Criteria for Smart City Identification: A Systematic Literature Review

Abstract

The transition towards greater smartness is an emerging trend in the development of modern cities. This transition manifests itself in the widespread adoption of information and communication technologies (ICTs), cloud computing, Internet of Things (IoT), and other technological tools aimed at improving the level of city smartness. Although numerous studies have focused on the smart city (SC) phenomenon, knowledge about empirical criteria that can be used to define a city as “smart” and to measure the degree of a city’s “smartness” remains limited. The present study aims to bridge this knowledge gap by a systematic literature review of recent studies, in which various empirical criteria are used for SC identification. The study helps to identify a total of 48 SC identification metrics, which are further split into three main categories—smart digital technology, living conditions, and environmental (ecological) sustainability. Among these groups of criteria, the “smart digital technology” group of metrics appears to be the most popular, while criteria pertinent to “ecological sustainability” are applied considerably less often. As the study also reveals, only about half of the criteria used by empirical studies for SC identification actually relate to urban residents’ needs, with the rest being general technological measures. Therefore, for a balanced SC assessment, we suggest a ranking system based on the nine most important metrics, which equally represent all the main aspects of the SC phenomenon while placing an emphasis on the improvement of the quality of life of local residents. The proposed system is applied to several major cities across the globe to demonstrate its use and usefulness.

1. Introduction

In 2018, there were some 7.6 billion people in the world, of which 4.2 billion (or ~55%) lived in urban areas. By 2050, the global population is projected to increase to around 9.8 billion, of which some 6.7 billion will be living in cities [1].

According to the UN estimates [2], by 2030, the number of megacities with more than 10 million residents is expected to exceed 40. It is also expected that such megacities will consume 81% of the world’s resources, while by 2050, energy demand might rise to about 620 exajoules (EJ) [3].

Although the rate of urbanization in Europe has been slowing in recent years, ~72% of the continent’s population currently lives in urban areas, and in some countries, such as the Netherlands, the rate of urbanization already exceeds 90% [4].

As urbanization progresses, many cities accumulate problems, such as environmental degradation [5,6], deteriorating infrastructure [7], poverty [8], and societal inequality [6,8]. Solving these problems directly affects the level of well-being of citizens, the environmental situation, and the sustainable development of the city as a whole.

According to the UN, cities play a key role in promoting sustainable development, and the UN Sustainable Development Goals (SDGs) explicitly refer to Sustainable Cities and Communities as the 11th SDG, while the UN Agenda 2030 emphasizes an objective of making cities more inclusive, safe, resilient, and sustainable so as to improve people’s lives [9].

In order to address urban issues, the smart city concept has emerged as one of the possible solutions.

On the positive side, many cities, especially in high-resource countries, have begun the process of transition towards greater smartness. This process manifests itself, inter alia, in a widespread adoption of information and communication technologies (ICTs) and a rapid spread of “big data”, cloud computing, and Internet of Things (IoT) as ways of modernizing and improving urban performance and services [10,11,12].

Although the concept of smart cities (SCs) is still evolving, it has gained considerable interest among researchers [12,13] as well as among ordinary web users (Figure 1).

As is well documented, new technological tools associated with SC transition help to monitor air pollution [14,15], optimize motor traffic [16], increase resource use efficiency [17], stimulate citizens’ participation in urban affairs [18,19], and improve human welfare [20].

Previous studies of the SC phenomena looked into the nature and origin of SCs [21,22], quality of life in SCs [23], and ethical aspects of using artificial intelligence and big data technologies [24]. Yet only a handful of studies have attempted to determine empirical criteria that can help to identify SCs and gauge their progress towards greater smartness.

As of today, eight literature surveys have been carried out on the SC topic [25,26,27,28,29,30,31,32]. Thus, in an early survey on the topic, Lombardi et al. [25] identified five main groups of components that characterize a city as smart: citizen participation (smart governance), people’s capabilities (smart human capital), efficiency of resource use (smart environment), quality of life (smart lifestyle), and competitiveness (smart economy).

