An Indoor Space Model of Building Considering Multi-Type Segmentation
Round 1
Reviewer 1 Report
I reviewed the previous submission of this paper and provided some comments. Obviously, many of these problems have been properly addressed. However, the main problem regarding research motivation and contribution has not been well addressed.
1 Motivation and contribution. In this version, the main logic remains the same: the current classification cannot meet the needs of indoor space analysis -> propose a new classification by extending IFC -> conduct experiments to make sure the new classification system works. There are mainly two problems. (1) It is not sufficiently explained why the current classification cannot meet the needs of indoor space analysis; they just made a statement without fully investigating that. That is why I said this study does not have a strong motivation. (2) For such a study, just showing that the new system works is not enough. It needs to show that the new system is better than others, under a certain context, because we already have working classification systems, if you want to propose a new one, you need to show it is better. Otherwise, you cannot persuade other researchers to use your classification system.
2 The new classification system. The new classification system, i.e., building space, floor space, household space, room space, and functional subspace, is just based on the authors’ observation of a residential apartment building. They have not provided the answer to ‘Why’, why do we need that? Also, this classification system may not be applied to standalone houses, or commercial buildings due to the inclusion of ‘household space’.
4 Others
1) line49-53: please mention that the newly released CityGML3.0 now only has four levels of LoD.
2) line 86-87: it is claimed that ‘however, the segmentation of room space by furniture is ignored (by IFC)’. I do not think this is a problem, as that is not the original purpose of furniture. Furniture can be used as a way of separation but that is not standardised.
3) Table 3. Connectivity and adjacency information has been described by the IfcRelConnects relationship or its subclasses, such as IfcRelSpaceBounday. For the ‘segmentation’, there is an attribute defined in IfcSpace, i.e., ‘CompositionType’, which indicates if a space instance is a space group (COMPLEX), a single space (ELEMENT), or a partial space (PARTIAL), which means a space can contain one or more subspaces.
Suggestion: Please focus more on a use case, and then develop a classification system that suits that use case, just like other studies. It will show a solid motivation for your study, otherwise, it will be like I make the classification system more comprehensive just because I want to make it more comprehensive.
Author Response
Response to Reviewer 1:
We appreciate the reviewer’s valuable comments. We have addressed the comments as follows and have revised the manuscript accordingly.
Point 1:
It is not sufficiently explained why the current classification cannot meet the needs of indoor space analysis; they just made a statement without fully investigating that. That is why I said this study does not have a strong motivation.
Reply: Thanks for your helpful suggestion. The motivation of this paper is to solve the problem of incomplete spatial model in the application of indoor analysis and management. The contribution of this paper is to extend the existing IFC spatial model to include the representation of household and functional subspaces from entity, relationship, property. We have made additional explanations in this paper, especially the "97-111" line.
“For example, fine navigation based on functional subspace division, payment management based on household space, etc.”
“How can the existing IFC specifications be extended to include the expression of household space and functional subspace, so that it can support refined analysis and management applications, is the research question that this article needs to focus on.”
Point 2:
For such a study, just showing that the new system works is not enough. It needs to show that the new system is better than others, under a certain context, because we already have working classification systems, if you want to propose a new one, you need to show it is better. Otherwise, you cannot persuade other researchers to use your classification system.
Reply: Thanks for your helpful suggestion. In order to support the application analysis on household space and function subspaces, this paper proposes the new extended space model based on IFC. The model divides indoor space objects into five levels, namely, building space, floor space, household space, room space and functional subspace. Compared with the CityGML model and the IFC standard, it increases the level expression of the household space, and at the same time extends the smallest granular indoor space to the level of indoor functional subspace.
Point 3:
Also, this classification system may not be applied to standalone houses, or commercial buildings due to the inclusion of ‘household space’.
Reply: Thanks for your helpful suggestion. The indoor space model proposed in this paper is constructed based on the IFC standard, and the spatial structure elements in the standard are aggregated through the ‘IfcRelAggregates’ relationship. In the absence of ‘household space’, floor space also can be aggregated directly with common room space through this relationship, and thus model can be applies to standalone houses and commercial buildings.
