Versioning of 3D City Models for Municipality Applications: Needs, Obstacles and Recommendations

Round 1

Reviewer 1 Report

This paper reviews the current requirements and challenges of enabling a version management system for 3D city models in the use case of Sweden. These are divided into four layers: data collection, building theme, city model and application. Each layer requires some changes and restructuring of the employed tools, but the common denominator is to enrich and handle temporal information properly. Based on their findings, the authors then proceed to recommend linking the 3D city models to the national building registers to enable more complex use in e.g. building permits, 3D cadastre, facility management as well as in municipal base maps.

The manuscript reads well. It shows that the authors have done extensive research in the state-of-the-art literature and software. The paper also offers an interesting insight into the current situation in some Swedish municipalities regarding the development of a version management system for 3D city models.

General comments:

• You mention in line 34 that city models “are often recreated instead of updated” (and should be avoided). But then in line 38, you write “it is important to save and archive the documents” (i.e. “snapshots” should be created). These might contradict each other, but of course they do not, as they represent two different use cases, one is change detection and update of only real changes on top-level features and sub-objects of top-level features, one is versioning of top-level changes only. I assume only the latter is discussed in your paper? I suggest reformulating these sentences to make the use case clearer and avoid confusion to the readers.
• I believe CityJSON is currently not a standard (see line 70, 489, 579 as well as Table 3 City model layer).
• Can the versioning information regarding the fifth issue mentioned in line 85 – 86 be propagated in both directions between layers without information loss (i.e. from data collection to application layer and vice versa)?
• You write in line 95 – 96 that “the conditions are similar in many countries, which makes the case study general”. Do you have a source for this? Because even within Europe, some countries are vastly different from the others regarding the infrastructure, digitalization, choice of hardware and software, etc. In addition, as mentioned in line 290 – 291, you only conducted interviews with a limited number of municipalities and half of them responded. I would therefore say that your findings are very interesting for countries that are somewhat similar in specific aspects, but not generalize the study for all countries.
• I do not think Git is popular mainly for programming only (see line 184), but also for work coordination and project management, etc.
• What do you mean by “some unclear definitions” in line 193? Change detection in CityGML is difficult due to many reasons, some of which are: CityGML covers a wide range of top-level features defined in a complex hierarchical structure, it combines both graphical and semantic aspect of city objects into one place, and syntactic and geometric ambiguities exist along with five Levels of Details (LOD 0 – 4). These ambiguities are neither the result nor the cause of “unclear definitions”, they allow users and surveyors “much” flexibility and freedom while defining objects but also at the same time must all be considered by developers (which is a challenge).
• Regarding the data provider incentives, have you also considered some other causes that might prevent a municipality from implementing a version management system, such as for security and data privacy reasons (city – city, city – private company, city – public users), monetization (city – private company), competition between municipalities or software vendors, or technical competence of municipality staff?
• Similarly, you recommend linking 3D city models to building registers. Are there any issues regarding data privacy, security, or conflict of interests between municipalities and private companies as data providers?
• How relevant is the open-source 3D City Database (3DCityDB) regarding the application tools for 3D city models in section 3.8?
• Have you tested your recommendations against real-world data, or developed a prototype? From what I understand the NS building is used exclusively in Sweden? So your recommendations can only be applied to cities in Sweden only? What should be done so that your results can also be applied in other countries as well?

 

The following lines contain some typos and/or need to be reworded (not exhaustive list):

30           and the number is [1, 2] -> make this sentence more clear

31           later -> latter

32           comma

34           comma

40           updates  -> updating

50           makes it possibility -> makes it possible

56           linked. [6, 7] -> linked [6, 7]

69           building permit process etc. -> building permit process, etc.

96           make -> makes

106         both these -> both of these

152         varies?

164        included -> included in

193         sematic -> semantic

196         neo4j , -> neo4j, (without space)

235         3DCityDb -> 3DCityDB

276         data providers -> data providers’

284         information,11 -> information, 11 (with space)

297         if and in that case?

440         inherits form -> inherits from

597         enable that?

 

Further comments:

• Your references/citations are sometimes used in plural and sometimes in singular form. Please use a consistent form.
• Section 2.3.5 should be renamed to 2.3.4.

 

In general, I find this manuscript well-written, well-structured and informative. It does not provide any implementation but it can partly help future developers implement a version management system for 3D city models, especially in countries and cities that are similar to the ones surveyed. I would suggest accepting this paper with minor revision.

