1 Introduction

Smart city development is a complex, multidisciplinary problem that requires adaptive resource use and context-aware decision-making practices to enhance human functioning and capabilities while respecting societal and environmental rights and ethics (March, 2019; Townsend, 2013). Although there is no universally accepted definition of smart cities, a broad and inclusive framing of the concept would link and localize various UN Sustainable Development Goals (SDGs): Goal 3 (good health and wellbeing), Goal 4 (quality education and lifelong learning), Goal 7 (affordable and clean energy), Goal 9 (industry, innovation, and infrastructure), Goal 10 (reduced inequalities), Goal 11 (sustainable cities and communities), Goal 13 (climate action), Goal 16 (peace, justice, and strong institutions). To achieve these goals, urban areas must take urgent actions to (a) enhance the health and wellbeing of city residents while ensuring inclusivity in urban life (e.g., through the intelligent design of public spaces, mobility, and transportation) and (b) ensure that urban development and planning contribute positively to resilience and sustainability (e.g., through better disaster management, planning of city logistics, and waste management).

Indeed, the concept of smart cities is also related to the notion of Society 5.0, which was proposed in Japan to form a sustainable societal environment and enhance residents' comfort with enriched technological opportunities, where physical space and cyberspace are integrated in a balanced way through technological advancements such as the Internet of Things (IoT), blockchains, edge computing, and machine learning algorithms (Kiruthika et al., 2024). Another term used for Society 5.0 is human-centric super-smart society. This concept has also impacted Industry 4.0, which primarily focuses on technological advances to form smart infrastructures such as smart transportation, smart buildings, smart factories, and smart healthcare, transforming it into Industry 5.0, where the wellbeing of humans is essential by respecting their working conditions (Coronado et al., 2022). In other words, all the smart infrastructures are established with a human-centric focus, paying attention to human-machine interaction and interfaces. This, in turn, brings the topics of societal values and human wellbeing issues to the forefront, in addition to the economic considerations and efficiency issues of Industry 4.0 (Alter, 2020). It has been well understood that significant technological advances in areas such as transportation, healthcare, communication, education, and manufacturing are insufficient and ineffective for sustainable development without considering human factors. Within the scope of smart cities, it is crucial to integrate the human dimension alongside various technological tools and systems such as IoT, big data, and machine learning algorithms (Komninos and Kakderi, 2019). This necessity brings neurotechnology into play.

In recent years, experts across various fields have acknowledged the significant potential of real-time sensing designs and neuro-inspired control algorithms for local interventions, thereby supporting the development and planning of urban systems that are healthy, inclusive, safe, and resilient. Ancora et al. (2022) highlight substantial advancements in our understanding of brain functioning, with neuroscience expanding its applications across various domains, such as marketing, economics, decision sciences, and educational sciences. This expansion, aided by technological advancements such as wearable devices and software applications, has also given rise to a novel field known as “neurourbanism” which integrates theoretical perspectives and analytical methods inspired by the brain. The aim is to deepen insights into human needs, behaviors, and decision-making processes in urban environments, ultimately enhancing service design and implementation throughout cities in a context-aware manner (Makanadar, 2024).

The study by Gu et al. (2024), for instance, investigates the impact of sidewalk ground murals, focusing on color (warm, cool, or achromatic) and pattern (rectilinear or curvilinear), on mood states and perceived restorativeness, especially under stress. Using a 3 × 2 × 2 mixed design, 112 students were divided into stressed and non-stressed groups and assessed across six mural designs and one uncolored condition. Findings indicate that ground murals enhance mood and perceived restorativeness compared to uncolored sidewalks. Cool colors were most effective in promoting restorativeness, especially for stressed individuals, while warm colors reduced relaxation, and achromatic colors decreased energetic arousal and were least restorative. Pattern type did not enhance mood but curvilinear patterns were perceived as more restorative than rectilinear patterns. The study supports urban design strategies that incorporate systematic use of color and pattern in ground murals to improve mental health. For example, two Superblock initiatives in Barcelona utilized ground murals to create a more inviting environment for pedestrians and cyclists, making the neighborhood more engaging and suitable for community activities (Archdaily, 2020).