In a separate study, Anthopoulos [28] suggests that a modern SC should include the following attributes—resources, transportation, urban infrastructure, living, government, economy, and coherency—all of which address the digital divide, social relations, and ICT connectivity. This classification of SC identification criteria corresponds to that suggested by Tregua et al. [26], who define the following three-tier structure of SCs: environmental component (smart environment and smart mobility), social component (smart people, smart lifestyle, and smart management), and economic component (smart economy).

In another literature survey, Albino et al. [27] examine the meaning of the word “smart” in the context of cities, emphasizing the following four most common SC characteristics: efficient network infrastructure, business-led urban development and creativity that fosters urban growth, social integration, and environmental preservation. This definition corresponds to that suggested by Silva et al. [29], who define four main attributes that create an SC: resilience (infrastructure and management, pollution and waste, energy and climate change, social issues, economy, and health), quality of life (emotional and financial well-being), level of urbanization (technology, economy, infrastructure, and governance), and intelligence (social, environmental, and economic indicators).

According to another recent study by Marchetti et al. [30], an SC is a place where government transparency is a common asset, where people are policy-driven and interested in participating in decision making, where shared cultural and recreational infrastructure is offered, where everyday needs are supported by urban amenities, and where a high level of comfort is provided. In this conceptual framework, an SC helps to reduce inequality and increase social, territorial, and economic integration by building effective relationships between a city’s assets, services, and social environment, citizens’ participation in urban affairs, and governmental transparency. These SC components are similar to those identified by Lim et al. [31]: ICT infrastructure and human, social, and institutional capital. In another recent survey, however, Li et al. [32] single out a different set of SC identification criteria, viz.: technological innovation, a smart economy, smart infrastructure, smart services, smart mobility, and a smart environment.

There are several international rankings that define criteria for classifying a city as smart, such as, e.g., the Italian ICity [33], the IMD Smart Cities Index [34], IESE City in Motion Index [35], CITYkeys [36], the European Smart Cities Ranking [37], and others. Yet these classification systems differ regarding how city smartness is assessed and measured. Such classifications are also not always balanced in terms of the social, economic, and environmental components that are key dimensions of sustainability. In particular, while one group of studies place an emphasis on ICT and infrastructure [21,27,30,31,32], other studies emphasize the social performance of cities and their physical environment [25,26,28].

These differences can be attributed to the fact that the SC phenomenon crosses different scientific fields, and studies covering specific research fields can lead to different results. Therefore, a holistic approach, based on an interdisciplinary review of the relevant literature, is needed to identify the full range of attributes that underline the existence and functioning of SCs.

The main goal of this study is to systematically review the recent literature on the SC phenomenon so as to develop a comprehensive and balanced set of empirical criteria that can help us to distinguish between SCs and the remaining urban localities and assess a city’s progress towards greater smartness.

To achieve this goal, the present study employs the Preferred Reporting Items for Systematic Reviews (PRISMA) tools. In contrast to traditional survey methods, such as content analysis [38], analysis of joint citations [39], or citation-based content clustering (see, e.g., [40]), the PRISMA methodology prevents arbitrary decision making regarding the selection of studies [41,42]. In particular, this survey method uses the Boolean search query, which helps to identify relevant studies while providing fully replicable results [42]. This approach provides an important advantage over traditional surveys, which mainly use a random search method or “snowballing”, and can be selective in terms of search scope and thus potentially overlook important bibliographic information [42,43]. Importantly, the PRISMA method employs the screening and data extraction tool known as “Covidence” [44], which helps to reduce the potential selection bias associated with prioritizing studies concomitant with a particular research perspective [41].

Previous studies of the SC phenomenon employed some elements of the PRISMA methodology to conduct a general analysis of SC development [31], identify different components of SC governance [45], and summarize support tools for active ageing in SCs [46]. However, to the best of our knowledge, no studies that conducted to date have attempted to use the PRISMA methodology to determine a comprehensive set of criteria that can help in distinguishing between SCs and the remaining urban localities and assess a city’s progress towards greater smartness, which the present study aims to accomplish.

More than 40 different definitions of SCs are currently found in the literature (see Appendix A). While some definitions emphasize ICT and modern infrastructure [14,16,24,47,48,49,50,51,52,53], other SC categorizations emphasize the importance of human capital and quality of life [54,55,56,57,58,59,60,61,62] (see Figure 2A).