Point 4:
Line 49-53: please mention that the newly released CityGML3.0 now only has four levels of LoD.
Reply: Thanks for your helpful suggestion. We have added the newly released CityGML3.0 as follow:
“CityGML supports the description of city objects from semantics, geometry, and to-pology. As one of the most detailed themes defined in CityGML 3.0, the building extension module allows building elements to be represented with four levels of detail from LOD0 to LOD3. The current version of CityGML 3.0 only regards the smallest granular indoor space type as the feature (Room), and internal facilities such as furniture can be represented in all LODs 0-3.”
OGC. Ogc CityGML. In Document No. 0-010, Available online: https://docs.ogc.org/is/20-010/20-010.html#toc30.
Point 5:
Line 86-87: it is claimed that ‘however, the segmentation of room space by furniture is ignored (by IFC)’. I do not think this is a problem, as that is not the original purpose of furniture. Furniture can be used as a way of separation but that is not standardised.
Reply: Thanks for your helpful suggestion. We have revised the related sentence from
“However, the segmentation of room space by furniture is ignored.”
To
“However, due to the placement of indoor furniture in the interior space, a smaller-grained functional space is formed, which is ignored by the current space model.”
Point 6:
Table 3. Connectivity and adjacency information has been described by the IfcRelConnects relationship or its subclasses, such as IfcRelSpaceBounday. For the ‘segmentation’, there is an attribute defined in IfcSpace, i.e., ‘CompositionType’, which indicates if a space instance is a space group (COMPLEX), a single space (ELEMENT), or a partial space (PARTIAL), which means a space can contain one or more subspaces.
Reply: Thanks for your helpful suggestion. IfcRelSpaceBounday is defined as an objectified relationship that handles the element to space relationship, it can only express the relationship between space and components. Adjacency and connectivity relationships refers to the relationship between different indoor spaces, such as functional subspaces. The segmentation relationship is the association relationship between indoor components and indoor space. And the attribute ‘CompositionType’ indicates if a space instance is a space group, but it confuses the hierarchical relationship between spaces,
Point 7:
Please focus more on a use case, and then develop a classification system that suits that use case, just like other studies. It will show a solid motivation for your study, otherwise, it will be like I make the classification system more comprehensive just because I want to make it more comprehensive.
Reply: Thanks for your helpful suggestion. We have developed a classification system that suits additional use case. This procedure is used to demonstrate the multi-level indoor space model proposed in this paper (Figure 17).
Figure 17. Indoor space classification system.
Author Response File: 
Author Response.docx
Reviewer 2 Report
Dear Authors,
Thank you for addressing all my concerns and I don't have any further comments on your paper. The paper is accepted from my side.
Best Regards
Author Response
Thanks for your valuable comments!
Reviewer 3 Report
I cannot accept several suggested responses to my review report for the first submission. Especially my points 2, 3, 9 and 13. See attached file.
Comments for author File: 
Comments.pdf
Author Response
Response to Reviewer 2:
We appreciate the reviewer’s valuable comments. We have addressed the comments as follows and have revised the manuscript accordingly.
Point 1:
The lines you refer to describe a statement, not a research question. You have described the contribution in lines 106-121, but you also need to describe a research question that defines what you want to investigate. Proposing a model is not enough, you need a question you can compare your results against. Something like “How can existing specifications be extended to …”
Reply: Thanks for your helpful suggestion. In order to express the research question of this paper clearly, we have added the following text to the introduction part of the paper:
“How can the existing IFC specifications be extended to include the expression of household space and functional subspace, so that it can support refined analysis and management applications, is the research question that this article needs to focus on.”
Point 2:
You need to answer by improving the paper, not by explaining to me. You have not described in the paper what specifications you have considered, and most important: you have not described why you have selected the three specifications for further studies. Before you start describing CityGML in line 48, you need something like “Several specifications for building spaces … exists … where the most prominent are considered … “. Add references to the specifications when you start describing them.
Reply: Thanks for your helpful suggestion. Before start describing CityGML, we have added the description of the specifications considered in this paper:
“Several specifications for building spaces have also been widely concerned, and the most prominent are considered include CityGML, IndoorGML and IFC.”
