Low-Cost Data, High-Quality Models: A Semi-Automated Approach to LOD3 Creation

Round 1

Reviewer 1 Report (Previous Reviewer 3)

Comments and Suggestions for Authors

Thanks for the revised manuscript, however, there are still many problems within this paper. Some problems remained from the previous version.

1 The abstract should contain the research context, research problem, research method, research results, and research impact.

2 The structure of the paper should be improved to be more logical; Many of the sections are written in a discursive way.

3 The use of language should be greatly improved, as well as the format of the manuscript, there are many formatting issues.

4 line 17: FME is short for ‘Feature Manipulation Engine’, not ‘Feature Modulation Engine’.

5 Figure 1 is not cited in the main text.

6 Line 381: it appears this is not the final version for submission, a citation is missing.

7 For the discussion section, use subsections to group contents.

8 For the conclusion section, focus on the findings and significance of your study.

Overall, this paper mainly presents what the authors did without sufficiently explain why they did the research, which is also important.

Comments on the Quality of English Language

The quality of English language should be greatly improved.

Author Response

Thank you for your thoughtful review of our manuscript. We appreciate your valuable feedback on the clarity of our paper's objectives and motivations. Your insights have been instrumental in guiding us toward improving the overall quality of our manuscript, and we are committed to addressing this concern effectively. Once again, we sincerely appreciate your time and effort in reviewing our work.

Author Response File: Author Response.pdf

Reviewer 2 Report (New Reviewer)

Comments and Suggestions for Authors

The manuscript reads very well, has a clear structure and interesting results. The paper fits well with the current theme of integrating theoretically two dichotomous technologies, BIM and GIS. However, a small revision is necessary. Please make the following comments:

- the acronym LiDAR should be developed in the introduction,

- the integration of BIM into GIS is a leading scientific topic and should therefore be emphasised more strongly in the introduction,

- the introduction lacks at least one paragraph on BIM; the author(s) go straight to the BIM thread without explaining to the reader how to understand it, for which I recommend reading:

https://doi.org/10.36680/j.itcon.2023.034

- the author(s) use the CityGML convention, but the BIMusually uses UK or US standards, at least one sentence should mention this,

- in the context of results and/or discussion: BIM models need to be semantically consistent and predictable to be useful (pragmatic), this is worth mentioning, I recommend reading: https://www.sciencedirect.com/science/article/pii/S2095263523000420

My decision is 'minor revision'. Once all comments have been made, the article can be published.

Author Response

Authors are grateful for your valuable feedback and constructive comments, which have greatly enhanced the quality and depth of the research.

Author Response File: Author Response.pdf

Reviewer 3 Report (New Reviewer)

Comments and Suggestions for Authors

This paper presented a workflow combining Apple LiDAR and Photogrammetric point clouds to create a BIM. The innovation of this paper is insufficient, but its industrial application is worthwhile. The manuscript still needs to be revised.

The following are my comments:

(1) In line 95, I suggest that this part be changed to CONTRIBUTION. And how did you resolve these issues;

(2) The key step in the application using LiDAR and Photogrammetric point clouds is how to unify them into the same coordinate system, how to solve the unavoidable coordinate deviation;

(3) In the Result, there is a lack of data description and there is insufficient experimental data, so I suggest adding more data for the experiment.

Author Response

Thank you for taking the time to review our paper. We appreciate your feedback regarding the workflow presented in our study, which combines Apple LiDAR and Photogrammetric point clouds for BIM creation. While we acknowledge your perspective on the level of innovation portrayed in the paper, we believe that the practical implications and industrial applications of our proposed methodology are indeed significant. We understand your suggestion, and we have attempted to enhance the clarity and depth of our manuscript. Our aim was to ensure that our work meets the highest standards of academic rigor and contributes meaningfully to the field.

Once again, we thank you for your valuable comments and constructive criticism, which have undoubtedly aided us in refining our research for the benefit of the scientific community.

Author Response File: Author Response.pdf

Round 2

Reviewer 1 Report (Previous Reviewer 3)

Comments and Suggestions for Authors

Thanks for the revised manuscript. A response letter for addressing comments from another reviewer has been mistakenly submitted. So, I cannot see how each of my previous comments is addressed from the response letter. I read the paper again and found the quality has been improved but many problems have not been properly addressed. The detailed comments are as follows.

1 The title ‘Creating LOD3 with Apple LiDAR & UAV Photogrammetry’ is vague. The title should show the key elements of your research.

2 The abstract should be further improved. Use 1-2 sentences to describe research context, research problem, research method, research results, and research impact.