Following our understanding of SCs that emphasizes a human-centric rather than a techno-centric approach to city smartness (see Figure 2B), the following operational definition of SCs is adopted in this study:

SCs are cities that balance economic, environmental, and societal advances to improve the wellbeing of residents through a widespread introduction of ICT and other technological tools.

According to this definition, a particular urban feature (e.g., smart sensors or improved monitoring) contributes to city smartness only if it effectively improves the quality of life (QoL) of local residents; otherwise, it is not considered to contribute to city smartness.

Although the present study focuses on SC identification criteria, its methodological approach can be used for similar surveys on a wide range of socio-economic, technological, and environmental topics pertinent to urban development, such as healthcare, security, computing, environmental pollution, city infrastructure, residential reassurance, and others.

The remainder of the paper is organized as follows. In the next section (Section 2), we explain our methodological approach and define the sequence of research phases. In Section 3, we describe the results of the analysis, while in Section 4, we discuss our key findings. Conclusions and recommendations of the study are formulated in Section 5, which also outlines directions for future research.

2. Materials and Method

The literature survey approach we used for this study generally followed the PRISMA protocol guidelines [63] and was carried out in the following sequence: (1) selecting the search keywords; (2) selecting the databases and defining the search parameters; (3) building the initial search string and adapting it to specific databases; (4) conducting the primary search and assessing the abstract and full text for eligibility; and (5) evaluating the results. These phases are discussed, in some detail, in separate subsections below.

2.1. Selection of Search Keywords

Following our definition of the search topic as “SC identification criteria”, the following search components were defined a priori: the search object (“City”), as well as alternative terms (“Urban area*”, “Settlement*”, “Urban region*”, “Metropoli*”, “Township*”), modifiers (“Smart”, “Sustainable”, “Green”), and measures (“Criteri*”, “Measure*”, “Index*”, “Metric*”, “Parameter*”). In addition, to focus the search, we also defined several exclusion terms that are not directly related to our focus on urban development issues: “Biodiversity”, “Ecology”, “Culture*”, “Nation*”, “Polic*”, “Design*”, “Health”, “Climate”, and “Educat*,” where * represents a wildcat search term. Figure 3 features the main search terms used in the present systematic literature review of SC criteria.

2.2. Selection of Search Databases and Definition of Search Parameters

According to a recent study by Gusenbauer and Haddaway [64], 14 out of the 28 popular academic search engines meet the necessary requirements for systematic reviews and are suited for synthesizing evidence by employing PRISMA tools. These search engines include: ACM Digital Library, BASE, ClinicalTrials.gov, Cochrane Library, EbscoHost, OVID, ProQuest, PubMed, ScienceDirect, Scopus, TRID, Virtual Health Library, Web of Science Core Collections, and Wiley Online Library. Among these search engines, Scopus, Web of Science Core Collections, and ScienceDirect are the most popular and widely adopted for systematic literature surveys [22,23,26,28,31,46,65,66]. These three databases are used in the present study.

According to Anthopoulos [28], the first articles on the SC topic appeared in 1999. In the following decades, the number of articles on the SC topic has started to increase [65], mainly due to the implementation of various SC projects worldwide [31]. Therefore, the present study covers the period from 1999 on. However, the search was effectively limited to the 2011–2019 period because no studies on SC identification criteria were found before 2011 and because, at the time of its initiation, in August 2020, it was technically impossible to correctly analyze the data for the entirety of 2020.

The survey covers only original articles published in English, the main language of modern science [67], focusing on papers published in peer-refereed journals. Concurrently, the survey excludes the so-called “grey” literature, which includes government reports, policy statements and issues papers, conference proceedings, pre-prints, post-prints, professional bulletins, etc. [68]. However, we did not limit the search by the geographic origin of the publication or the organizational affiliation of the authors, considering such limitations irrelevant.

2.3. Building the Initial Search String and Adapting It to Specific Databases

The initial search string used Boolean syntax based on a combination of the “AND”, “NOT”, and “OR” operators [69]. However, due to the specificity of individual databases [70,71,72], the search string was adopted for each database separately (see Appendix B).