Author Response

Please see the attachment

Author Response File: Author Response.docx

Reviewer 2 Report

This article discusses the subject of versioning of 3D city models and tries to identify the challenges that will be undertaken when such a solution is going to be implemented at a national level. The authors use Sweden's initiative to integrate information of buildings between different parties so that the building permit process is being followed by all stakeholders. They identify six main challenges that need to be undertaken and try to evaluate the issues that will arise when such an initiative will be implemented.
The structure of the article is good and the text quality is decent. Certain parts of English can be further improved, though (see minor suggestions below) as there are some issues with syntax occasionally. The research conducted is rather interesting and, in particular, the proposed formulation of six issues that need to be addressed in order to tackle the problem of versioning of 3D city models between organizations is a going to be a very useful insight to others in this field. Also, the recommendations can be particularly useful in the context of building permit management at the national level.
However, there is a major points that needs to be addressed. The article's scope seems to be more specific than the introduction and motivation suggest. While the title and introduction imply that this research undertakes the problem of versioning of 3D city models in general, the actual work conducted is focused too much on the particularities of the specific use case. For instance, the authors only discus issues that are related to buildings and they focus on how the problem occurs when the specific Swedish organizations are involved in the NS building specification. However, this is too specific and the authors do not clarify how the findings of this research can generalise to apply to others. Not all 3D city models are about monitoring building permits in a national level. So, the authors will either need to restructure the title, introduction and motivation so that it is clear that this is mainly tackling the problem of versioning of 3D city models in the context of building permit monitoring, taking the example of Sweden; or they will have to revise the conclusions so that they can provide enough insight on how the findings of the specific use case can be applicable in other use cases.
For this reason, I suggest that the paper is majorly revised before it gets accepted.
Other smaller issues:
- In line 40 you state some conclusions. In my opinion, this is too early to conclude things and a reader will be confused as to how you resulted there. - In section 2.3.3 you are missing the *author* property of a version. - In section 3.5.1, it would be more clear if you used the term "schema" instead of "structure", as this is the only thing that is common between a specification and a database structure. - The Conclusions are presented in a bulky way that is hard to follow. It would be easier for the reader to follow your conclusions if they were divided into paragraphs, ideally linked with the previous discussions or the recommendations that preceded this chapter. - In line 30: the sentence ends with "[...] the number is [1, 2]." Are you missing a verb there? - In line 125: "[...] that require*s* temporal information [...]". - In line 140: "[...] where *a* geographic phenomenon [...]". - In line 164: "[...] included *in* CityGML 2.0 [...]". - In line 173: "[...] which objects that belong [...]", I suggest that you remove the word "that". - In line 219, the sentence "Still, a majority of openly available 3D city models are never changed, or recreated instead of updated" is a bit confusing. Could you, please, rephrase it? - In line 288: "[...] international geodata standards *were* investigated".

 

Author Response

Please see the attachment.

Author Response File: Author Response.docx

Reviewer 3 Report

This paper proposes to make a survey of the versioning capabilities in both theoretical and applicated manner. It gathers information and stories from swedish administrations and provide a brief summary of how versioning can be handled.

We do not catch the contribution of the paper from a research viewpoint. Even if an inventory of what is currently achieved in Sweden can be easily generalized to most of the developed nations, it does not provide any advance or concrete proposal.

Before moving on to some more specific examples, we would like to point out that the paper needs to be deeply proofread before resubmitting it: many subject-verb agreement errors, some words are missing (ex. The first line of the intro is missing a verb), use of references should be harmonized, ... The recommendations and conclusions state general truths without any fruitful discussions.

1. The "Building theme layer" notion is difficult to catch. If the figure 1 is at the center of the paper, please explain it more.
2. Section 2.3: If the previous paragraph already speak about the reference 10, why developing it later? This paragraph is punctuated by errors that devalue the quality of the section. Is it really necessary?
3. Line 160: does the reference 18 really state such things?
4. 2.3.1: It is a list of attributes. Where is the added value?
5. 2.3.2: CityGML is a Data Model. INSPIRE is an european directive which does not provide any real formalism. Please ensure the relation between the two and their potential dependencies (Is CityGML related to INSPIRE, is it the inverse, are they really speaking about the same things, are their creation purpose evet the same?).
6. 2.4 : ok. So what? It is already clear from the introduction and does not provide any added value.
7. 3.3 same remark as point 6.
8. Sentence starting at 219 is difficult to write. Some can answer you to search more. It deserves you.
9. Table 1 is mainly empty. Why using a table? Aren't there any other presentation mode to highlight your purpose?
10. Fig 4: elements are not comparable or linkable. Can you please express links or modify the elements representation? UML perhaps?
11. Table 2: Around the terms "validTo", "validFrom". We suggest you to harmonize your "NS" with the official standards.
12. Section 3.8: these tools are not even created to handle 3D natively. So you cannot expect them to handle subcapabilities of 3D.
13. CityJSON is not a standard. Moreover, you wrote that the reference 10 provide an implementation for CityJSON. There is no clue of such an implementation in the reference or drafts.
14. Figure 1 is not an architectural model but a schematic diagram that needs to be reworked.