Similarly, Mikuni et al. (2024) are interested in whether artistic interventions in urban street environments enhance wellbeing and how these improvements relate to aesthetic evaluations of the environment. After examining three hypotheses, the authors find out that engaging with artistic installations in urban street settings alleviates subjective experiences of anxiety, stress, and negative emotions. Negami et al. (2018) examine how urban design elements such as colorful crosswalks and greenery impact mental wellbeing, sociability, and environmental stewardship. Participants in Vancouver's West End evaluated six sites, three with urban interventions and three without, using a smartphone app to express their emotional reactions. Sites with greenery and colorful designs were linked to higher happiness, trust, and attraction. The findings suggest that urban design changes can enhance city residents' wellbeing and sociability.

In another study, Dimitrov-Discher et al. (2022) assess the impact of green space, air pollution, and noise pollution at the place of residence on neurofunctional activation during social stress. Based on fMRI data from 42 participants, the study revealed that exposure to green spaces correlates with increased parietal and insular activation under stress. In contrast, air pollution is related to diminished activation in these brain regions. These outcomes could provide valuable insights for upcoming research in the developing discipline of “neurourbanism” and highlight the critical role of environmental factors in urban design. The findings suggest that simple urban design changes can enhance city residents' wellbeing and sociability. Neurotechnology, in this context, by interacting with the human brain and nervous system, enables technological systems to become more aligned with human needs. Consequently, this integration has the potential to allow smart cities to achieve both efficiency and human-centricity.

In the context of smart cities, “neurochallenges” involve using principles of cognition and neuroscience to comprehend and anticipate human behavior and requirements within urban environments. Appropriate solutions are then devised and implemented utilizing neurotechnologies to enhance the quality of life and human wellbeing. This smart city vision, depicted in Figure 1, positions itself at the nexus of neuroscience and technology, urban space, and societal interactions. Realizing this vision requires co-producing knowledge toward a hybrid intelligence, whereby education and research, technological innovation, and societal innovation go hand in hand, addressing key focus areas in line with the SDGs.

Figure 1. Smart cities as hybrid intelligence in the nexus of neuroscience/technologies, urban space, society.

It goes without saying that, “going smart” raises significant ethical and privacy concerns that must be carefully managed. Therefore, implementing this vision mandates an interdisciplinary and integrative approach to address the critical issues and concerns of urban development and planning by allowing the consistent and coherent communication of multiple codes and perspectives (Kourtit and Nijkamp, 2012; Castells, 2000) and strong collaboration among governments, industry, academia, and civil society organizations (Breuer et al., 2019). Again, the ultimate goal is not just to improve effectiveness but to ensure that technological advances serve human wellbeing and promote healthier and more sustainable urban planning and design purposes (Pykett et al., 2020).

Against this background and motivated by the potential interplay between societal challenges and emerging technologies, in this paper, we undertake a bibliometric analysis of the literature on smart cities through the lens of neuroscience and neurotechnology. Using data from 2018 to 2022 extracted from the Scopus database, based on a list of terms and themes identified within the scientific community, we provide a comprehensive overview of the research landscape around smart cities to address the following four research gaps: (1) mapping the knowledge structure in the literature around smart cities and exploring emerging topics in neuroscience and neurotechnology as they apply to smart cities; (2) evaluating to what extent technological solutions and advances effectively address societal and environmental challenges in the 21st century; (3) discussing, using key research streams, how to better create synergies and complementarities to contribute to the overarching goals of health, inclusivity, safety, and resilience in urban development; (4) providing insights for strategic planning and future research directions to policymakers, funding agencies, and institutions.