OGC. Ogc CityGML. In Document No. 0-010, Available online: https://docs.ogc.org/is/20-010/20-010.html#toc30.
OGC. Ogc indoorgml. In Document No. 14-005r4, Available online: http://www.opengeospatial.org/standards/indoorgml (accessed on 12 April 2017): 2014.
BuildingSMART. Industry foundation classes (IFC4.2). Available online: https://technical.buildingsmart.org/standards/ifc/ifc-schema-specifications/
Point 3:
You have described what IFC can do, but not why you chose to extend IFC, and not CityGML or IndoorGML.
Reply: Thanks for your helpful suggestion. In the Section 3 of this paper, we have added the following text to explain why we chose to extend IFC:
“The reason why this paper chooses to expand based on the existing IFC standards is that compared with other standards, IFC has more indoor space objects and relationships than CityGML and IndoorGML. ”
Point 4:
The fragment and the extended text is accepted. Note: I believe there is a wrong reference in line 554. Figure 15(a) should probably be figure 16(a).
Reply: We have revised this language mistake.
Point 5:
You have described what you have done, but not how it answers the intention of your research (the research question – see my point 1).
Reply: The same response with Point 1.
Author Response File: Author Response.docx
Round 2
Reviewer 1 Report
Thanks for providing the revised manuscript, many problems have been properly addressed, but not all of them.
For those unsolved comments, I would like to provide more comments.
1 Point 1 – This paper does not have a strong motivation, because the following questions have not been properly answered in the introduction and literature review sections. (1) Q1 – the authors only concluded that ‘the spatial organization of buildings, floor, and rooms can no longer meet the needs of refined indoor analysis, and a more detailed subdivision framework of indoor space model is required in indoor navigation analysis’, without explaining why previous studies made that statement. It is mostly because some types of information required by certain applications do not exist in the current standard. (2) Q2 – Based on Q1, the second question is, what kind of information is missing from the existing framework/standard? This is to be concluded from existing studies/applications. (3) Q3 - Based on Q2, the third question is, what are the requirements of the new framework? These questions need to be properly answered to justify your proposal and contribution.
2 Point 5 – if you use furniture to segment room space, the following questions need to be clarified. (1) Should every piece of furniture in an area be used for that purpose or just some of them? (2) If it is only some of them, how can we determine those to be used as boundaries?
3 Point 6 – In openBIM, it is common to extract the adjacency and connectivity information from IfcRelSpaceBoudnay, for example, http://dx.doi.org/10.1016/j.autcon.2016.04.001, which explicitly states that ‘In IFC, adjacency of two spaces can be defined through the space boundary relationship (IfcRelSpaceBounday)’.
4 There is a grammar error in the title ‘An building indoor space model’.
5 The authors stated in the abstract that ‘this study has advantages in terms of indoor model completeness’, the question is how do you assess if a framework is complete?
6 Line 43: the full version of ‘IFC’ should be given.
Author Response
Response to Reviewer 1:
We appreciate the reviewer’s valuable comments. We have addressed the comments as follows and have revised the manuscript accordingly.
Point 1:
This paper does not have a strong motivation, because the following questions have not been properly answered in the introduction and literature review sections. (1) Q1 – the authors only concluded that ‘the spatial organization of buildings, floor, and rooms can no longer meet the needs of refined indoor analysis, and a more detailed subdivision framework of indoor space model is required in indoor navigation analysis’, without explaining why previous studies made that statement. It is mostly because some types of information required by certain applications do not exist in the current standard. (2) Q2 – Based on Q1, the second question is, what kind of information is missing from the existing framework/standard? This is to be concluded from existing studies/applications. (3) Q3 - Based on Q2, the third question is, what are the requirements of the new framework? These questions need to be properly answered to justify your proposal and contribution.
Reply: We appreciate your helpful comments. According to your suggestions, we have made the following revision to the original text:
First, at the beginning of the introduction, we explain why previous studies no longer meet the needs of refined indoor analysis.
“With more and more environmental simulation and application analysis of indoor scenes, the spatial organization of buildings, floors, and rooms is difficult to meet the needs of refined and diversified indoor space analysis. It is mostly because some types of information required by certain applications do not exist in the current models and standards, for example, a more detailed subdivision information of the indoor space model framework is required in indoor navigation analysis [2-4].”
