Urban Smartness and City Performance: Identifying Brazilian Smart Cities through a Novel Approach

Abstract

While smart city transformation is a remarkably popular topic among urban policymakers across the globe, there is little evidence on how to evaluate a city’s smartness level accurately. This study aims to bridge this knowledge gap by applying a novel assessment framework to a case study context and generating useful insights. To achieve this aim, the study evaluates the smartness levels of 27 Brazilian state capital cities through the indicators of productivity and innovation, livability and well-being, sustainability and accessibility, governance and planning, and connectivity and innovation. This urban smartness analysis is conducted through a smart city assessment framework that brings up three categories of smart city performance types—i.e., leading, following, and developing. The findings of the analysis revealed that the common characteristics of cities with leading smartness performance are having: (a) a strong innovation ecosystem; (b) Specific legislation for developing entrepreneurship; (c) Training opportunities for skilled labor; and (d) Conditions for knowledge-based development and digital transformation offerings and readiness. The analysis identified the smartest cities in Brazil as follows: Florianópolis, São Paulo, Vitória, Curitiba, Porto Alegre, Brasília, Belo Horizonte, Rio de Janeiro, and Cuiabá. This study offers insights from the application of a novel method in the Brazilian context for the local authorities to consider adopting for smart city performance and progress analyses and subsequently making necessary interventions to transform their smart city policy and practice to realize their desired goals.

1. Introduction

The concept of a smart city, from the perspective of City 4.0, is characterized as a system of systems with sustainable and balanced practices in the four dimensions of economic, societal, environmental, and governance, generating outcomes for human and non-human outcomes [1]. Over the past decade and a half, as part of the smart city agenda, smart urban technologies have inundated our cities with the intention of forming the backbone of a large and intelligent infrastructure. Along with such a development, the dissemination of a sustainability-driven ideology has had a significant impact on the planning and development of our cities. Today, the concept of a smart city is regarded as a vision, a manifesto, or a promise with the objective of constituting the 21st Century’s sustainable and ideal city. In other words, a smart city is an efficient, technologically advanced, green, and socially inclusive city. This is to say, smart city applications place a particular technology focus at the forefront of generating solutions for ecological, societal, economic, and management challenges [2]. Fundamental smart city themes are shown in Figure 1.

Despite the criticism of the technocentricity of smart cities, it is common sense among scholars that it is a good thing to rethink our cities’ planning and development paradigms and processes in an era of digital disruption and climate change. This belief has turned the notion of a smart city into an important urban innovation agenda [4,5]. In this sense, smart cities actively adopt new technologies with the objective of achieving desired urban outcomes, the most common of which involve productivity, sustainability, accessibility, wellbeing, livability, and good governance [6].

Today, many cities around the world are developing their own smart city agendas. Since smart city transformations require high investments, the key to success in such a transformation is to be ‘smart’ about smart city development. This means that understanding smart city transformation readiness is the key, given that smart city investment is not an inexpensive one [7]. Nevertheless, there is little evidence on how to accurately evaluate a city’s smartness level to understand how such a transformation could be driven.

A key aspect of smart city readiness is the availability of open data to developers, who leverage it through applications such as mobile apps to enhance the efficiency and effectiveness of urban functions. Similarly, reliable and representative data inputs are essential for its success. To effectively drive urban innovation, it is imperative to have access to relevant data that accurately reflects the socioeconomic realities of cities. This data serves as the foundation for informed decision-making, enabling urban planners, government professionals, and technologists to develop impactful strategies and solutions.

Without comprehensive and contextually accurate data, the potential of these networked infrastructures and technological platforms to drive positive change within cities would be severely limited. Therefore, the integration of high-quality socioeconomic data into these innovative systems is paramount to ensuring their effectiveness and enabling the realization of their full potential. After all, data is the new oil of the contemporary economy, and as much as data has enhanced analytical system capability, in many cases, artificial intelligence (AI)-driven systems are critical [8,9]. In other words, there is a lack of knowledge in academic literature and practice on which data or indicators and which metrics or analytics can be used to understand smart city readiness [3,10,11].

Addressing this knowledge gap is the raison d’être of this paper. In that perspective, to evidence the findings in a case study context, we adopted the research question, ‘What are the smartness levels of Brazilian state capital cities?’ In a nutshell, his study places Brazil’s 27 state capital cities under the smart city readiness assessment microscope to understand their performance and potential to really become smart cities, and compares their performance with other cities from the same country context by also considering the goal of sustainable urban development.

To sum up, the paper addresses the concept of sustainable urban development and the trend of using advanced information and communication technologies (ICTs) to achieve sustainability goals in cities, resulting in the emergence of the smart city concept. While smart city applications have been criticized for their technocentric approach, they offer opportunities to rethink urban planning and development paradigms in the face of digital disruption and climate change. The paper highlights the importance of understanding a city’s smartness level to drive a successful smart city transformation and addresses the knowledge gap by evaluating the smartness levels of Brazilian state capital cities. The paper’s originality and novelty lie in its focus on evaluating the smartness levels of Brazilian state capital cities, which can serve as a model for other cities globally, and its contribution to policymakers’ decision-making towards making their cities smarter.

