RTPO: A Domain Knowledge Base for Robot Task Planning
Round 1
Reviewer 1 Report
The authors preset a work for arranging the knowledge used for robot task planning. This approach is named Robot Task Planning Ontology (RTPO) and is divided in three main parts: task ontology, environment ontology, and robot ontology. Inside each of these, the knowledge is hierarchically organized in sub classes. Also, the authors present an experiment to support the idea that this organization helps to perform a robot planing task. Moreover, the development of this work is based in useful technologies, like ROS to the node and communication implementation and prolog-protege for the query definition and knowledge organization.
But after all, it is not clear the contribution of this work. In case the contribution is just the knowledge organization, the authors must quantitatively demonstrate that this knowledge organization is better than other schemes. As the authors note in the manuscript, there are more works addressing same problem: KnowRob, ORO, SWARMs (cited by them). Therefore the authors have to used at least one of this as baseline.
The use case devote to prove the performance of this scheme is not supporting the thesis of authors. Mainly because they do not demonstrate the flexibility of this approach, what is actually the most desirable skill in this kind of contributions.
Remarks.
Figure 4 does not correspond with the labels (neither what it is talked in the following paragraphs). It seems that the label has a correlation with figure 11. Figures 6 to 9 give marginal information to the manuscript due to the arrow mess Manuscript writing should be improved, sometimes it is difficult to follow the ideas due to the writing. Also, there are multiple typos in the paper, e.g. Figure 2 “sofeware” ; page 7 paragraph 1 line 242 “ .. The second kind of knowledge is present in causal knowledge…”Author Response
Please see the attachment.
Author Response File: Author Response.pdf
Reviewer 2 Report
Dear Author,
In my opinion the paper is well written, the problem is described correctly and enough references have been included. The experiments validate the robustness of the algorithms proposed.
But, I would like to include some comments and suggestions:
The contributions and novelty of the work must be clearly mentioned in the manuscript. I propose to include some KPIs (key performance indicators) to measure the efficiency of the proposed high-level decision-making algorithm.Best regards
Author Response
Please see the attachment.
Author Response File: Author Response.pdf
Round 2
Reviewer 1 Report
Dear Authors
Thank you for answer me carefully all my suggestions and do this hard work in the improvement of this manuscript “RTPO: A Domain Knowledge Base for Robot Task Planning”. I have not more comments to add for this work.
Regards,
A reviewer