2.4. Conducting the Primary Search and Assessing the Abstract and Full Text for Eligibility

An initial search was conducted in October 2020. According to the selection criteria defined for the study (see Section 2.1), a total of 2140 publications were identified in Scopus, 2127 publications were identified in the Web of Science Core Collection, and 3880 publications were identified in ScienceDirect—8147 publications in total. Figure 4 reports the flowchart diagram of the publication selection and validation process, along with the number of entries retained (or excluded) at each stage of the reviewing process.

Following Moher et al. (2009), with the search performed in the “Scopus”, “Web of Science—Core Collections”, and “ScienceDirect” databases in October 2020.

The titles and abstracts of the pre-identified articles were screened next, with duplicate items and the following items excluded as not directly relevant to the research subject:

• Studies of longevity and health in SCs;
• Studies of societal inequality in SCs;
• Studies examining logistics and entrepreneurship in SCs;
• Studies of Big Data use in SCs;
• Studies of artificial intelligence (AI) in SCs;
• Studies on data protection and cyber security in SCs;
• Studies of navigation and urban topography;
• Studies of pricing in SCs;
• Studies of smartphone apps;
• Energy studies;
• Studies of social networks in SCs;
• Studies related to the physical parameters of urban space and patterns of land use;
• Studies related to specific ICT technologies, such as data collection, data storage, and data analysis.

Applying these exclusion criteria resulted in the selection of 296 papers eligible for the full-text analysis, which was our next step. At the final step of the screening process, only the papers reporting empirical criteria, measures, indexes, or quantifiable metrics of city smartness were retained, leaving us with 51 papers eligible for further categorization and analysis (see Figure 4).

2.5. Evaluation of Results

The 51 publications identified were analysed to extract the following information: year of publication, geographic origin, field of study, and empirical SC criteria used. Next, we grouped the extracted SC criteria into categories, and the frequency of their use was calculated.

3. Results

3.1. PRISMA Flow Diagram

The PRISMA flow diagram in Figure 4 illustrates the flow of information, details the number of records identified at each stage of the analysis, and reports the number of records included and excluded at each stage. As Figure 4 shows, the search started, as previously mentioned, with 8147 publications and then narrowed down to 51 publications, identified as relevant for subsequent reviewing.

3.2. Database Split

Out of the total number of 51 articles selected for full-text review, 12 articles (24%) came from Scopus, 7 articles (14%) from the Web of Science Core Collection, and 32 articles (63%) from the ScienceDirect. These differences in numbers can be explained by the fact that Scopus covers approximately 20% more journals compared to the Web of Science Core Collection [73]. Concurrently, the Scopus database mainly focuses on natural sciences, social sciences, humanities, engineering, medicine, and the arts [74], while ScienceDirect is a multidisciplinary platform which covers computer science, energy, engineering, life sciences, health sciences, economics, econometrics and finance, and social sciences [74,75], that is, topics directly relevant to ICT in general and SCs in particular.

3.3. Temporal Trends

Appendix C reports the thematic split of the studies based on SC identification criteria, while Appendix D summarizes the relevant studies in detail. Concurrently, in Figure 5, Figure 6 and Figure 7 we report the temporal trends in the annual number of publications on SC identification criteria (Figure 5), publication distribution across research fields (Figure 6), and geographic origin of the publications (Figure 7).

As evidenced in Figure 5, the first paper we identified on SC criteria was published in 2011, and research interest in the topic increased in 2013–2016, when 3–4 papers on the topic were published each year. After 2017, the number of publications on the topic surged, with eight papers being published in 2018 and twice as many in 2019. This increase of interest in SC identification is likely due to a number of recent developments, such as the initiation of Cisco’s USD 1 billion City Infrastructure Financing Acceleration Program [76]; the beginning of the Neom Smart City construction in Saudi Arabia, termed “the most ambitious project in history” [77]; the initiation of Japan’s national strategy for transitioning towards a Super Smart Society, or Society 5.0 [78]; and the implementation of Huawei’s SC concept in 60 cities worldwide [79].

3.4. Research Fields and Geographic Coverage

As Figure 6 shows, publications on SC criteria cover six main areas: computer science and engineering; economy, finance, and management; social science; energy; environmental sciences and social sciences; and arts and humanities. However, publications in computer science and engineering appear to be most frequent, with 20 papers out of the total 51 papers analysed, while only one paper on the topic was published in the field of arts and humanities.

