good addition to that section.

The author’s answer:

I have rewritten the discussion and conclusion sections, detailing the development challenges of the technology in the discussion and pointing out in the conclusion that in response to the challenges future.

6)

The reviewer’s comment:

Minor comments and suggestions

All mentions of "et.al" should be replaced by "et al.".
There are several instances where there is no space between a word and the citation, for example "multiple robots possible[21]". I point some below but a full check has to be made whenever there is a citation.
Some figures have very small text (e.g. Figure 9).
The tables look a bit strange to me, specially the ones that break pages. This is more a matter of style, but nonetheless tables over page breaks should definitely be avoided (perhaps
Page 1, line 40: the physical of the elderly -> the physical health of the elderly
Page 2, line 61: of operators[10] -> of operators [10]
Page 2, line 70: it is very necessary -> it is necessary
Page 2, lines 72-74: I would change the sentence: "Due to space limitations, this work summarizes only the research progress of cooperative operation, one of the key technologies of agricultural multiple robot systems.". Journal papers do not really have a limit for space; I think it is better to just say that the work focuses on cooperative operation.
Page 3, line 99: Multiple rotary tillers were plowed -> Multiple rotary tillers plowing
Page 3, line 100: Multiple seeders were sowed -> Multiple seeders sowing
Page 3, line 100: Multiple drones were sprayed -> Multiple drones spraying
Page 3, line 101: Multiple harvesters harvested -> Multiple harvesters harvesting
Page 3, line 109: are undulating, uneven and uneven -> are undulating and uneven,
Page 3, line 117: with high quality[20] -> with high quality [20]
Page 3, line 119: in the ground cross The -> in the ground cross. The
Page 3, line 125: To solve above problems -> To solve the above problems
Page 5, line 174: (These sensors -> these sensors
Page 12, line 344: Distributed distribution -> Distributed allocation
Page 16: has a large empty space that should be filled

The author’s answer:

The problems you pointed out I have changed, as for the large blank because the table is too long or the picture is too large, using the page break layout caused, I have tried to modify.

Round 2

Reviewer 3 Report

In general, I am fine with the current version of the paper. The authors have made great work to improve it. In fact, they did not exactly and completely follow all my recommendations. Nevertheless, my notes and recommendations did not go unanswered or other reactions. The authors corrected some places of the paper according to their own opinions and I am generally fine with them. They also have made some extra reasonable corrections and improvements, possibly, according to other reviews.

Currently, the paper looks much better regarding its completeness, soundness and regarding to its logic. The only remark is that the link indicated in the end of the paper (in line 1046) still does not work.