While some earlier studies offer reviews of smart city research, they do not explicitly tackle the questions and issues discussed here. Zhao et al. (2021) provide a systematic review of highly cited smart city literature based on the analysis of 191 publications. Allam and Dhunny (2019) focus in their review on the opportunities and challenges of big data and artificial intelligence and discuss how to govern them so that they are integrated into the urban societal fabric. Ruhlandt (2018) analyzes the relevant body of literature on smart city governance to better understand its different components and the metrics used to measure outcomes. Mora et al. (2019) examine a group of literature reviews and studies to highlight the lack of consensus on selecting an approach to effectively manage smart city development, arguing that this undermines smart city practice and the potential for urban sustainability. Özdemir et al. (2019) conduct a qualitative review of smart city literature, concluding that rapid ICT development has exacerbated developmental gaps within and between urban areas. They emphasize that the social dimension is often neglected, underlining the need for smart city strategies to address both socioeconomic and quality of life dimensions. In this context, Figure 2 illustrates the scope and structure of this paper, emphasizing the unique contributions and perspectives explored in this study.

Figure 2. Purposive areas for neuroscience and neurotechnologies in smart cities.

In Section 2, we present the state-of-the-art in the literature by offering a bibliometric analysis of neurochallenges within the context of smart cities. After explaining the methodology and describing the data sources used for the bibliometric analysis, we first present our findings and insights, initially on a broad basis, and then focus on four specific sub-disciplines (computer science, engineering, social sciences, and environmental sciences). Section 3 illustrates the potential use and practical applications of neurotechnological solutions in the context of smart cities and aims to discuss how better challenge-technology synergies might be built. In particular, we consider the implications of neurotechnology in two purposive areas aligned with the SDGs (represented in blue color in Figure 2), which are (i) human health, wellbeing, and social inclusiveness and (ii) resilience and sustainability. For the former, we focus on neurodesign, followed by a discussion on mobility and access to transportation. For the latter, we investigate disaster and waste management for ecosystem resilience and health.

Then, Section 4 emphasizes some specific areas that require the attention of academic and policy circles, particularly emphasizing three interrelated concerns (represented in pink color in Figure 2) that must be concurrently addressed from a systemic standpoint to ensure that societal goals and technologies are not overlooked or sidelined. These concerns involve acting and establishing (a) institutions promoting data-driven governance, transparency, and accountability to support public decision-making processes in these two purposive areas, necessitating collaboration among multiple stakeholders (b) new educational and research initiatives like NeurotechEU, the European University of Brain and Technology, which represents an alliance of nine European universities focused on neurotechnology (c) regulatory frameworks and ethical codes to ensure the responsible development and application of these technologies. Lastly, Section 5 provides insights on future research directions and concludes the paper.

2 A bibliometric analysis of neurochallenges within the context of smart cities

In this section, we introduce the details of our research, presenting the methodology and initial outcomes. We first explain how we conducted the bibliometric analysis and the sources from which we extracted data. This is followed by the presentation of the co-word analysis conducted on all selected articles. We then shift to a more discipline-focused perspective, performing a co-word analysis with a particular disciplinary emphasis. Finally, we present the critical insights derived from our bibliometric analysis.

2.1 Methodology and data sources

The main goal of our bibliometric study is to reveal the intellectual structure and important themes in the field of research. To achieve this, we initially generated a draft list of terms and refined and augmented them into a finalized list through consultation with experts, as can be seen in Table 1. This represents the first step in our bibliometric methodology. In the second step, we ensured the relevance of our bibliometric analysis to smart city research by narrowing our search to those papers where the “smart cities/city” keyword appeared, along with terms and themes representing Neuroscience and Neurotechnology in relation to Smart Cities (listed in Table 1) or terms and themes characterizing a Smart City concept (listed in Table 2). When one searches these lists with the concept of a “smart city,” the evolution of the idea becomes apparent. It shifts from being primarily about technology to focusing on people, eventually embracing inclusive and participatory governance.

Table 1. Terms and themes representing neuroscience and neurotechnology in relation to smart cities.