Following this introduction, Section 2 presents a concise review of the literature, the methodological approach, and the case study context. Section 3 reveals the results of the analysis, while Section 4 offers a discussion and concludes the paper with key findings and future research directions.

2. Materials and Methods

2.1. Literature Background

Since the turn of the century, the impacts of global climate change have become more catastrophic. To tackle such an issue, advanced information and communication technologies (ICTs) are now seen as a potential panacea; in other words, ICTs may somehow reverse or ease the impacts of our unsustainable urbanization, industrialization, and consumerism practices [12]. The potential of advanced ICT applications in environmental decision-making is widely recognized. Due to the technology currently at hand, many governments—at local, regional, state, national, and supra-national levels—around the world have relied on technology to solve problems, which has originated the ‘smart city’ concept [1].

The concept of sustainability has become a dominant policy underpinning the planning agendas of both developed and developing countries [13]. This is mainly due to externalities—e.g., climate change, non-renewable resource depletion, air, water, and land pollution, pandemics, rapid and sprawling urbanization, and social inequalities—which are usually disregarded until they reach the level where their consequences may threaten citizens’ overall wealth and wellbeing [14]. The trend of growing urban populations and the citizens’ needs thereof have reinforced the importance of actions towards reaching the goals of sustainable communities and cities [15].

The popularity of the sustainability concept has spawned a new development type, namely sustainable urban development [16]. The term sustainable urban development is contradictory since it joins words with completely different meanings—that is an oxymoron. Sustainability implies preserving the ecosystem and its services while human needs are met; in contrast, urban development refers to any activity that improves people’s quality of life by depleting natural resources and devastating natural areas. Nonetheless, while urban development cannot be fully sustainable, sustainable urban development, in general, refers to a less harmful or intrusive type of development for natural ecosystems [17].

In the sustainable urban development perspective, cities are made by and for people, but keeping in mind the balance between economic development, environmental responsibility, and social justice is of utmost importance [18]. In this scenario, urban planning—which includes identifying conditions of population density and demand for housing, infrastructure, energy, and mobility—becomes essential. Given the continuous growth of the urban population on our planet, cities must find solutions to provide public services, achieve social equality, and improve citizens’ quality of life. At this very moment, technology has become the tool to increase efficiency and optimize the proposed solutions, which then led to the conceptualization of a new city brand, namely ‘smart city’ [19].

The evaluation of smart cities has become an increasingly important topic in recent years as urban areas around the world undergo rapid technological advancements to improve the quality of life for their residents. Smart cities are characterized by their innovative use of digital technologies to optimize infrastructure, services, and resources for sustainable urban development [20]. Evaluating the effectiveness of these technological interventions in achieving the goals of a smart city requires a multidimensional approach that encompasses various aspects, such as technological, social, economic, environmental, and governance dimensions. This evaluation process involves collecting and analyzing data from diverse sources, including sensors, networks, and citizens’ feedback, to assess the impact and outcomes of smart city initiatives [21]. As the concept of smart cities continues to evolve, robust evaluation methods and frameworks are essential for guiding policymakers, urban planners, and stakeholders in making informed decisions for the sustainable development of urban areas in the digital era.

The many definitions of smart cities found in the literature mention the use of technology to improve communication across companies, collectives, institutions, and individuals; to solve environmental issues such as waste management and energy production; to improve access to healthcare, education, and transport by expanding the operational functioning of the network; and to increase efficiency in the way services are provided and controlled [22].

Cities face challenges that are so complex that the need for innovation in all aspects of policymaking and public service leads to seeking new solutions. Such challenges demand a transformative change in the way society works, lives, and designs a new future; in turn, this change imposes a particular burden on those responsible for governing such processes with an optimal use of available public resources [23]. To contribute to this transformation, the International Standardization Organization (ISO) is developing a new set of international standards designed for a holistic and integrated approach to sustainable development and city resilience.

Various indicators and standards can be used to assess smart cities’ performance and sustainability. City measures are of interest to all city stakeholders, as they can assist target setting, monitoring, and decision-making [24]. However, different perspectives and contexts result in a great variety of city measures and classification models [25], with no ‘one-size-fits-all’ solutions. In this sense, cities should choose the indicator sets most suitable to their specific context and needs [26].

Internationally standardized frameworks of smart city indicators were developed by institutions such as ISO, the International Telecommunication Union (ITU), and the coalition of the European standardization organizations CEN, CENELEC, and ETSI [26]. There are also globally renowned research-based and market-oriented indices, such as the Cities in motion index (CIMI), the digital city index (DCI), the global e-government survey, the innovation cities index (ICI), the smart city governments (SCG), and the smart city index (SCI) [27]. Indices specifically tailored to Brazilian reality are the Brazilian smart city maturity model (SCMM), the urban systems ranking connected smart cities model, and the Brazilian network of smart and human cities model (RBCIH) [25].

Smart cities respond to issues such as climate change, rapid population growth, and political and economic instability by primarily improving how they involve society, applying collaborative methods of leadership, working across disciplines and city systems, and using data information and modern technologies to deliver better services and quality of life to people in the city (residents, businesses, and visitors), now and for the foreseeable future, without unfair disadvantage to others or degradation of the natural environment. In addition, there is an important document entitled ‘Brazilian Charter for Smart Cities’ [28] that provides its own definition of smart cities: cities committed to sustainable urban development and digital transformation in their economic, environmental, and sociocultural aspects, which act in a planned, innovative, inclusive, and integrated manner. Smart cities ensure the safe and responsible use of data and ICTs and promote digital literacy, collaborative governance and management, and the use of technologies to solve concrete issues, create opportunities, offer services in an efficient way, reduce inequalities, and improve the quality of life of all people [29].