Then, after research and analysis to the existing frameworks/standards, the conclusion is what types of information are missing. In the text on lines 104-108, we added a description of this information.
“The semantic, relationship and attribute information related to the household space and the functional subspace are missing from the current standard. Focusing on the expression of a single space, and lacking a description mechanism for multi-level complex indoor spaces under considering spatio-temporal segmentation.”
Finally, this paper proposes a new framework, which requires these entities, relationships and attributes to be included.
“In view of the above-mentioned problems existing in the existing indoor space model, the contribution of this paper is to propose a multi-level unified expression indoor space data model considering the multi-type segmentation, which covers household space and indoor functional subspace.”
Point 2:
If you use furniture to segment room space, the following questions need to be clarified. (1) Should every piece of furniture in an area be used for that purpose or just some of them? (2) If it is only some of them, how can we determine those to be used as boundaries?
Reply: Thanks for your helpful suggestion. We have added related text in Line 217. We figured that every piece of furniture should be used to divide the interior empty space of room in this paper. In Section 2.2, we introduced two types of furniture entities participating in the subdivision of interior space: vertical division and horizontal division.
Point 3:
In openBIM, it is common to extract the adjacency and connectivity information from IfcRelSpaceBoudnay, for example, http://dx.doi.org/10.1016/j.autcon.2016.04.001, which explicitly states that ‘In IFC, adjacency of two spaces can be defined through the space boundary relationship (IfcRelSpaceBounday)’.
Reply: Thanks for your helpful suggestion. Indeed, the adjacency and connectivity relationships between spaces can be inferred from the space boundary relationship, which is an implicit expression of those types of relationships. However, when we pay more attention to the spatial model, we need to express the relationship directly and explicitly, so that it can be easily and quickly applied to indoor analysis. We have added this reference to the original text and explained it in Section 3.2.2.
Skandhakumar, N.; Salim, F.; Reid, J.; Drogemuller, R.; Dawson, E. Graph theory based representation of building infor-mation models for access control applications. Automation in Construction 2016, 68, 44-51, doi:10.1016/j.autcon.2016.04.001.
Point 4:
There is a grammar error in the title ‘An building indoor space model’.
Reply: Thanks for your helpful suggestion. We have revised this grammar error.
“An building indoor space model considering multi-type segmentation”
To
“An indoor space model of building considering multi-type segmentation”
Point 5:
The authors stated in the abstract that ‘this study has advantages in terms of indoor model completeness’, the question is how do you assess if a framework is complete?
Reply: Thanks for your helpful suggestion. We have revised the description in the summary:
“This study has advantages in terms of indoor model completeness”
To
“This study enrich the granularity of existing indoor models”
Point 6:
Line 43: the full version of ‘IFC’ should be given.
Reply: Thanks for your helpful suggestion. The full version of IFC is 4.3. We have added the related text of the version in Line 45.
Author Response File: Author Response.docx
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
The paper describes an extended model for indoor spaces, mainly by extending the IFC model. The purpose and results are original and significant, but there is a need for improving several parts. Besides, the language must be improved. Several vital paragraphs are hard to understand due to ambiguous language.
- It is difficult to identify the practical use case that the improved model is meant to serve. This should be described early in the introduction. Why is such an improved model needed? For what will it be used?
- The research question(s) is not clearly stated. Therefore, it is difficult to see if the results meet the purpose of the paper.
- There should be a more specific description of state of the art concerning existing specifications and prior research. For example - what existing models have been considered, and why were IFC, CityGML and IndoorGML selected? How was previous research identified? What are the weaknesses of prior work?
- In section 2.1, you describe the indoor space as separated from the outdoor natural environment. Is this true? What about an outdoor built environment such as a playground, a pedestrian zone etc. How is the indoor built environment different from the outdoor built environment?
- In section 2.2, the description of virtual segmentation should be illustrated in a figure.