3 Create a flowchart for the methods. The current section for methods is not well organised, it is more like a record.

4 For the discussion section, use subsections to group contents.

5 Line 577: ‘This article presented a 3D modelling workflow to employ a variety of readily available 577 techniques for acquiring, disseminating, and reconstructing an information model in 578 LOD3 within a GIS environment’ So, what is the contribution of this study if everything is off-the-shelf. Are there any new algorithms developed for better geometry generation?

6 Point cloud can be directly used to generate CityGML models, why are they converted into IFC models first and then converted into CityGML?

7 Others

Line 13: The first time an acronym is used, the full term should be given, e.g., ‘LOD3’ in line 13.

Line 16: IFC is short for ‘Industry Foundation Classes’, not ‘Industry Foundation Class’

Line 17: ‘Feature ManipulationEngine (FME)’

Line 20: ‘CityGML presently includes three levels of detail (LODs)’, CityGML2.0 has 5 levels of detail (LoD0-LoD4), while CItyGML3.0 has 4 levels of detail (LoD0-LoD3).

Overall, the authors should think about the scientific contribution of this study and revise the introduction, literature review, and method sections accordingly surrounding that.

Comments on the Quality of English Language

The use of English should be improved.

Author Response

Thank you for your message, and we apologize for any confusion caused by the mistakenly submitted response letter. We understand the importance of addressing each of your previous comments thoroughly, and we appreciate the opportunity to clarify how they have been incorporated into the revised manuscript. It has been resubmitted and notified to the publisher, meanwhile we are reattaching here again for your reference (in the Last page of the attachment). We appreciate your valuable feedback and your continued support throughout the review process. We have prepared a detailed summary of how each of your comments has been addressed in the attachment.

Author Response File: Author Response.pdf

Reviewer 3 Report (New Reviewer)

Comments and Suggestions for Authors

All my questions are answered. 

Author Response

We appreciate your correspondence and are delighted to learn that every one of your inquiries has been adequately responded to. Our utmost dedication is to guarantee the accuracy and comprehensiveness of our replies to your remarks. We express our gratitude for your conscientiousness and comprehensive evaluation of our manuscript.

This manuscript is a resubmission of an earlier submission. The following is a list of the peer review reports and author responses from that submission.

Round 1

Reviewer 1 Report

Comments and Suggestions for Authors

Summary:

This paper provides a complete semi-workflow from field photogrammetric point by Apple LiDAR measurements to BIM (Building information modeling) then to the generation of CityGML (Global Markup Language) models with geometric information and semantics using Feature Manipulation Engine 80 (FME Software). The detailed background knowledge of Industry Foundation Classes (IFC) and FME has been illustrated sufficiently, including a description of the software functions and methods. The implementation of new technologies (Apple LiDAR,IFC and FME) is sound, especially the latest Apple LiDAR hardware is used to assist.

Although some errors were found in the information transferring between models in quality assessment session, alternative solutions are given for some of the cases for the 3D modelling. The interoperability between IFC and CityGML data models was met during the conversion process through the validation from BIM and GIS. Eventually, the defects with the input IFC model and the conversion workflow can be fixed by identifying the errors in the CityGML output.

 

Here are some questions I had thrown the paper:

-Section 2.1, “Images have been acquired using DJI Phantom 4 Pro which comes with a pre-installed 20 190 MP FC360 camera, in an oblique mode to cover the roof as well as façade information of 191 the building.” DJI Phantom 4 pro is another device that uses its own language, how does it collect the photogrammetric points needs to be elaborated. I think transferring data between various software systems (such as from BIM to IFC) is hard but possible and they are likely a time-consuming task. But obtaining raw data from site is more significant to any project. How is the accuracy, reliability from the data of “pre-installed 20 190 MP FC360 camera’’? Does the flying vibration affect the data collection (wind effect)? Always have Enough GPS signal for the Drone? How can we confirm the data once the building is half built and the drone needs to fly into the rough-in building? Data will be reliable? It is a GIS paper, but is it far from the real industry application?

-The method of how the digital model is built in the Revit environment should be explained in more detail (in section 2.1-2.2) leading up to the information transferring of this paper, as it is the basis for the geometric information and subsequent steps. A clearer explanation on any data transform to materialize the data transfer would be helpful.

-This article could have been more readable if it is focused more on the methodology and technical details. Enough presentations have been given that these could be reduced and reordered. Figure 1,6,9,10,13,14 is very clear for the purpose of this paper, but other figures serve as a user manual for the CityGML, such as figure 11,12. I am not saying they are not useful for explaining purpose of the information, but it is over-done. I’d like to see more results figures from the paper. Also, worth noting is the order of the graphs. Some of the graphs are incorrectly positioned. More details could be added.

-It would also make sense to put more effort into correcting the output CityGML model since the authors noticed some errors in the results part. How those errors were found? From the raw data or from defects in information transferring and sharing? How to address the defects? Percentage numbers, analysis results would be helpful for support the results.