We have adopted a conceptual smart city framework based on an input-process-output-impact model that is widely used in urban and regional planning (Figure 2). This framework was used in the report ‘Smart Cities Down Under: Performance of Australian Local Government Areas’, and it advocates the importance of a smart city as an organic whole of a network that may benefit from technologies and innovation but does not depend exclusively on them [3]. We underscore that these definitions bring this discussion closer to the Right to the (Smart) City and imply that the perspective of smart cities is in line with the development of the social functions of the city towards their convergence, reinforcing the understanding that technology must be used to achieve more human development and to ensure that no one is left behind.

2.2. Methodology

The concept of a smart city has been a topic of much academic discussion in recent years. However, there is no clear consensus on what exactly makes a city ‘smart’. Some argue that it is all about technological advancements, such as the Internet-of-Things (IoT) and big data, while others point to factors such as sustainability, livability, and citizen participation. One of the main reasons why it’s difficult to measure the ‘smartness’ of a city is the lack of a universal definition and standardized metrics. Different cities have different priorities and goals, and what may be considered smart in one context may not necessarily apply to another. Additionally, measuring the impact of technological innovations on a city’s overall performance and well-being is a complex task that requires sophisticated data collection and analysis.

As such, the debate about what makes a city truly smart remains ongoing and requires ongoing interdisciplinary research and collaboration. Given the complex nature of measuring the smartness of a city, the current research assessed the feasibility of finding convergences by using similar indicators applied to cities with different realities in Australia and Brazil. Although the study did not make direct comparisons between cities in Australia and Brazil, by applying a similar model used in a previous study in Australia to Brazilian cities, it is possible to evaluate the feasibility of using similar indicators and identify the need for adjustments to suit each reality. This can help consolidate a group of common indicators, leaving room for specific indicators that reflect the particularities of each city. The resulting dataset can contribute to the development of a more standardized and applicable model for assessing the smartness of cities in different contexts.

In our research, we have embraced the notion of accuracy in data analysis, which refers to how closely the results derived from an analysis align with the actual or true value of the data being analyzed. The criteria employed to define accuracy in data analysis vary based on the nature of the data and the specific analytical method being utilized. To ensure the accuracy of our study, we have included specific data related to the socioeconomic characteristics of Brazil. Moreover, we have incorporated other criteria, such as precision, which measures the consistency of the results over repeated analyses, and validity, which assesses whether the analysis measures what it intends to measure. By taking these additional criteria into account, we have been able to obtain more reliable and precise results.

Smart city is a multi-faceted concept; however, the main available studies focus on key domains or provide global rankings, e.g., cities in motion [30], European digital city index [31], innovation city index [32], smart city strategy index [33], global cities index [34], and global livability index [35], easy parks’ smart city index [36], and the top-50 smart city governments index, developed by Eden Strategy Institute and ONG&ONG [37]. Although they present relevant methodologies for an overview of the evolution of Smart Cities in different places on the planet, such reports do not present a holistic perspective of the performance of smart cities or even their development potential. Therefore, we adopted a comprehensive smart city assessment model—the smart city assessment model—whose systemic foundation is based on the conceptual framework of smart cities. The model evaluates the achievements of smart cities and urban regions with a large base of multivariable indicators according to the conceptual perspective presented in Figure 2. This methodology was adopted in 2020 to analyze cities in Australia and proved to be valid for evaluating the performance of smart cities in relation to the proposed set of indicators [38].

The analysis of information and the choice of indicators were based on the study carried out in Australian cities [1]; however, they were made in such a way that the final selection reached the idiosyncrasies of the Brazilian context. As a result, the smart city model originally used in the Australian context is now adapted to a Latin American smart city model with a selection of indicators consistent with reality that allows the creation of reliable and authentic portraits of cities. In addition to adapting the indicators, this study included the technology dimension and also new indicators closely associated with Brazilian reality, such as slums, sanitation, and education. In our Brazilian study, we analyzed the indicators of the capital cities of all 27 Brazilian states. The data were collected from official open databases, and a total of 25 indicators were analyzed for each of the 27 cities. Once the data were obtained, they were normalized on a scale of 0 to 1 and distributed across the five dimensions of the model: productivity and innovation, livability and wellbeing, sustainability and accessibility, governance and planning, and connectivity and innovation. By analyzing these indicators, we were able to gain insights into the performance of each city across the various dimensions of the model, providing a comprehensive view of the smart city landscape in Brazil.

The methodological steps include the following:

• Adapting an indexing framework for smart city assessment;
• Determining indicators of the framework;
• Determining the weightings of the indicators;
• Collecting data via primary and secondary data collection techniques;
• Using statistical techniques to scale and normalize data for comparison;
• Conducting statistical and descriptive analyses of the findings

The specific indicators of Brazilian reality were selected from the prominent smart city literature on the basis of the following key principles: measurability; analytical soundness; comparability; geographic coverage; data availability, and; relevance and suitability. The rationale behind the selection of these indicators is listed in

Table 1. Indicator descriptions and references used in the study [39].

