Clustering of European Smart Cities to Understand the Cities’ Sustainability Strategies

Comments from Reviewer #1

Responses

-       The given abstract does not reflect with MDPI format. I strongly suggest you refer to MDPI research paper and rewrite the abstract without any citation.

-       We rewrote the abstract following MDPI format. Please see lines 11 to 22 of the manuscript.

-       How about the conclusion of the research

-       We included a section “Conclusions” to strengthen the results. Please see lines 497 to 517 of the manuscript.

-       The list of references also does not follow the MDPI format.

-       We reviewed the references following MDPI format with the bibliography software Zotero.

-       When examining the above points, it seems some major points are remaining to be fixed before the next level.  If this is your first manuscript with MDPI, better to follow the author's guideline.

-       We restructured the manuscript following MDPI format and author's guideline.

Comments from Reviewer #2

Responses

-      As an introduction, the reader learns that Angelidou considers that a technocentric vision cannot reflect the reality of a smart city. Immediately afterwards, the authors explain that “to enhance the understanding of a smart city, knowledge management seems to be a relevant framework which is being employed in academic studies.” This is a weak explanation of a possible gap in the literature. The authors then describe the operational and core capabilities. While the intention is laudable, we would need more information to understand why it is interesting and necessary to do this analysis. In other words, what need does this article address? More precisely, which theoretical needs and actually existing needs? Although the journal does not require a literature review, the introduction could have referred to the literature to further justify the relevance of this study, especially since Angelidou’s work has given rise to a great deal of critical research on this technocentric vision of the smart city. At the end of the introduction, for example, the reader knows neither why it would be interesting to “question the smart strategies of cities” nor why it would be interesting or even necessary to discuss the results in light of the work of Dameri and Ricciardi.

-      The introduction has been rewritten to better highlight this issue. Please see lines 25 to 84 of the manuscript.

I therefore recommend that the authors respond to the following questions:

-       What theoretical problem does this article respond to?

-      Academic studies are often centered on one or two cities and on part of the seven dimensions of smart cities (e.g., Anthopoulos, 2019). Even if these analyses are essential to understand the smart city approach, they do not enable to understand an overall vision of the smart city strategy. We consider dynamic capabilities approach to develop an empirical analysis for a sample 40 cities to highlight the seven smart city dimensions and how these strategies make cities sustainable.

-       What actual problem in reality does this article respond to?

-      This article analyzes the complementarity of the different dimensions of smart cities. The study provides a global framework for implementing sustainable smart policies.

-       What is the scientific and social relevance of this article?

-       From a scientific point of view, the relevance of our article emphasizes two points. The first one is to enrich the understanding of smart cities by showing the interest of considering all the seven dimensions and not focusing on a few. The second point lies in the application of knowledge management, by showing that this is a relevant framework to understand smart cities.

The main social relevance is to provide a global framework for stakeholders (authorities, decision makers, academia, etc.) on the implementation of sustainable smart city strategy.

-       What are the contributions of this article?

-       The main contribution of our article is to identify the three smart city clusters of European cities: cities with emerging smart strategies, technology-oriented cities, and quality-life oriented smart cities. These strategies are steps to include in the smart city approach and its implementation a realignment towards sustainable smart city strategies. Thus, we also show that it is relevant to address smart cities in plural, depending on the context, and not as a generic smart city approach.

The second part (from smart cities themes to core capabilities) constitutes, in my opinion, more of a methodology whereas it is presented as a kind of a literature review. Since the proposed analysis is quantitative, I think that a more fleshed out methodology section would be necessary, and I would eliminate this part. To consider it as a methodology is all the more necessary since the authors evoke a whole series of choices in this section—without this being adequately justified elsewhere at all times. The table in the appendix gives us all the information we need. Hence, why make a separate section of it? Or, why not incorporate it into the appendix? I therefore recommend that the authors address the following questions:

We included the reviewer request, and we modified our introduction and our second part. All aspects relating to the description of the smart dimensions are in part 2.

-       How is the methodology innovative?

-       This methodology innovates by the classification and identification of homogeneous clusters of smart cities. We therefore develop a new analysis of sustainable smart cities in Europe.

The classic methodologies for analyzing smart cities have as objective the production of rankings of smart cities. For example, the article of Akande et al. (2019) uses a similar methodology, based on a PCA and a HCA, but their objective is a ranking of smart cities. They look forward to reinforcing competitiveness.

-       What are the limits of this methodology?

-       The limits of our methodology were added to the discussion section of the article. Please see lines 473 to 485.