- Section 2.3 introduces "household space" as one hierarchical level and relates it to property rights. This is a division based on legal principles, not on physical characteristics. This approach seems to narrow the scope of the model. What if there is a single owner of a whole floor or an entire building, but different users? I would prefer another term and definition for this level.
- For building space in section 2.3, you provide a list of eight categories. What is the origin of this list, and what is the basis for the classes? The class "office" overlaps several of the other classes.
- For the floor space in section 2.3, you describe a rule for refuge floors in buildings that are more than 100 m high. Is this a national requirement? It should have a reference in the paper.
- In clause 3, the extensions of IFC are described. But there is no explanation in the paper on why you selected to extend IFC and not any other specification.
- In line 369 (page 10), you describe IfcProxy as abstract. Technically, this doesn't seem right. The entity is not abstract in the EXPRESS schema.
- You have extended IfcProxy instead of defining new entities. The arguments for this approach are hard to understand, especially as the changes you suggest to IfcProxy creates a whole new model for IfcProxy.
- The case studies and the example in the discussion are simple illustrations. Data fragments should support them to show that they are based on actual tests of the model.
- The discussion and conclusion should point back to the initial purpose of the paper and describe how the results align with the research question(s).
Reviewer 2 Report
The paper proposes a multi-level indoor space framework based on different segmentation types, as well as an extension of indoor space definition using IFC. While the topic of the paper is significant and interesting, the authors must consider revising the manuscript significantly before it can be published in IJGI.
First, the abstract seems insufficient to describe the goal of the paper. Briefly and concisely, it must attract readers' attention while sufficiently discussing the paper's significance, objectives, and results. The abstract also contains statements that are not found or are insufficiently discussed in the paper (such as cases and discussions).
The introduction section lacks sufficient discussion in the motivation and background of the study. After a single paragraph, it jumps straight into discussing existing data models for spatial data. However, this discussion may be better placed in a separate section, along with the extensive discussion on related studies on the topic- which this manuscript currently lacks. Is the "article" in line 34 talking about the entire paper or only the section? Additionally, the paper does not sufficiently discuss the significance or the reason for pursuing its objectives.
Section 2.1 enumerates characteristics of Spatial Segmentation and Temporal Segmentation but does not define the meaning of these concepts. The divisions of temporal segmentation do not seem to be temporal characteristics too, and these are also spatial characteristics (building components and furnishing components). Figure 1 does not also reflect the contents of Section 2.1 fully as it only focuses on the "temporal segmentation." Overall this figure contributes little to the discussion of the section.
The discussion in Section 2.2 contains details that are essential to the rest of the manuscript, so it is important that readers understand this section well for the paper to make sense. It is recommended that a figure be inserted to serve as a guide to readers in understanding this lengthy section. It is also recommended that the discussion flow of this section be improved. The relationship of cognition and accessibility to space segmentation must also be improved as this is the basis of how segments are compartmentalized in the contained sections. What does "renovation" in Line 188 have to do with the rest of the section's discussions? There are too many statements like this, which may be related to the component being discussed, but have little to no relation to the paper's topic. It may help the brevity and conciseness of the paper if these statements are eliminated.
In Section 2.3, please improve the presentation (headings, statements) in Table 1 or, better yet, use a visual figure to illustrate your point. This section also contains some content found in Yan et al. (2019) * https://doi.org/10.1111/tgis.12574, especially Figure 3. Statements starting Line 296 must also be illustrated through Figures.
Section 3 begins with introducing IFC, but the earlier sections only discussed CityGML and IndoorGML. Some discussions about the IFC standard must be introduced in the early sections. Lines 350-351 talk about the limitations of IFC in describing indoor space, but the next statements do not provide a solution or a closure for this. The table in Line 361 is erroneously numbered. What is the purpose of illustrating the EXPRESS definition of IFCProxy entity (and another one in the succeeding pages) in the context of the paper? Similarly, Table 3 contains indoor relationships that do not have corresponding IFC relationships, but the succeeding discussions are silent about this. The section ends with the proposed model, which is the most important part and contribution of the paper, still unclear to the reader. As this section contains definitions in the form of enumerations, the section must be summarized and synthesized, and additional value-adding visuals or tables must be added to help readers' understanding.