Indicator Area	Indicator	Description	Rationale and Reference
Productivity & Innovation	Economic Productivity	Gross domestic product (GDP) per capita at current prices	To make smart cities more effective for higher economic development, macroeconomic factors must be linked with regular urban policies [40].
	Labor Force Participation	% of formal workers in working population	Smart cities provide increased employment opportunities in the knowledge and service sectors [41].
	Talent pool	% of labor force with university education	Highly educated workers are the backbone of smart cities, stimulating economic growth and vibrancy [42].
	Innovation Industries	% of companies categorized as knowledge-intensive	Innovation industries form the economic core of smart cities [43].
	Income	Average monthly salary of formal workers	Smart cities are claimed to be prosperous locations that generate wealth and disposable household income [44].
Livability & Wellbeing	Health Status	% of the population covered by a health insurance according to data from National Health Agency	Smart cities develop and implement policies to improve the health conditions of their residents [45].
	Education	Basic education development Index (IDEB) score	For Smart Cities, the need to educate all citizens is a basic element of development [46].
	Safety and Security	Deaths from external causes (i.e., accidents, violence) per 100,000 people	Digital security, health security, infrastructure, and personal safety are integral elements of smart cities [47].
	Housing Affordability	% of households in irregular settlements	Housing affordability is a critical element in facilitating the varied skill sets that support sustainable innovation economy of smart cities [48].
	Socioeconomic Progress	% of individuals categorized as low income	Smart economy of smart cities should be socially inclusive to address the urban inequity issue [49].
Sustainability & Accessibility	Sustainable Commuting	Bus fleet per 100,000 people	Smart cities aim to develop innovative services for sustainable transport and mobility [50].
	Sustainable Vehicles	% of vehicles categorized as electric or hybrid	Mobility strategies in smart cities promote cleaner mobility options [51].
	Sustainable Energy	Installed power of solar radiation per 100,000 people	In realizing the energy supply of a smart city, it is essential to maximize the use of renewable energy sources [52].
	Sustainable Buildings	Buildings with sustainability certifications per 100,000 people	Smart cities contain buildings that are designed, built, and utilized to consume less energy and facilitate efficient building operations [53].
	Sanitation	Basic sanitation index (i.e., water, sewage, solid waste) score	Smart Cities need to find a creative, innovative, and useful way to expand infrastructure (water and sanitation, energy, transportation, housing, information, and communications) [54].
Governance & Planning	Smart Policy	Existence of municipal smart city or urban innovation policy	A smart city policy is necessary to establish a shared democratic approach to engage leaderships from local institutions and to priorities local issues [55].
	Sustainability Policy	Existence of municipal sustainable development policy	A smart city affects sustainable planning through changes in urban infrastructure (energy, land-use, water, sanitation and waste management, and transportation) and the structure of urban governance [56].
	Participation	Urban civil associations per 100,000 people	A smart city listens and gives voice to everyone [57].
	Research Support	Research and development (R&D) funding (FINEP)	Public R&D is important because it can create advances in the underlying technologies of smart cities that all smart city stakeholders can benefit from [58].
	University	International ranking of the most prestigious university (QS World University Rankings 2022)	The universities have a diverse influence on the development of society. Today, this also includes countless smart city and community initiatives all over the world [59].
Connectivity & Innovation	Broadband Internet	Broadband internet coverage	World-class broadband provides opportunities for inclusion in the innovation economy, which is the core economic activity of smart cities [60].
	Public Wi-Fi	Free Wi-Fi spots per 100,000 people	Smart cities offer public Wi-Fi networks to increase connectivity and access to smart services [61].
	Innovation	Patents registered per 100,000 people	Policy makers need efficient and effective tools to measure and monitor innovation-related performance so that they can develop new measures and policies and evaluate current approaches [62].
	Research capacity	Research grants per 100,000 people (PQ-CNPq)	Universities act as knowledge intermediaries, knowledge gatekeepers, knowledge providers, and knowledge evaluators for smart cities [63].
	Media	Local digital press media per 100,000 people	a smart city listens and gives voice to everyone [64].

, and supported with relevant references. The model, as its default, uses an equal weighting for its indicators.

In this study, all 25 indicators in Table 1 were equally weighted at 0.04 (25 of them adding to 1). Alternate weighting assignments are also possible within the model. Table 1 illustrates the model structure and indicator descriptions.

As shown in the last column of Table 1, we adopted studies that used or advocated the use of the same smart city indicators and concepts as those used in this study as references. The table also shows the technology dimension included in the model with the broadband internet coverage indicators: free Wi-Fi hotspots per 100,000 people; patents filed by 100,000 people; research grants per 100,000 people (PQ-CNPq); local digital press per 100,000 people.

In addition to the technology dimension, some indicators revealing the singularities of Brazilian cities were included. For example, to measure housing accessibility, the percentage of residences in subnormal agglomerations over the total number of residences was applied. These houses are characterized by an irregular urban pattern and a lack of basic sanitation and are usually located in favelas. Favelization is an indicator of urban chaos and reveals local idiosyncrasies that continue to be a fundamental challenge for the transformation of smart cities.

