Automatic Design of Collective Behaviors for Robots that Can Display and Perceive Colors
Round 1
Reviewer 1 Report
The paper describes a research in the field of swarm robotics. It proposes and analizes TuttiFrutti: an automatic design method for robot swarms. It focuses on the use of colors, to widen collective behaviors.
It is interesting and well-written and goes in depth with the description of the whole process, including methodology.
My only suggestions to enhance this work are the following:
- try to connect the visions of colours with what happens in nature, for example with the advantage for animals and animals who live in groups
- widen the discussion to include a reference to what can be achieved above the missions you cite
Author Response
We thank the reviewer for the positive and constructive feedback. Please find below a detailed response to the comments in the review.
Point 1:
The paper describes a research in the field of swarm robotics. It proposes and analizes TuttiFrutti: an automatic design method for robot swarms. It focuses on the use of colors, to widen collective behaviors.
It is interesting and well-written and goes in depth with the description of the whole process, including methodology.
Response 1:
We thank the reviewer for the comments.
Point 2:
Try to connect the visions of colours with what happens in nature, for example with the advantage for animals and animals who live in groups
Response 2:
Although a considerable part of the literature in swarm robotics it is tightly connected to biological systems, in the last years the field has shifted to become an engineering discipline that stands on its own [4]: much more research is devoted to the development of methodologies and tools pertinent only to robotic systems, which need not be inspired or related to what is observed in nature [3]. Accordingly, we contend that references to biological systems are not necessary for developing the ideas we present in the paper. Indeed, literature in swarm robotics already provides many examples that demonstrate how groups of robots can use color-based information to collectively perform tasks.
On the other hand, the focus of the journal is on engineering and we build our arguments in the paper with a focus on engineering too. We do not have the expertise in biology to make convincing arguments that could relate the behaviors observed in biological systems with those designed by TuttiFrutti. We believe that including references to biological systems could deviate the attention of the reader and led to comparisons that we are unable to address properly in the paper.
For the reasons mentioned above, we wish to not include references to biological systems in the document.
These references are already cited in the document:
[3] Dorigo, M.; Birattari, M.; Brambilla, M. Swarm robotics. Scholarpedia 2014, 9, 1463. doi:10.4249/scholarpedia.1463.
[4] Brambilla, M.; Ferrante, E.; Birattari, M.; Dorigo, M. Swarm robotics: a review from the swarm engineering perspective. Swarm Intelligence 2013, 7, 1–41. doi:10.1007/s11721-012-0075-2.
Point 3:
Widen the discussion to include a reference to what can be achieved above the missions you cite
Response 3:
As suggested by the reviewer, we extended the discussion in Section 5 (lines 563 - 593) in a new subsection.
In the new subsection, we first address the research questions presented in Section 1, and afterwards, we highlight what we consider are important considerations that go beyond the missions we studied: TuttiFrutti, and automatic design methods with shared characteristics, can close the gap between the complexity of the collective behaviors that have been designed manually for robot swarms, and those that can be designed with automatic methods.
Reviewer 2 Report
The topic is important and relevant.
Structure and content: The introduction provides sufficient background and the literature review is pretty comprehensive. The method is adequate and results clearly described. The manuscript is well-organized and the supplementary material is also helpful.
Overall, this study illustrated how TuttiFrutti, an instance of AutoMoDe specializing in the design of robots swarms that act according to color-based information, advanced previous instances of AutoMoDe, by performing missions in complex environments and enlarging the variety of collective behaviors. Suggestions and question are as followed.
- Conclusions section: In additions to the summary of key findings and future work, a paragraph related to main contribution as well as room for improvement, or possible real-life application, might help readers to grasp how to comprehend the significance of this study within the literature of swarm robotics.
- Four research questions were presented on P. 2, from Lines 56-60. I am not quite sure whether they were answered in Results and discussion, or in Conclusions.
- Abstract: The current version is short and concise. But it might be nice to add a sentence about results.
Author Response
We thank the reviewer for the valuable feedback. Please find below a detailed response to the comments in the review.
Point 1:
The topic is important and relevant.
Structure and content: The introduction provides sufficient background and the literature review is pretty comprehensive. The method is adequate and results clearly described. The manuscript is well-organized and the supplementary material is also helpful.
Response 1:
We thank again the reviewer for the comments.
Point 2:
Conclusions section: In additions to the summary of key findings and future work, a paragraph related to main contribution as well as room for improvement, or possible real-life application, might help readers to grasp how to comprehend the significance of this study within the literature of swarm robotics.
Response 2:
As suggested by the reviewer, we extended Section 6 (lines 618 - 622) with relevant findings and room for improvement of our research. In addition, we extended the discussion in Section 5 (lines 563 - 593) in a new subsection, and it now includes key elements that we think could be subject of further study with TuttiFrutti, or with automatic modular design methods that share its characteristics. We think that providing such discussion could help the reader to find the significance of our research within the existing literature in swarm robotics.
We do not wish to include references to possible real-life applications since such discussion is far beyond the scope of the paper. Indeed, we think that building arguments about how the automatic modular design of robot swarms can be further used in real-life missions requires an in-depth comparison of the specifications of those missions and the missions we consider in our experimental set-up. We agree that the discussion is interesting and necessary, however, we believe that it deserves its own development in future work.
Point 3:
Four research questions were presented on P. 2, from Lines 56-60. I am not quite sure whether they were answered in Results and discussion, or in Conclusions.
Response 3:
Indeed, we did not answer the questions explicitly. In the revised version of the paper, we address these questions in a new subsection in Section 5 (lines 563 - 593).
Point 4:
Abstract: The current version is short and concise. But it might be nice to add a sentence about results.
Response 4:
As suggested by the reviewer, we extended the abstract and included relevant findings (lines 19 - 22).