Geographically, most papers on SC identification criteria originated in Spain (13%), USA (12%), India (7%), China (6%), and UK/Italy/Portugal/UAE/Japan (4% each) (see Figure 7). The predominance of papers originating in these countries can be explained by the extraordinary popularity of the SC concept in Europe, the Far East, and North America, specifically in well-developed countries, which have the necessary economic, technological, and social resources to build a modern SC. Concurrently, the contributions of India and China can be explained by their rapidly increasing technological capabilities and the magnitude of the urban problems these countries face [80].

3.5. SC Categories and Metrics

Figure 8 categorizes the SC identification criteria, starting from three main categories (economy and technology, environment, and society) to the nine generalized subgroups (smart digital technology, smart use of resources and energy saving, competitiveness, economic development, environmental (ecological) sustainability, social capital, leisure, smart management and policy making, and living conditions), as well as the 48 specific empirical metrics identified.

Somewhat surprisingly, only 29 metrics used by the empirical studies for SC identification (60%) actually relate to citizens’ needs—broadband subscriptions, access to free public Wi-Fi, the number of electric-vehicle charging stations, etc. The rest (i.e., 40%) are general technological or socio-economic performance measures, including the web use index, the number of startups, the share of the city’s water distribution network monitored by smart water systems, etc.

Empirical metrics: (1) Broadband subscriptions per 100 inhabitants (*); (2) Percentage of households with access to Internet (*); (3) Web Index; (4) Share of people who order goods or services over the internet (*); (5) Access to public free Wi-Fi (the number of wireless access points) (*); (6) Number of users of public transportation per 100,000 (*); (7) Percentage of public parking spaces equipped with real-time availability systems (*); (8) Percentage of public transport lines equipped with a real-time information system (*); (9) Number of startups; (10) Innovation Cities Index; (11) Labor productivity; (12) GDP per capita; (13) Change in gross household income; (14) Hourly wage (*); (15) Purchasing power parity (*); (16) Number of jobs created (*); (17) Unemployment rate (*); (18) Share of the city water distribution network monitored by smart water systems; (19) Percentage of rain and grey water re-used to replace potable water; (20) Percentage of the urban population that has door-to-door garbage collection with individual telemetering of household waste quantities (*); (21) Proportional share of the wastewater pipeline network monitored by a real-time data tracking sensor system; (22) Percentage of street lighting remotely managed by light management systems; (23) Number of electric-vehicle charging stations per registered electric vehicle (*); (24) Percentage of buildings (or housing units) with smart energy or water meters; (25) Number of real-time remote air quality monitoring stations per km² (*); (26) Proportional share of public buildings equipped for indoor air quality monitors; (27) Environmental Performance Index (EPI); (28) Proportion of population with secondary and higher education; (29) Number of universities in the city (or number of students per 1000 (*); (30) Expenditure on education per capita (*); (31) Expenditure on leisure and recreation per capita (*); (32) Corruption perceptions index; (33) Share of residents participating in online platforms (*); (34) E-Government Development Index (EGDI); (35) Number of online government services (*); (36) Extent to which public amenities are available within 500 m (*); (37) Decreased rate of travel time (*); (38) Happiness index (*); (39) Access to basic health care services/waiting time (*); (40) Percentage of the city area covered by digital surveillance cameras; (41) Emergency service response time (*); (42) Number of transportation fatalities per 100,000; (43) Number of acts of violence, annoyances, and crimes per 100,000; (44) Access to public outdoor recreation space or public outdoor recreation spaces (m²) within a 500 m radius from residential areas (*); (45) Increase in ground floor space for commercial or public use (*); (46) Life expectancy (*); (47) Morbidity and mortality (*); (48) Social inequality (GINI index or similar) (*).

(*) marks items that are oriented towards the needs of local residents (as opposed to general performance measures).

The breakdown of the SC identification criteria by research field is shown in Figure 9. As evidenced in this figure, papers on the economy and technology criteria compose 47% of the total number of publications, studies on societal criteria account for 40% of all publications, while environmental (ecological) criteria only appear in 13% of the publications. This uneven split effectively contradicts the sustainable development concept, which assumes a balance between the three main pillars of sustainable development—social, economic, and environmental [81].