-       Finally, the last section is the most interesting; however, the results are not sufficiently highlighted. One would expect the results to be discussed in more breadth and detail. You infer—too quickly—that the formation of a cluster of sustainable cities is based on the assumption that “sustainable equals fight against pollution.” The political context is in no way presented or discussed. There is no criticism. The postulated difference between some European cities and global cities could be further exploited. The authors do not indicate the limits of their approach. It is likewise odd that there has been no analysis of smart city policies or strategies.

-       We rewrote the discussion considering the remarks of the reviewer. We underlined that smart cities are sustainable in the sense of knowledge management. They are organizations that adapt to the evolution of their environment. Thus, all three clusters of cities are considered sustainable, but not in the same way.

To avoid confusion, we have changed the label of cluster 2, it is no longer specifically sustainable cities, but cities that have a quality-of-life oriented strategy. These cities consider environment as a main issue of their strategies. We have also reinforced the discussion around clusters 2 and 3 and the results of Perboli and Rosano (2020).

Comments from Reviewer #3

Responses

-       Title: “Mapping of European smart cities to understand the cities' sustainability” is not enough developed in the paper. The Authors should consider the relationships between: the title – the research goal – the research process and – finally - conclusions. I think, that the research goal was set too ambitiously.

-       We have taken into account the reviewer's remark and changed the title of our article, it is now “Clustering of European smart cities to understand the cities’ sustainability strategies.”

-       Literature analysis – is treated to a large extent as an introduction to the smart cities problem. It is correct, but – on the other site - not sufficient in the context of the provisions in points: 2. Materials and Methods and 3. Results.

-       We entirely rewrote the introduction and made a focus on the presentation of the scientific and social relevance of our work. Please see lines 26 to 84 of the manuscript. We also reorganized sections 2 and 3: the description of the different dimensions has been replaced in part 2, which is pertinent in the context of our methodological approach.

-       Point 2. Materials and Methods - does not get acquainted the reader sufficiently with the implemented methodology. It is too general, it lacks the characteristics of significant steps in research activities carried out by the Authors. In many places they refer to the research methods used by other authors. The reader should be given the opportunity to identify and understand the most important steps in the research that is undertaken in this article.

-       To clarify the methodology, we have reorganized the section 2 by defining three steps: the definition of smart dimensions, the interpretation of core capabilities through a PCA, and the identification of clusters with a HAC.

The section 3 has also been restructured around these three stages with the creation of a section and subsection specifically related to each smart dimension. The Figure 1 “Methodology for clustering sustainable smart cities” was modified in consequence of this change.

-       There is no reliable input data characterization. The role of these data in research procedure should also be discussed. In this form (without individual characteristics of the features) entries 239-334 are not acceptable (please notice, that I do not question their actual value, computational quality - because these can be really good).

-       Following the request, we restructure section 3. The article develops an analysis of smart city strategies through a clustering. Seven smart dimensions for 40 cities in Europe are characterized by 79 variables to define operational capabilities. These variables contribute by a PCA to identify core capability at the city level. We interpret these core capabilities and generate a clustering by a HCA. As a result, a classification has revealed homogeneous clusters and three kinds of sustainable smart strategies. We included on the file Data.xlsx a sheet describing every variable of the analysis presenting: sources, estimations, and extraction information. Please see lines 257 to 367 of the manuscript and the supplementary material.

-       Supplementary material - is a compilation of various research results and various figures. The question is, as a result of which operations did they arise? On the other hand, the "Data" file contains many numerical values ​​(these are only numbers without mathematical formulas) - I want to know how they were obtained (at least a general overview of the research operations carried out is necessary). The main text should convince the reader of its correctness and research quality. The auxiliary files can only supplement it.

-       We rewrote section 3 presenting results and include as supplementary material in file Data.xlsx a new sheet namely “data_characterisation” describing each variable used on the methodology. The figures and tables on the supplementary material are cited on the manuscript.

-       I recommend that the main text should be clearly combined with the results of the research.

-       We have rewritten the discussion section to emphasize the relevance of our results and relation with previous studies on smart cities. Please see section 4, lines 368 to 496 of the manuscript.

-       The authors did not convince me of the legitimacy of choosing these 40 specific cities. When formulating this observation, I still remember reading about a few selected cities in the "discussion" point 4. Maybe it is worth focusing on urban spaces that are less "researched" so far?

-       We integrated this point on the limits of the methodology, specifically on the data collection. Please see section 4 “Discussion” from lines 476 to 485 of the manuscript.

-       Moreover, with such extensive data and research, I would expect suitably advanced conclusions and recommendations.

-       To respond to this point, we added a conclusion section which allows us to reinforce our main results and show the research perspectives. Please see lines 486 to 496 for the future research guides and section 5 “Conclusions” from lines 497 to 517.