The set of indicators adopted for the analysis of Brazilian capital cities can be seen in Table 1. Once the raw indicator data have been attained, the ‘Smart City Assessment Model’ standardizes the data so they can be used in the index.

This procedure converts raw indicator values to a standard scale (ranging between 0 and 1). The index uses the ‘min-max normalization’ technique to reflect the best and worst performers. The min–max normalization of the data is calculated in accordance with the following equation:

$$I new=\frac{I raw-I min}{I max-I min }$$

The result (I new) corresponds to the city’s indicator value; I raw, I min, and I max denote normalized, original minimum, and maximum scores for each indicator, in the order mentioned. After each city receives its composite indicator scores, three performance categories (also called clusters) are formed using the quantile method—that is, the cities are sorted in ascending order and divided into three equal groups. After the clusters are formed, the three performance categories or groups are labeled as: (a) leading, corresponding to the best performing cities; (b) following, corresponding to cities with good results and potential but not as much as leading cities; and (c) development, corresponding to cities with progress and potential not as substantial as the other city categories.

In our analysis, we recognize that a smart city cannot be evaluated based on a single indicator, such as the presence or absence of universities. Instead, we need to consider a range of indicators that capture different aspects of urban life, including social, economic, and environmental factors. While the quality of universities is certainly an important indicator, it is just one piece of the puzzle. By taking a comprehensive approach to assessing cities, we can better understand their strengths and weaknesses and identify opportunities for improvement. This approach also allows us to consider the complex interrelationships among different indicators and how they contribute to the overall well-being of the city and its residents. Ultimately, by evaluating cities based on a set of diverse indicators, we can gain a more complete and nuanced picture of what it means to be a smart city and how we can work towards creating more sustainable and equitable urban environments.

There are a limited number of smart city assessment approaches developed so far, e.g., [10,11,26,65,66,67,68,69,70]. Nevertheless, these approaches are not comprehensive enough to adequately evaluate the smart city readiness of urban locations, as discussed by Yigitcanlar et al. [7]. Additionally, “the factors affecting urban smartness is still an understudied area of research. Besides, there is limited empirical evidence in the literature on what the exact factors of urban smartness—a key indicator for smart city transformation readiness—are”. The methodology adopted in this study and introduced above offers comprehensive coverage of smart city readiness aspects.

2.3. Case Study Context

Brazil is a three-tiered federation: the Federal Government at the national level, the States at the regional or state level, and the municipalities at the local level. Brazil is formed by 26 states and 5570 municipalities. The 27th one, which is the Federal District, where the city of Brasília—the country capital—is located, accumulates both state and municipal competences. The 27 cities home to the states’ political and administrative centers are ‘state capitals’. The 27 state capitals—which are the object of this study, alongside the country capital, Brasília—are regional localities of political, institutional, and economic relevance; however, they have the same legal status and competences as any municipality in Brazil.

According to the 1988 Brazilian Constitution [71], urban planning is essentially a local, or municipal, issue. As such, municipalities have exclusive powers to approve their own master plans in compliance with the guidelines laid out by the national statute of the city. Brazil’s constitution also grants municipalities the power to provide access to culture, education, science, technology, research, and innovation at the local level.

Technology is only mentioned by the national statute of the city—entered into force by a national law promulgated in 2001 [72]—as applied to civil construction: it is recommended that sustainable technological solutions be used to reduce environmental impacts in the city. In contrast, the National Science, Technology, and Innovation Code, edited in 2016 [73], provides for the creation, implementation, and consolidation of environments that promote innovation, such as technology parks, centers, and incubators, by municipalities to foster technological development, competitiveness, and local-level interaction between the private sector and scientific, technological, and innovation institutions.

Many municipalities have passed local regulations on innovation and technology, encompassing not only incentives and fiscal grants but also the application of new technologies in urban planning, disaster management, resource management, mobility, and climate change adaptation, among other fields. Numerous smart city projects and initiatives have been developed by large and medium-sized cities in the country with the aim of attracting investors and enhancing urban quality of life.

In this sense, it is worth mentioning the connected smart cities (CSC) Ranking, an annual report published by the private consultancy urban systems, which ranks Brazilian cities according to 70 indicators distributed across eleven sectors: mobility, environment, urbanism, innovation and technology, health, public safety, education, entrepreneurship, energy, governance, and economy. The CSC Ranking is currently in its 7th edition [74] and has consolidated its position as a national reference on smart city development in the public, private, and third sectors.

In 2018, to replace the former Digital Cities Program, Federal Decree No. 9612 [75] determined the creation of a Smart Cities Program to be managed by the Ministry of Science, Technology, and Innovation (MCTI) and jointly developed by the National IoT Plan and the National Cities 4.0 Chamber. A technical cooperation agreement was signed by the MCTI and the Ministry of Regional Development to set up a national strategy on sustainable and smart cities. In 2020, this national strategy was published in the form of the Brazilian Charter for smart cities [76], a political document that provides for a public agenda for the digital transformation of Brazilian cities and relies on eight strategic objectives and policy recommendations.