 

-Overall, it would have been better if the paper had focused more on technical details and corresponding analysis. The background (literature review) can be streamlined as some of them are unnecessary for this level of paper. I always have the feeling of read a “user manual” for the entire section 2. “Result and discussion” is insufficient to me. 

-Lack of numerical supports in conclusion and discussion part, or to the entire paper. The paper indeed is focus on “Creating 3D model”, there are plenty of jargon about software input and robust features in the paper (JSON, SIFT, SURF, ORB, AKAZE, FLANN, BOW), but what I expect to see is not clear yet, that is how a set of raw data from drone is transferred to a “CityGML” model.

 

Comments on the Quality of English Language

The paper needs to be edited, there are some missing words in the sentences. 

Quality of Figures are low, some screenshots showing important data structures or unique methods are not legible.

Author Response

Authors appreciate the efforts and time provided by the reviewer from their busy schedule for reviewing this article. Authors are grateful for excellent feedback and have tried to revise the manuscript with reference to reviewer’s suggestions and recommendations. A detailed response has been provided below with specific comments of the reviewer in reference. Your thoughtful and insightful comments have been invaluable in improving the quality and impact of our research. We assure you that we have carefully considered each of your comments and suggestions and necessary revisions have been made to address the identified concerns and improve the clarity of the manuscript.

Author Response File: Author Response.docx

Reviewer 2 Report

Comments and Suggestions for Authors

The paper explores the potential of constructing a three-dimensional (3D) representation of urban entities by leveraging a combination of various software tools. Its primary aim is to address the challenges associated with developing an urban digital twin, which is accomplished through the creation of a 3D model within a Geographic Information System (GIS) environment. This model is generated using cutting-edge techniques in 3D geospatial data modeling, specifically utilizing Apple LIDAR technology. By adopting such an approach, the paper seeks to provide insights into how to effectively establish a comprehensive and accurate digital twin with a Level of Detail 3 (LOD3) directly from the acquisition process. The paper describes a semi-automatic workflow, articulated into three parts. The first step, concerns the acquisition of a point cloud by using Apple LiDAR. Then a BIM model was created in Revit sw and exported to IFC. Finally, the IFC file is transformed into a GML file using Feature Manipulation Engine. The process is interesting and can indicate a good way to obtain a GML model by using simple enough tools and sw. A critical remark could concern the fact that the tools and the software are not very cheap and moreover, they don't belong to the open-source world. However, the overall opinion of the paper is certainly positive.

Comments on the Quality of English Language

The quality of English is good. In some cases, it would be useful to use different phrases, but this does not affect the quality of the paper.

Author Response

Authors are grateful for your valuable feedback and constructive comments, which have greatly enhanced the quality and depth of the research. We are delighted to hear that you found the article to be positively impactful. We agree that inclusion of the open-source software and algorithm will better the implication of the acquired workflow and are working towards developing such workflows in future developments. We firmly believe that open-source technologies play a crucial role in advancing research, fostering collaboration, and democratizing access to knowledge. 

Author Response File: Author Response.docx

Reviewer 3 Report

Comments and Suggestions for Authors

This paper presents the authors’ work on creating BIM models and using them in GIS by jointly using Apple LiDAR, UVA photogrammetry and FME. There are mainly three steps involved. (1) Apple LiDAR was used to scan the indoor space, while UAV photogrammetry was used to take photos of the exterior of building. (2) BIM model was manually created from the point clouds and exported as IFC (Industry Foundation Classes), and (3) FME was used to convert IFC into CityGML. Overall, this paper explores an important topic and is easy to understand. However, there are some problems about paper structure and methodology. The detailed comments are as follows.

1 Paper structure

This paper has an acceptable overall structure, but there are problems with some sections.

1) The introduction section has provided background of the research. There are many concepts involved but not well organized.

2) A section for literature review is missing. The literature review section should provide an overview of previous studies and identify a research gap, which is used to justify the method.

3) The discussion section is simply an overview of the introduction section and the method section.

2 Methodology

1) This paper proposed to use cutting-edge mapping techniques to establish a methodology for creating BIM models and use them in GIS. While a new methodology is proposed and tested, a comparison with the traditional methodology is required.

2) A systematic view on IFC-to-CityGML conversion is required. Please refer to ‘BIM/GIS data integration from the perspective of information flow’

3) A better understanding of IFC is required.

a) Line 343: ‘the opening is the child of the wall’, ‘opening’ is not the child of ‘wall’, it is just related to the wall, please refer to Figure 2 in https://doi.org/10.3390/rs13224727.

b) Line 416: ‘IFC is based on “ConstructiveSolidGeometry” and “SweptGeometry” whereas CityGML on “BoundaryRepresentation’. IFC uses all the three types of geometry.