Moreover, it seems that this paper takes the reference 10 from the same authors as a reference of major importance. Adding administation stories from it is not an added value.

We realise that our remarks could be seen as nasty but it is necessary to rethink the added value of this paper in depth. We encourage the authors to look at the evolution of the notion of identity in GIScience to improve their insight of what versioning is and what it is based on.

Author Response

Please see the attachment

Author Response File: Author Response.docx

Reviewer 4 Report

This paper performs a systematic investigation of performing versioning on the 3D city models used in national building registers. Specifically, the paper motivates the needs of utilising 3D versioning in certain situations, identified the difficulties of implementations, and provided recommendations on addressing the versioning problem. The paper had reviewed existing usage of version control tools in other related areas, provided the architectural model for managing 3D city model information, and proposed ways to incorporate the version control tools into existing management framework.

The motivation is grounded, and the approach used to evaluated are also scientifically appropriate. The problem formulation is mainly based on the current city registry situation in Sweden, but most of the approach that the paper described can be generalised to other similar cities management. The authors had identified 6 main issues associated to be the potential obstacles for 3D city models versioning. All of the issues had been well elaborated and the recommendations provided at the discussion of the paper was able to provide an insight on how to tackle those associated issues. Surveys had been used to identify some of the issues with the current situation of 3D city model versioning in municipalities, which is able to provide an empirical evaluation of the potential difficulty. Overall the case study for the six issues was able to answer a lot of the potential concerns when one wants to implement 3D model versioning under the existing national building register framework.

Overall the methodology, evaluation and the recommendations provided are well written. However, there are parts in the paper where some sentences don't flow well, contain grammar mistakes, or should be stated more clearly. The authors should go through the text and proofread the manuscript.

For example, Line30: "...and the number is [1,2]" seems to be missing some part of the sentence. Line56: citations should be followed with the previous sentence. Line125, 136, etc.: A sentence should not begin with a numeric citation. Rather than "[3] propose", use the author's name as a subject. Line121: Avoid beginning a sentence with a numeric number. Line252: sentence structure is weirdly constructed and unclear. And most of all: Line 414-416: "This section may be divided by subheadings. It should provide......that can be drawn".

The listed lines are not exhaustive by any means, so a proofread and fix on related issues should be performed on the entire document.

Author Response

Please see the attachment.

Author Response File: Author Response.docx

Reviewer 5 Report

Section 2.3. This section containing too many detailed information of other versioning methods, but lack of compar comparative analysis of advantages and disadvantages.

Page 9, Table 1. The table shows the additional activities and version information to the new national platform. the attributes only consider the building as 2D object, lake of 3D detail information, such as floors number, population capacity, building space, room capacity. Maybe you could reserve some area for future improvement.

Page 11, line 363. this paragraph could be merged with the next paragraph.

Page 11, line 365. Give a suggestion expensive and time cost for this work according as a reference.

Page 18, Second paragraph is a too long paragraph, simplify the paragraph to let your opinion more distinct.

Author Response

Please see the attachment.

Author Response File: Author Response.docx

Reviewer 6 Report

In this work, the authors aim to search for the causes and obstacles that have not allowed a continuous updating of 3D digital Cities versions, in the Swedish context but clearly extendable to a general level. Through the discussion of the current status of six issues, the authors propose an analysis of the actual model version management and what should be implemented in order to recognize/fix an updated version of the digital structure and links between the various versions. The structuring of a version management model for 3D Digital Cities would allow to magnify the potential and to return highly useful tools for professionals and simple users.

In order to improve the presentation, I recommend carefully reviewing the English language version.

The work does not present heavy revisions in conceptualization and structuring but I report some minor revisions in order to be fully accepted:

An improvement of the Introduction is recommended. It is too dependent on the following paragraphs, not giving importance to the topic addressed. The consecutive repetition of concepts and terms reduces the readability of the work.

Line [30]: unclear the expression "[...], and the number is [1,2]". Explicate the content of the Reference or explain what is meant.

From Line [35] to [49]: It is suggested to support the statements with appropriate literature to strengthen the conceptualization of the work.

Line [56]: Remove point in "[...] be linked. [6,7].". Check other writing errors.

From Line [56]: In all those cases where the Reference is reported at the end of a sentence, my suggestion is to report at least one name of the author, for example "Authorname et al. [11] describes...". This can improve the readability of the text.

From Line [205]: Check the formatting in the text of the figure references (capitalize them).

Line [209]: Format the literature reference better.

Line [247]: Remove brackets to the year 2020.

Line [248]: Remove repetitions (see "section").

Author Response

Please see the attachment.

Author Response File: Author Response.docx