Nonetheless, data on internet connectivity in Brazil for the year 2019 shows there is a huge sociotechnical gap that needs to be bridged [77]. Around 17.3% of Brazil’s households, or an estimated 12.6 million, are not connected to the internet. The main reasons given by people living in those households for their lack of internet connectivity are lack of interest in connectivity (32.9%), digital illiteracy (25.7%), and high prices for services and equipment (25.7% and 5%, respectively). Only 6.8% of the households reported service unavailability as the main reason for their lack of connectivity. Another important factor is regional disparity: the highest percentages of households without internet access (24% and 25.7%, respectively) were found in the country’s north and northeast regions.

In conjunction with Brazil’s enormous socioeconomic inequities and regional disparities, the Brazilian charter for smart cities is strongly positioned in favor of promoting equality and social inclusion by fostering the digital transformation of Brazilian cities. The charter points out the traits of smart cities it aims to implement: diverse and fair, alive and for people, connected and innovative, inclusive and welcoming, safe, resilient, and self-regenerative, economically abundant, environmentally responsible, articulating different notions of time and space, aware, independent in the use of technologies, watchful, and responsible towards its principles. Hopefully, these very goals will be incorporated into the National Policy for smart cities, currently under discussion in the Brazilian Congress (Bill No. 976/2021) [78].

Against this backdrop, the Brazilian capital city context forms an ideal case for the adoption of the smart city assessment model to evaluate smart city transformation readiness.

3. Results

3.1. Locational Performance Variances

The datasets for our study were collected from official government databases at the national and local levels in 2022. We used a quantile method to categorize the progress and achievements of each city based on their normalized index scores. This method involved dividing the frequency distribution into three equal performance groups, or clusters: leading, following, and developing. Our analysis revealed that while some cities performed well in certain dimensions of the smart city assessment model, others struggled in different areas. These findings highlight the need for a comprehensive and nuanced approach to evaluating smart cities, one that recognizes the interrelationships among different indicators and the importance of balancing economic, social, and environmental goals.

Table 2. Smart city performance clusters of BCCs. Reprinted/adapted with permission from Ref. [39]. 2022, Fachinelli, A. et al.

Capital City	State	Region	Population	Score	Cluster
Florianópolis	Santa Catarina (SC)	South	508,826	0.72	Leading
São Paulo	São Paulo (SP)	Southeast	12,325,232	0.60	Leading
Vitória	Espírito Santo (ES)	Southeast	365,855	0.59	Leading
Curitiba	Paraná (PR)	South	1,948,626	0.58	Leading
Porto Alegre	Rio Grande do Sul (RS)	South	1,488,252	0.56	Leading
Brasília	Distrito Federal (DF)	Midwest	3,055,149	0.51	Leading
Belo Horizonte	Minas Gerais (MG)	Southeast	2,521,564	0.47	Leading
Rio de Janeiro	Rio de Janeiro (RJ)	Southeast	6,747,815	0.46	Leading
Cuiabá	Mato Grosso (MT)	Midwest	618,124	0.45	Leading
Palmas	Tocantins (TO)	North	306,296	0.44	Following
Goiânia	Goiás (GO)	Midwest	1,536,097	0.41	Following
Campo Grande	Mato Grosso do Sul (MS)	Midwest	906,092	0.38	Following
João Pessoa	Paraíba (PB)	Northeast	817,511	0.34	Following
Recife	Pernambuco (PE)	Northeast	1,653,461	0.31	Following
São Luís	Maranhão (MA)	Northeast	1,108,975	0.31	Following
Natal	Rio Grande do Norte (RN)	Northeast	890,480	0.31	Following
Teresina	Piauí (PI)	Northeast	868,075	0.30	Following
Rio Branco	Acre (AC)	North	413,418	0.28	Following
Fortaleza	Ceará (CE)	Northeast	2,686,612	0.27	Developing
Aracaju	Sergipe (SE)	Northeast	664,908	0.26	Developing
Macapá	Amapá (AP)	North	512,902	0.24	Developing
Boa Vista	Roraima (RR)	North	419,652	0.24	Developing
Porto Velho	Rondônia (RO)	North	539,354	0.23	Developing
Salvador	Bahia (BA)	Northeast	2,886,698	0.22	Developing
Belém	Pará (PA)	North	1,499,641	0.22	Developing
Maceió	Alagoas (AL)	Northeast	1,025,360	0.21	Developing
Manaus	Amazonas (AM)	North	2,219,580	0.18	Developing

shows the 27 Brazilian capital cities (BCCs) in one of three clusters based on their indicators.

As can be seen in Table 2, the best-performing capital cities in Brazil are the most densely populated metropolises. Moreover, the top capital cities are mainly distributed in the southern region, although two of the nine cities are in the central region. Consequently, capital cities located in the North and Northeast regions perform poorly on the indicators of the smart city assessment model. Some factors that may explain these results include:

• Basic educational development;
• Development of the public health care system;
• Amount of employment and income opportunities;
• Presence of knowledge-intensive enterprises and innovation results;
• Existence of a municipal sustainable development policy.

3.2. Overall Findings

Figure 3 contains the two graphs that illustrate the performance levels of Brazilian capitals as smart cities. In the top chart, the lines show the averages for each of the ‘Leading’, ‘Following’, and ‘Developing’ clusters in the six dimensions studied. In the bottom graph, it is possible to see these performances by indicator. The performance diagrams of all 27 BCCs are presented in Appendix A.