Table 2. Terms and themes characterizing smart city concept.

In the third step, we searched the Scopus database, preferring it over the Web of Science due to its greater number of indexed journals. Using the query function, we looked for articles containing these terms in the article title, abstract, or keywords, covering all research institutions for the last 5 years, from 2018 to 2022. For this step, following typical literature review practices, we restricted our search to peer-reviewed journal articles and conference publications, thereby excluding gray literature and books. Furthermore, we only included articles in English. Our search criteria and terms yielded 27,346 peer-reviewed manuscripts. In the final stage, to streamline the data in our bibliometric analysis, we also created a thesaurus and combined or recoded some words following bibliometric analysis guidelines (Donthu et al., 2021), addressing variations in term usage across different publications, such as the singular or plural forms or the use of acronyms. For instance, we made no distinction between “smart city” and “cities” or between “mobile crowdsensing” and “mcs”. This enabled us to generate a reliable list of author keywords in our dataset.

We also analyzed the disciplinary foundations of the bibliometric data at our disposal. Figure 3 illustrates the diversity of fields covered by the literature, each highlighting different aspects of smart cities and reflecting the varied goals and priorities that stakeholders may have.

Figure 3. Bibliometric analysis: share of various disciplines (2018–2022).

The most prominent subject area is computer science, which contains 19,910 documents, accounting for 31.0% of the overall documents. Engineering ranks second with 11,893 documents, constituting 19.7% of the overall number of documents. Therefore, half of the global research in Smart Cities is highly technical. The social sciences discipline ranks third with 6,138 documents (10% of documents). Due to its high relevance, we also studied environmental sciences, which contains 2,433 documents. This distribution signals opportunities for dialogue and collaboration between the dominant disciplines and those with smaller representations. Neuroscience does not appear to be a field dominating the scene presently (appearing within the group of subject areas referred to as Others). Still, it might become prominent, given its potential for new collaborations with other disciplines.

2.2 Co-word analysis for all articles selected

We conducted a co-word analysis, also known as “author keyword co-occurrence”, on the selected articles in the VOSviewer software. This method identifies keywords that frequently appear together in the content of the articles, indicating thematic relationships. Figure 4 presents one such bibliometric analysis result, a general one conducted without disciplinary filtering. While generating images for different periods is always possible, the image below represents the last 5 years, from 2018 to 2022. We used the VOSviewer clustering algorithm (with the counting method fractional counting and the normalization method based on association strength), which allowed us to generate a number of clusters based on topics. In our analysis, it seemed adequately meaningful to interpret the thematic groups within five clusters.

Figure 4. Co-word network visualization (2018–2022).

Each node in this network represents an author-defined keyword, such as “smart city,” “internet of things,” “machine learning,” or “urban planning.” The size of each node represents the frequency of the keyword occurrences. As the node size increases, it indicates higher keyword frequency. Links between nodes imply that corresponding keywords co-occur within articles, with the link thickness representing the strength of co-occurrence. Each color represents a distinct thematic cluster. The nodes and links within a cluster help explain the range of topics (nodes) covered by that theme (cluster) and how those topics (nodes) are interconnected (links) (Donthu et al., 2021). To fully interpret the results of the bibliometric analysis, our team convened to discuss the interpretation of each thematic cluster and the significance of the keywords within publications in that cluster. The following are some preliminary observations drawn from these results.

The red cluster in Figure 4 focuses primarily on sustainability issues, exploring strategies for more sustainable urban development, planning, and governance, which are key driving forces in existing smart city literature (Caragliu et al., 2011). This includes leveraging innovative technologies such as e-government, geographic information systems, and information and communication technologies (ICT) to manage and operate a city more effectively and improve the quality of life. The green cluster centers around using specific digital technologies for addressing particular problems, such as energy consumption and efficiency, smart grids, smart meters, electric vehicles, and edge computing. The blue cluster highlights the “smart city” concept, emphasizing techniques from areas like “machine learning” and “artificial intelligence.” The yellow cluster addresses privacy and security concerns, touching on “authentication” and “smart contracts.” Lastly, the purple cluster may be considered a niche theme, delving into network and gateway protocols like LoraWAN and LPWAN (see Supplementary Table SA for a detailed breakdown of the keywords in each of the five clusters, and their frequency of occurrence).