The section on the case study must discuss the procedures first before presenting the results. Furthermore, the experiment and results presented do not utilize the proposed model extensively, and its role in the extraction of the results is unclear. The discussion in the succeeding section does not correspond to the contents of the case study. It also contains content that compares existing standards with each other, as well as some limitations, which must be discussed in much earlier sections of the paper in order to establish the gap in the body of literature that the paper is addressing.
The paper also contains uncited content from https://doi.org/10.3390/ijgi7080321 and https://doi.org/10.3390/ijgi6110343. Please make sure to cite references both in-text and in the bibliography section.
The paper lacks organization and structure. Numerous grammar errors (beginning from the title), punctuation errors, and spelling errors are present makes the manuscript confusing and difficult to read. Abbreviations must be defined during the first mention. Statements bearing different ideas must be separated to improve the readability and organization of the paper.
Reviewer 3 Report
Dear Authors,
Please find the attached file for my concerns.
Best Regards
Comments for author File: Comments.pdf
Reviewer 4 Report
This paper presents the authors’ work on developing a classification system for indoor space. In this system, the indoor space is divided into five levels, including building space, floor space, household space, room space, and functional subspace. Overall, this paper does not have a strong motivation, and its contribution is not clear. The detailed comments are as follows.
(1) Even though the authors came up with a classification system for indoor space, but the question is why do we need it and what for? In terms of ‘why’, the authors did not give a very strong motivation for conducting this research. The authors stated that ‘the spatial organization of buildings, floors, and rooms can no longer meet the needs of refined indoor space analysis’, this seems to be the motivation of this study, but the authors did not further explain that statement, i.e., what kind of indoor space analysis and what they need. In terms of ‘what for’, the authors only proposed a classification system but did not demonstrate how this classification is different from existing ones and the practical implications imposed by this ‘difference’. For example, this study divided the bedroom into four basic functional areas: sleep, dressing, storage, and audiovisual, but did not explain what kind of applications would need this classification. Without properly answering these two questions, the motivation and contribution of this study cannot be justified.
(2) The use of English should be improved. The overall usage of English is fine, but the use of terminologies is terrible. It seems that the authors have made a lot of terms, such as ‘split-level space’, which do not have a clear meaning. Also, many grammar errors are noticed in this manuscript, e.g., line 354: The mapping relationship between spatial objects in indoor scenes and IFC spatial structure units, which is not even a sentence. Not to mention that many parts of this manuscript appear to be translated by machine, which is difficult to understand. The language must be improved before this paper can be further examined.
(3) This study extended space by using ‘IfcProxy’. In fact, the type of ‘space’ can be defined in the ‘PredefinedType’ attribute of ‘IfcSpace’. Apart from the predefined ‘SPACE’, ‘PARKING’, ‘GFA’, ‘INTERNAL’ and ‘EXTERNAL’, user can define their own type of space.
(4) Section 4 ‘Case Study’ is more like a demonstration rather than validation. The authors should validate the ‘usefulness’ of this classification system, rather than ‘feasibility’.
(5) Specify the IFC version. There are currently two official IFC versions, IFC2x3 and IFC4. They are quite different. The authors should specify the IFC version referenced in this study.
(6) In IFC, the spatial structure element includes IfcSite, IfcBuilding, IfcBuildingStorey, and IfcSpace, which present the spatial structure of a construction project, while IfcSite was omitted from this study. You may refer to ‘Integration of BIM and GIS: Geometry from IFC to shapefile using open-source technology’ for more information, or https://standards.buildingsmart.org/IFC/RELEASE/IFC4/ADD2_TC1/HTML/schema/ifcproductextension/lexical/ifcspatialstructureelement.htm for more information about the spatial structure of IFC.
(8) Figure 7 and Figure 8: the relationship between entities should be specified in a clearer way. To be specific, the first information that would be obtained from Figure 7 is that ‘Sofa contains living room’, which is obviously not logical. UML can be used for clearly describing the relationships between entities.
(9) The ‘extended’ attributes/properties overlap with existing ones. Many of these ‘extended’ properties, such as lighting, humidity, temperature, have already been included in the various property sets defined by IFC for IfcSpace.