Comments from Reviewer #4

Responses

-       Perhaps, the authors could clearly highlight the importance of clustering and classifying smart city projects as they propose. As they state in the paper, there are many classifications and papers proposing different dimensions of what to consider a smart city. So, why do we need another classification? Emphasising this would also put more emphasis on the results of the paper. 

-       We have entirely rewritten the abstract to better present the problematic subject of our work. The aim is to provide an analysis of smart city strategies through a classification. We include the seven smart dimensions for all the cities in our sample and examine their complementarity. This constitutes a contribution to the existing literature on two points. First, we enrich the analyze presenting an overview methodology, instead to only focus on one or two dimensions for a specific city. Second, we develop a classification that reveals homogeneous clusters to point out how these strategies make cities sustainable. It is not a smart city ranking.

-       In line 77, authors say that they focus on internet access to all individuals and businesses at the national level to illustrate this dimension (smart architecture and technology). As they mention, this dimension includes many more issues. Just as an example, something very representative of a smart city is the deployment of sensor and actuator networks on the streets to gather city data that could be used in real time or exploited in non-real time to improve certain city operatives. This data-driven management approach where data is directly gathered from the streets is, perhaps, one of the most recognizable traits of this dimension. Therefore, the authors may want to justify how just focusing on internet access they can provide an evaluation of this dimension, which, intuitively, seems to be much broader.  

-       We took into account this remark. We have modified the presentation of the smart architecture and technology dimension by adding the following elements, please see lines 145 to 157.

-       Authors have explained all the dimensions separately. This gives the feeling that all the dimensions have the same importance in the subsequent clusterization. Perhaps, the authors could clarify how the performed analysis takes into account the importance of each dimension or the fact that some cities may score really high in just one dimension and really badly in other, and other cases as such.

-       We do not highlight the importance of one dimension over another in the definition of smart dimensions because the objective of the study is to understand the relative importance of each dimension in the implementation of smart city strategies. Through our three clusters, we clearly show that a smart city does not implement an extensive policy on all 7 dimensions but on certain of them. To clarify this point, we have rewritten the abstract, the introduction and the conclusions of the article to better expose the purpose of our approach.

Comments from Reviewer #1

Responses

-     The manuscript has been improved. English language and style are fine/minor spell check required.

-       We improved the English language and style.

-     Are the methods adequately described?

Can be improved.

-     We updated the section 2 “Materials and Methods” and expanded the methodology. We also improved the English language and style.

-     Are the results clearly presented?

Can be improved.

-     We updated the section 3 “Results” and every subsection. We expanded the discussion of our results in section 4.

We also improved the English language and style.

Comments from Reviewer #2

Responses

-     Does the introduction provide sufficient background and include all relevant references?

Can be improved.

-     We expanded the introduction, in particular the references.

We also improved the English language and style.

-     Are the results clearly presented?

Can be improved.

-     We improved the section 3 “Results” and every subsection.

We also included more details in the discussion of our results in section 4.

We improved the English language and style.

-     The authors have implemented every suggestion made by reviewers. The paper has improved a lot. From my point of view, two elements are still somewhat less convincing: the literature review and the discussion.

The reader is left with the impression that the contribution of the article is “only” methodological. I find it is missing something and that it is hence not offering convincing conclusions. A thorough review of the literature would fill this gap. I encourage them to revise their literature review.

-     We updated the introduction, the discussion and the conclusion presenting more details of the references.

-     Our research has two goals: implementing a method for the assessment of smart cities strategies and understanding the sustainability of smart cities using a clustering method.

Please see lines 67 to 80 in the introduction.

-     We rewrote the first paragraph of the discussion to emphasize the results.

-     We rewrote the second paragraph of the conclusion to highlight the results. Please see lines 498 to 508.

Comments from Reviewer #3

Responses

-       Does the introduction provide sufficient background and include all relevant references?

Can be improved.

-     We expanded the introduction, in particular the references.

We improved the English language and style.

-       Is the research design appropriate?

Can be improved.

-       Our research has two goals: implementing a method for the assessment of smart cities strategies and understanding the sustainability of smart cities using a clustering method.

Please see lines 67 to 80 at the introduction.

We rewrote the first paragraph of the discussion to emphasize the results.

We also improved the English language and style.

-       Are the methods adequately described?

Can be improved.

-     We improved the section 2 “Materials and Methods” and added more details to the methodology.

We also improved the English language and style.

-       Are the results clearly presented?

Can be improved.

-       We updated the section 3 “Results” and every subsection. We also expanded our results in section 4.

We also improved the English language and style.

-       Are the conclusions supported by the results?

Can be improved.

-       We rewrote the second paragraph of the conclusions to highlight the results. Please see lines 498 to 508.

We also improved the English language and style.