This observation is confirmed when the relevant publications are disaggregated even further, down to the subgroup level (see Figure 10). At this level, papers on innovative technologies form 30% of the publications, with criteria pertinent to digital technology being used most often (23% of studies), while leisure-related criteria are considerably less frequent (2% of studies).

Somewhat unexpectedly, several SC metrics, such as the use of renewable energy sources, urban waste recycling, and green space expansion, which are important components of sustainable urban development in general appear in less than 2% of the studies identified and analysed in this review.

4. Discussion

The main objective of this study was to identify and categorize the main empirical criteria that help to differentiate between SCs and the remaining urban localities and to monitor a city’s progress towards greater smartness. The review helped us to identify 48 empirical metrics falling into 9 subgroups and 3 main groups—economy and technology, environment, and society. The study also enabled us to rank the identified criteria according to the frequency of their use in empirical studies, thus reflecting the current state of knowledge about metrics commonly used for SC categorization.

Sustainable development assumes an explicit balance between social, economic, and environmental development objectives [81]. Yet the present systematic literature review indicates that most previous studies placed an emphasis on the economy and technology and the society groups of the SC categorization criteria (87% of the studies analyzed), while the use of environmental criteria was less frequent (13% of studies).

SCs are meant to be designed and built around the experience of the people using them [82], with the main priority being citizens’ needs [83]. However, as the present study shows, only 60% of the empirical metrics we identified as commonly used for SC identification actually relate to citizens’ needs; the other indices are general technological measures. Most previous studies on the SC phenomenon also appear to have overlooked several important SC metrics, such as the use of renewable energy sources, urban waste recycling, and green space expansion, which are important components of sustainable development but appear in less than 2% of the studies identified and analyzed in this review.

The 2030 Agenda for Sustainable Development [9] specifies 17 Sustainable Development Goals. However, as the present study reveals, many of these goals are not covered by commonly used SC identification criteria. Such omitted categories include gender equality, peace and justice institutions, and partnerships to achieve the goals.

The results of this analysis correspond in part with the results reported by Tregua et al. [26], who define three main groups of criteria for SC categorization (environmental, social, and economic), and with the results reported by Lim et al. [31], who reveal that the availability of ICT infrastructure plays the key role in SC categorization. The results of this study are also similar to those reported by Ruhlandt [45], who emphasizes the importance of smart technology and smart governance, and to those reported by Marchetti et al. [30], who emphasize social and human capital, smart governance, and competitiveness as the most important factors for SC categorization.

Most previous reviews of the SC phenomenon placed an emphasis on resource management and ICT implementation [22,28,31,46,66], while this analysis was aimed at identifying empirical criteria for SC ranking and categorization.

In particular, the study identified 48 operational criteria for SC performance assessment (see Figure 8). The analysis also helped us to rank the SC criteria according to the frequency of their use, showing that the smart digital technology criterion was used most often in empirical studies (23% of the papers analyzed), while the use of the city leisure criterion was least frequent (2% of studies). This outcome contradicts the findings obtained by Ojo et al. [22] and by Lim et al. [31], according to which the most frequently used criteria for SC identification and categorization were participatory governance and ICT infrastructures, while the least frequently used criteria were developing an active self-decisive citizenry and social capital.

We explain these differences in findings by the fact that previous studies of SC metrics used survey approaches that do not set formal criteria for the literature selection; they may thus have introduced a selection bias. In particular, there is a possibility that a researcher performing a survey may be selective in collecting evidence based on the perspective adopted while ignoring evidence that points the other way [38]. By contrast, a systematic survey such as was used for this study provides an important advantage compared to other survey methods, such as content analysis [38], analysis of joint citations [39], or citation-based content clustering [40], that have no established study-section protocol and might thus overlook important bibliographic information [43].

The human-centric conceptual model of SCs proposed in this study (see Figure 2B) helps to differentiate urban features that contribute to city smartness from those that do not. In particular, according to our conceptual approach, a particular urban feature (e.g., sensors or improved monitoring) is considered to contribute to city smartness only if it effectively helps to improve the quality of life of local residents; otherwise, it is not considered relevant to city smartness.