2.3 Co-word analysis with a disciplinary focus

In recognition of the integral link between smart cities and the vision of context-aware decision-making, we also aimed at establishing stronger connections to specific SDGs, in particular focusing on SDG3 (good health and wellbeing), SG7 (affordable and clean energy), SDG9 (industry, innovation, and infrastructure), SDG10 (reduced inequalities), SDG11 (sustainable cities and communities), SDG13 (action on climate change) and SDG16 (peace, justice, strong institutions). To this end, we narrowed our search to papers within four primary disciplines: computer science (relevant to SDGs 3, 9, and 11), engineering (relevant to SDGs 7, 9, and 11), social sciences (relevant to SDGs 3, 10, and 16), and environmental sciences (relevant to SDGs 11 and 13).

When we limit the analysis to certain subject areas or subsets of disciplines, more specific thematic groups, or clusters, emerge. This is because each discipline—Computer Science, Engineering, Social Sciences, and Environmental Sciences—has its unique perspective on smart cities. The themes identified from the bibliometric analysis and how they are interpreted in these four disciplines give us which keywords are most frequently used in each field and how many are shared across disciplines. Each figure also provides insight into the meanings of the clusters. While some clusters in these figures seem more goal-oriented or challenge-driven, others are techniques, methods, or technologies. For each subject area, the top 15 keywords were classified as challenges or technologies and presented as a table to support the maps. As such, the most frequently addressed technologies and challenges were underlined. Indeed, this classification highlights the dominant characteristics of the field, as technology-dominant or challenge-driven and, in a way that supports the maps. Understanding key/notable technologies and challenges facilitates discussions about missing links and potential disciplinary partnerships.

2.3.1 Bibliometric analysis for computer science

The analysis under computer science was based on 19,910 documents (including peer-reviewed articles and conference papers) and 38,189 author keywords. In Figure 5, we visualize the top 100 keywords with the highest occurrence (the minimum number of author keyword occurrences being 66) under five clusters. This gives us a general overview of the main topics that authors explore in computer science within the context of smart cities. The blue cluster reveals that a significant portion of research in computer science, as expected, relates to the “Internet of things” and techniques such as machine learning, deep learning, image processing, and classification. The red cluster focuses on digital technologies and their infrastructure, encompassing aspects such as cloud computing, wireless sensor networks, edge computing, fog computing, 5G, and the internet of vehicles. The green cluster corresponds to research on smart cities' practical concerns and challenges. Specifically, it represents areas where digital technologies and AI techniques from the blue and red clusters can be applied, such as public transport, smart mobility, information and communication (ICT) technology, e-government, smart governance, smart community, and augmented reality. The yellow cluster emphasizes privacy and cybersecurity concerns. Lastly, the fifth cluster, indicated in purple, centers around a niche topic and explores issues around gateway protocols (see Supplementary Table SB for a detailed breakdown of the keywords in each of the five clusters, along with their frequency of occurrence). Overall, computer science in smart city research seems a technology-dominant subject area, not challenge-driven and not so much integrated into environmental and societal goals.

Figure 5. Co-word network visualization with Computer Science focus (2018–2022).