The study highlights a stronger high contrast in productivity and innovation (particularly in industry—existence of knowledge-intensive companies and economic productivity—GDP per capita) between leading BCCs and the following and developing BCCs, as these two clusters are closer in the performance of indicators in this area. The analysis points out a general weakness in connectivity and innovation (particularly in research capacity and free wi-fi spots) and sustainability and accessibility (particularly in sustainable energy and sustainable commuting/transit) across all the investigated BCCs. Overall, the findings suggest a relatively stronger performance of leading BCCs in livability and wellbeing (particularly in progress and housing) areas and productivity and innovation (particularly in labor force and knowledge-intensive companies) areas.

There are two key findings that need further elaboration. The first one concerns the increase in the number of favelas in Brazil, which has doubled in the last decade. Between 2010 and 2019, the number of subnormal agglomerations, such as slums and stilts, went from 6329 in 323 municipalities to 13,151 in 743 cities. These houses are characterized by an irregular urban pattern and a lack of basic sanitation. The favela presence is an indicator of urban disruption, which has existed for decades due to social, economic, territorial, environmental, and political inequalities. The second issue relates to connectivity and innovation scores. This indicator area was included in the smart city assessment model for the study carried out in the BCCs due to the specific context of Latin American cities pointed out by Marchetti et al. [54]. The study found that this was the area with the lowest scores in all clusters (normalized values for Leading = 0.411, Following = 0.230, and developing = 0.703).

3.3. Comparative Findings: Leading vs. Following

The performance levels of the following BCCs are significantly lower in all categories (productivity and innovation, livability and wellbeing, sustainability and accessibility, governance and planning, and connectivity and innovation) compared to the leading BCCs. Leading BCCs (normalized index score of 0.551), in general, have a statistically significant higher performance compared to the following BCCs (normalized index score of 0.343). Nevertheless, the following BCCs are not necessarily performing poorly in all indicators compared to the leading BCCs. For instance, the following BCCs average cluster performance is higher in sustainable energy (a normalized score of 0.227 against the leading BCCs 0.193) and sustainable commuting (0.374 vs. 0.357) indicators compared to the best-performing BCCs. This difference partly reflects the low incentives for the use of renewable energy and the deficiencies in public transportation that still exist in the best performing BCCs (leading).

An important point to highlight is that the findings of this study should not be considered as a definition of success or failure in the smart transformation journey of the cities evaluated, but rather as a guide to analyze the results of the 25 to identify strategies and opportunities, as well as to strengthen the weaknesses and sustain the strengths of each capital. This is justified when we analyze the similar performances of two BCCs. The combined effect of relatively weaker performances could raise one of the BCCs to the leading cluster and downgrade another to the Following cluster. For example, Cuiabá (MT) was classified as leading and Palmas (TO) as following, despite having a similar indexed score with almost zero differences in performance.

It is important to understand that the findings of this study should not be seen as a definitive measure of success or failure in the journey towards becoming a smart city for the 25 Brazilian state capital cities evaluated. Instead, the results can be used as a guide to analyze the performance of each city and identify strategies and opportunities for improvement. The idea is to focus on strengthening the weaknesses and sustaining the strengths of each city rather than seeing the results as a pass-or-fail evaluation. It is important to remember that the performance of a city in a single metric or index does not necessarily reflect its overall progress towards becoming a smart city. Therefore, it is necessary to use the findings of the study as a starting point for analysis and combine them with other relevant factors and information to develop a comprehensive understanding of the smartness of each city.

3.4. Comparative Findings: Following vs. Developing

Although the Following BCCs perform statistically significantly higher compared to the developing cities, the BCCs located in the developing cluster do not necessarily perform lower in all categories compared to those in the following cluster. This information is confirmed by analyzing Figure 3, which shows that the developing cluster is relatively close to the Following BCCs in four of the six study categories (productivity and innovation, livability and wellbeing, sustainability and accessibility, and connectivity and innovation).

The biggest difference between the following and developing BCC clusters is in Governance and planning, as this is the weakest performing area for developing BCCs. One possible reason for the lag in this area could be the low investment in science and technology, which is also reflected in the poor international recognition of universities in the northern and northeastern regions of the country. Developing a smart city strategy could help BCCs involve the community more in their priorities and plan smart city initiatives based on local needs and existing potential.

Socioeconomic progress, safety and security, and sustainable buildings, are the three individual indicators on which developing BCCs perform slightly better than Following cluster BCCs. This happens because Salvador and Fortaleza (two cities in the developing cluster) perform better on the sustainable building indicator than all the BCCs in the Following cluster. In addition, in the areas of socioeconomic progress, safety, and security, the BCC Recife (of the following cluster) performs relatively poorly in the indicators of violent deaths and in the indices of basic education compared to the BCCs of the developing cluster. Finally, all BCCs in the Following cluster are below average on the GDP per capita indicator, while there is one city in the Developing cluster, Manaus, that is above average on this indicator.

4. Discussion and Conclusions

Buying on the smart city’s premises, today, many cities around the globe are developing smart city agendas [79]. Nevertheless, there is a big gap between theory and practice. What smart cities are claimed to be is quite different from what they are in reality [80,81]. Hence, it is imperative to undertake empirical studies to understand what makes a city smart and what the existing performance and hidden potentials of a locality are to transform itself into a smarter one. To address these important questions, this paper adopted a smart city conceptual model and its metrics, previously applied to the Australian city context [3], to evaluate the smart city performances of Brazilian cities.