3 Writing and others

1) The first time an acronym is used, give the full term, e.g., IFC

2) Line 113: ‘in formulating the idea of ’ - contents are missing

3) Line 118: ‘Forstner and Wrobel. in 2015’ -> ‘Forstner and Wrobel in 2015’

4) Line 124: be consistent in terminology, such as ‘LiDAR’

5) Line 213: it should be ‘Figure 2’

6) Line 214: improper capital letter ‘Understand’

7) Line 239: improper capital letter

8) Over-explanation for ‘IFC’, line 241: ‘It is further exported in IFC format which is an open standard CAD data exchange data schema for BIM’, line 249: ‘IFC, which is an openBIM data exchange format keeps all the modelling information intact and provides user to view,’ IFC does not need to be explained in such a way.

4 Suggestion

Proofread the manuscript to improve the overall writing quality.

Comments on the Quality of English Language

Overall, this paper is easy to understand, and the quality of English is good, despite some minor problems with writing, such as improper use of capital letters.

Author Response

Authors truly appreciate the thoroughness with which you have reviewed the manuscript and provided constructive feedback. We wanted to express sincere gratitude for taking the time and effort to review our article. Your thoughtful and insightful comments have been invaluable in improving the quality and impact of my research. We are particularly grateful for the specific recommendations and suggestions you provided to strengthen various aspects of the article. Your guidance on refining the methodology, analyzing the results, and discussing the implications has been invaluable in shaping the article's overall structure and content. Your expertise in the field has certainly helped me gain new insights and perspectives. We assure you that we have carefully considered each of your comments and suggestions. Necessary revisions have been made to address the identified concerns and improve the clarity of the manuscript. 

Author Response File: Author Response.docx

Round 2

Reviewer 1 Report

Comments and Suggestions for Authors

Please see comments for Editor

Author Response

Your valuable insights have played a pivotal role in enhancing the quality and significance of our research. Rest assured that we have thoroughly considered each of your comments and suggestions, and as a result, we have made necessary revisions to address the identified concerns and enhance the overall clarity of the manuscript. We extend our heartfelt appreciation for dedicating your time and effort to review and provide feedback on our article. 

Reviewer 3 Report

Comments and Suggestions for Authors

Thanks for the revised manuscript. Many problems have been addressed, but not all.

1 The structure of the introduction section should be improved to be more logical.

2 The challenges listed between line 98-102 are just tasks to be completed.

3 The background section contains many elements that are simply listed without logic.

4 The discussion section looks like a combination of introduction and methods. The discussion section should cover how the research question has been addressed, any new problems discovered, the limitation of your study, etc.

5 The use of language should be greatly improved.

6 others

1) Line 81: IFC is short for ‘Industry Foundation Classes’

2) Line 82: GML is short for ‘Geography Markup Language’, not ‘Global Markup Language’.

3) Line 92: full term for ‘DBMS’

4) Line 328: ‘The Information retrieved’

Comments on the Quality of English Language

The quality of English should be improved.

Author Response

We wanted to express sincere gratitude for providing the time and effort to revise the changes in our article. Your insightful comments have been invaluable in improving the quality and impact of my research. We assure you that we have carefully considered each of your comments and suggestions. Necessary revisions have been made to address the identified concerns and improve the clarity of the manuscript. 

Author Response File: Author Response.docx

Round 3

Reviewer 1 Report

Comments and Suggestions for Authors

I couldn't find response or revision to previous comments; the relevant paper is still not cross referenced to show the originality, difference...

 

 

Reviewer 3 Report

Comments and Suggestions for Authors

Thanks for the revised manuscript. Many problems have not been addressed, and more problems/questions have been introduced during the revision. The unsolved/new problems are as follows.

1 The structure of the introduction section should be improved to be more logical.

2 The authors should clarify the following questions. Why Apple’s LiDAR-embedded products are used in this study? Can it solve problems that cannot be solved by conventional LiDAR device? Can it create new application cases that cannot be completed by conventional LiDAR device? Are there any differences in the data collecting procedure? Otherwise, the manuscript is more like a project report, rather than a scientific paper.

3 Line 144: Using FME to convert IFC into GML is not a research problem.

4 The use of language should be greatly improved.

5 Others

1) line 19: FME is short for ‘Feature Manipulation Engine’, not ‘Feature Modulation Engine’

2) line 45: ‘This study aims to establish a data-driven workflow for crafting digital twins by harnessing two 3D geospatial data modelling techniques: photogrammetry and LiDAR’. We already have such a workflow. I thought the innovation is about how to use the handheld device in this workflow, and how it compares with those conventional photogrammetry and LiDAR devices.

Comments on the Quality of English Language

The quality of English should be greatly improved.