2.3.2 Bibliometric analysis for engineering

The analysis for engineering was conducted on 11,893 documents (peer-reviewed articles and conference papers), and over 24,695 author keywords. In Figure 6, we then visualized the top 100 keywords, with the highest occurrence (the maximum number of keyword occurrences being 3,957 and the minimum being 41) falling into five clusters. The yellow cluster with “smart city” as its central node signifies some primary challenges for engineering encompassing sustainability, urban planning, mobility, resilience, and governance issues. The red cluster centered around the “internet of things” incorporates technologies such as object detection and techniques like deep learning, machine learning, and deep neural networks. These are deployed for specific purposes, such as in smart homes, smart parking, and classification, all represented within the same cluster. Though diverse, the green cluster contains many innovative ICT infrastructures and uses, such as 5G, fog computing, edge computing, blockchain, and the internet of vehicles. The blue cluster focuses on energy-related topics, such as energy efficiency, renewable energy, smart grid, smart meters, and electric vehicles. The purple cluster features a large central node that connects the “artificial intelligence,” to cyber security, and information models and systems (see Supplementary Table SC for a detailed breakdown of the keywords in each of the five clusters, along with their frequency of occurrence). While the engineering subject area in smart city research is also technology-dominant, the relatively frequent occurrence of sustainability keywords underlines its problem-solving-oriented character.

Figure 6. Co-word network visualization with Engineering focus (2018–2022).

2.3.3 Bibliometric analysis for social sciences

In social sciences, we analyzed 6138 documents (peer-reviewed articles and conference papers) encompassing 15,693 author keywords. In Figure 7, we visualize the top 100 keywords with the highest occurrence (the maximum number of keyword occurrences being 2,364 and the minimum 21) in five clusters. Three large clusters (red, green, and blue) with many keywords and two relatively smaller groups stand out. The green group considers the core of smart cities to be sustainable urban development, urban planning, governance, participation, and quality of life. These ideas are presumably made workable by connecting them to the other large cluster, the red areas, where techniques such as the internet of things, machine learning, and AI are used. The blue cluster in the social sciences covers issues surrounding intelligent transportation systems linked to smart mobility and public transport. The yellow cluster then examines the uses of technologies such as data visualization, virtual reality, and augmented reality. The purple cluster is small, featuring social media and smart tourism as niche topics (see Supplementary Table SD for a detailed breakdown of the keywords in each of the five clusters, and their frequency of occurrence). In the social sciences, while there is more emphasis on challenges compared to computer science and engineering, the fact that even in this subject area, smart city research emerges as a technology-dominant subject area underlines the need to create better synergies between technologies and challenges.

Figure 7. Co-word network visualization with Social Sciences focus (2018–2022).

2.3.4 Bibliometric analysis for environmental sciences

The analysis of 2,433 documents (peer-reviewed articles and conference papers) within environmental sciences gives us 6,288 author keywords. Figure 8 reflects the top 100 keywords with the highest occurrence (the minimum number of keyword occurrences being 8). The red cluster of the network represents challenges such as eco-cities and sustainable urban development from an environmental science perspective. It includes issues such as air quality, climate change, energy transition, urban transformation, and the circular economy. With sustainability and quality of life concerns, the green cluster focuses on planning and decision sciences, where optimization, indicators, and infrastructure policies play a role. The blue cluster is centered mainly around internet of things and machine learning, representing computer science solutions along with sensors, data visualization, and image processing. The purple and yellow clusters are small niche areas around energy and security (see Supplementary Table SE for a detailed breakdown of the keywords in each of the five clusters, along with their frequency of occurrence). While being still technology-dominant, environmental sciences in smart city research are where there exist signs of better synergies and integration between challenges and technologies.

Figure 8. Co-word network visualization with Environmental Sciences focus (2018–2022).