Many SC identification criteria that are frequently used by empirical studies are general technological measures not directly relevant to human welfare, such as water reuse and the size of urban areas covered by surveillance cameras. The existing classifications are also often imbalanced, placing a major emphasis on technological elements. Therefore, we suggest an alternative, more balanced approach to SC identification and categorization, with an emphasis on human-centric objectives—whether a particular metric measures a change that helps to improve human welfare.

Following this approach, we suggest an SC classification system based on nine metrics, which are directly relevant to human welfare and which evenly represent each of the three main groups of SC criteria—economy and technology, environment, and society—with three identification criteria in each group. These nine metrics were selected from the list of 48 frequently used SC categorization metrics (see Figure 8) after applying two selection criteria: (1) equal representation of societal, environmental, and technological development aspects, and (2) direct QoL relevance.

Table 1. The proposed SC ranking system with an example of a comparative evaluation of two cities—Dubai (UAE) and London (UK).

Main Category	Subgroup	Metric	Rank (1–3) *	City Grade (1–5) **
				Dubai (UAE)	London (UK)
Economy and technology	Smart technology	Percentage of households with Internet access	3	5	4
		Access to public free Wi-Fi	2	4	5
	Economy	Hourly wage	1	4	5
Society	Social capital	Number of universities in the city (or number of students per 1000)	2	4	3
	Smart government	Online government services	3	3	4
	Quality of life	Life expectancy	1	4	5
Environment	Environmental sustainability	Percentage of the city population that has door-to-door garbage collection with individual telemetering of household waste quantities	1	1	3
		Number of electric-vehicle charging stations per registered electric vehicle	2	3	5
		Density of real-time remote air quality monitoring stations per km2	3	1	4
Overall ranking:				29	38

Note: * Ranked according to frequency of use in the literature; larger numbers reflect the most frequently used metrics (within the main categories). ** Information sources: Aqicn.org, Country economy, Dubai Municipality, Electric Vehicle Council, IHS Markit, International Telecommunication Union, London Authority, the NationMaster database, NUMBEO, Our World in Data, The World Bank, United Nations Department of Economic and Social Affairs, QS World University Rankings, Waqi.info, WiGLE, Wi-Fi Map App, and Wi-Fi UAE initiative. Larger numbers reflect values close to the possible maximum (optimal) of the corresponding metrics.

features the proposed identification and ranking system, with two cities—Dubai in the UAE and London in the UK—used as evaluation examples.

As Table 1 shows, according to the proposed ranking system, both cities scored relatively highly—29 and 38 points, respectively (out of the 45 points possible). This corresponds to the results of other assessments which place these cities high on the SC ranking scale [84,85].

We argue that the proposed ranking system can help several SC shareholders. In particular, using this system, urban decision makers can streamline their municipal policies by setting appropriate targets and gauging the progress achieved by their cities during transition toward greater smartness. Concurrently, technology providers can design and develop SC technologies, software products, and solutions for “smartening” cities by focusing on the most important aspects of transition from conventional cities to SCs. Moreover, researchers can focus their studies on the least researched topics related to SC transition, thus helping to further our understanding of the SC phenomenon and formulate informed development policies.

5. Conclusions

The present study was aimed at determining a comprehensive set of criteria that can help us to distinguish between SCs and the remaining urban localities and to assess a city’s progress towards greater smartness, while focusing on the human-centricity of each criterion, that is, its potential impact on the quality of life of urban residents.

The proposed SC identification system includes nine identification criteria, equally representing three main sustainable development categories—economy and technology, environment, and society—and directly relevant to improving the QoL of local residents.

This classification thus differs from traditional SC categorizations in which some categories contain more metrics than others and many metrics are general technological measures not directly related to human welfare.

The study also helps us to identify topics which require further research attention. These topics include: (1) ranking SC criteria according to their temporal attributes (i.e., short-term progress evaluation vs. strategic long-term impact assessment), and (2) ranking SC evaluation metrics according to their performance in assessing the extent of a city’s “smartness”.

Lastly, we should remark that although the present study focuses solely on SC identification criteria, its methodological approach can be used in similar surveys on a wide range of urban development topics.