This paper explored the smart city performance and potential of 27 Brazilian Capital Cities. The clustering of the investigated cities under the Leading, Following, and Developing smart cities categories revealed the following insights into the smart city transformation readiness of Brazilian cities.

First, the leading cluster is mainly made up of southern and southeastern Brazilian capital cities, although some Midwest capital cities have made their way into the leading cluster. The other Midwest capital cities are in the following cluster. North and northeastern capital cities are scattered through the following and developing clusters. This clearly points out a regional divide in Brazil. Cities in the south and Southeast of the country lead the smart city development agenda, while the Midwest is catching up and the north and northeast far behind.

Second, across all investigated cities, the highest category of performance was livability and wellbeing. This is an indication of the unique Brazilian culture, nature, and lifestyle that shape the social sphere of Brazil’s smart cities as a critical asset. Studies [82,83,84] also highlighted the sociocultural assets of Brazilian cities as a core strength of vibrancy in the city.

Third, in all clusters, the weakest performance was in the sustainability and accessibility areas. This is a major challenge for achieving truly smart city development in Brazil. As cities cannot fully achieve their smartness goals without becoming sustainable. In other words, Brazilian cities need to seek ways to develop a sustainable urbanism approach to transform their cities and create new ones that are truly smart and sustainable.

Next, the analysis shed light on the governance and planning, and connectivity and innovation issues and challenges of the investigated Brazilian cities. In terms of connectivity and innovation, only a limited number of Brazilian cities have managed to build a functional innovation ecosystem, such as Florianopolis and Sao Paulo [85,86]. There exists a major socioeconomic disparity between cities that have a sound innovation ecosystem and those that do not. Additionally, while strong creativity potential exists in Brazil, the cities seem to not be tapping into the creative industry opportunities as much as other countries do [87]. In terms of governance and planning, the main issue is the lack of, or limited, good and effective governance. This problem is not only unique to Brazil but is widely common across all developing nations.

Last but not least, the findings revealed that the common characteristics of cities with leading smartness performance are: (a) a strong innovation ecosystem; (b) specific legislation for developing entrepreneurship; (c) training opportunities for skilled labor; and (d) conditions for knowledge-based development and digital transformation offerings and readiness. The smartest Brazilian state capital cities are identified in the following order: (1) Florianópolis; (2) São Paulo; (3) Vitória; (4) Curitiba; (5) Porto Alegre; (6) Brasília; (7) Belo Horizonte; (8) Rio de Janeiro and (9) Cuiabá.

In sum, we echo Desouza et al.’s [88] following considerations for Brazilian cities: “(a) Smart cities which focus only on technology seldom work; (b) Local governments should adopt the role of facilitator; (c) Risks need to be shared with the private sector; (d) Local governments should be open to new innovations and learn from mistakes; (e) Smart cities should focus on inclusivity; (f) Consumption of resources must be considered—particularly regarding the longevity of technological infrastructure; (g) Long-term sustainability is dependent on renewable resources, and; (h) Smart cities require a smart community that is knowledgeable, conscious, forward-thinking, engaged, united and active”.

We also note that this paper has been prepared to address the research question, ‘What are the smartness levels of Brazilian state capital cities?’ It also assists public organizations, including the Brazilian Local, State, and Federal Governments—and many others across the globe—in designing and improving smart city policies for their localities and communities. Although an analysis of clusters may have been performed to identify commonalities and differences between cities, it is important to note that for a city to perform its own analysis, it should be done in conjunction with other data and evaluations from other sources. This is because the smartness of a city is a complex and multidimensional concept that cannot be fully captured by a single set of data or analysis [89]. Therefore, to obtain a comprehensive understanding of the smartness of a city, it is important to consider a wide range of factors and information, including qualitative data, citizen engagement, and other relevant contextual factors. By combining and analyzing multiple sources of data and information, a city can develop a more nuanced understanding of its own strengths and weaknesses and identify opportunities to improve its overall performance and quality of life for its citizens.

Similar to all empirical studies, this one has some limitations that should be noted when interpreting the results. First, data availability may have impacted the study’s decision to adopt the most suitable indicators and instead use proxies. Second, the comparison made by a metric, while useful, may also ignore some of the soft factors cities have, which might be hard to quantify and compare. Third, the interpretation and communication of the study findings might involve some degree of unintentional researcher bias. Lastly, the research reported in this paper utilized a novel approach that has only been applied and tested in Australian and Brazilian smart city contexts (in this and our previous study) so far; hence, application of the same approach to other country contexts will require careful tailoring of the model, particularly indicators and associated datasets. Our prospective research will focus on addressing these limitations and will also undertake cross-country assessments.

We conclude the paper by stating Patrão et al. [90] (p. 1127), where this study contributes to “the development of a single tool, able to overcome the limitations of [smart city assessment approaches], is a quite ambitious task. Efforts should be made to at least develop an assessment tool capable of performing an evaluation track record, with, for instance, automated acquisition of several important indicators and the possibility to compare implementations with different scales”.