2.4 Key insights from the bibliometric analysis

Our analysis presents compelling evidence that smart city development encompasses diverse dimensions, necessitating the application of multiple perspectives to address the societal and environmental challenges in urban settings. Within our bibliometric analysis, numerous author keywords consistently emerged, signifying recurring themes and shared terminologies across disciplines at the intersection of neuroscience, technology, urban space, and society. A clear trend is that certain key technologies dominate smart city research across all subject areas: Internet of Things (IoT), AI-based techniques/technologies (including machine learning, artificial intelligence, and data mining), and information/network technologies (such as wireless sensor networks, cloud computing, edge computing, blockchain, smart grids, GIS, and mobile applications). These technologies stand as humanity's core infrastructures, and they are relevant for smart cities as well and will surely play a fundamental role in shaping urban processes. In contrast, core challenges aligning with SDGs, such as urban planning, sustainability, e-government, climate change, resilience, healthcare, smart mobility and intelligent transport systems, energy efficiency, air and water quality, and smart homes, are discussed less frequently. This disparity indicates a fragmentation in the interface between challenges and technologies. Therefore, creating better synergies at this intersection seems crucial to ensure that technological advancements address urban challenges effectively and in alignment with the SDGs. Naturally, there are variations in core focuses among disciplines, reflecting researchers' and stakeholders' diverse goals and priorities. For example, the computer science perspective predominantly revolves around the role of technological solutions and ICT infrastructures. Technological solutions and problem-solving approaches primarily guide the engineering perspective. The social sciences perspective emphasizes theoretical and conceptual frameworks that shape urban development and address organizational concerns such as e-governance and citizen participation. Yet, the coherence between challenges and technologies in the subject area of social sciences is remarkably and unexpectedly lacking, as, above all, this is also a technology-dominant subject area. Key technologies studied globally in social sciences are identical to those in the technical sciences of computer science and engineering, i.e., IoT and AI-based technologies (machine learning, artificial intelligence, deep learning). The environmental perspective integrates high-tech solutions with green interventions to achieve environmental goals. Nonetheless, a more integrated approach is needed to better align technologies with the pressing societal and environmental challenges of our time.

One potential risk in smart city development is the possibility of incoherent and inadequate implementation or a lack of coordination in addressing common urban challenges. The overall complexity of urban development poses a danger of prioritizing short-term gains through opaque technological solutions. Seeing the big picture across disciplines holds significant value for academics and policymakers, providing them with insights to establish synergies and better complementarities. Indeed, this brings us to the importance of transdisciplinarity, a process that is inherently complex and challenging. It was originally defined as the coordination of all disciplines and interdisciplines within an education/innovation system based on some common objectives to deliver on the purpose of societal continuous self-renewal (Jantsch, 1970). This critical concept and its implications at the interface of science and society will be explored further in Section 4.2.

It is important to note that this analysis is primarily quantitative and focuses on the frequency of keywords, which may not capture the full depth of the research landscape. A more nuanced question arises here: Do the outcomes of the most frequently studied three subject areas actually support the hybrid intelligence-based co-production of knowledge? Or do they reflect a technology-dominant approach conducted by technical experts, with results published in social sciences journals? A qualitative and systematic study that examines the content and context of the documents would complement this analysis and provide a richer understanding. Additionally, considering the temporal aspect of the keywords and documents could offer a more detailed understanding of emerging trends and shifting research priorities over time.

The acknowledgment of noticeable developments and what has been achieved so far, as well as the identification of the critical aspects that have been overlooked but need to be developed enable researchers and policymakers to be aware of potential disciplinary bridges. This comprehension is essential for formulating cohesive and integrated strategies and advancing smart city initiatives in alignment with SDGs.

Building upon these insights, Section 3 of our study presents specific perspectives—from various fields of study—regarding the implications of neurotechnology in smart cities. These perspectives encompass the associated challenges and neurotechnological solutions. Any unresolved and open issues in this context will then be discussed in Section 4.

3 Perspectives on implications of neurotechnology in smart cities

The discussion in this section is structured in two parts. The first part highlights specific challenges and corresponding neurotechnologies in smart cities aimed at enhancing human health, wellbeing, and social inclusiveness. The second part discusses, in a similar vein, neurotechnologies that contribute to the resilience and sustainability of smart cities, in particular, through improved disaster management and waste management.

3.1 Neurotechnologies in smart cities for human health and wellbeing 3.1.1 Challenges in the intelligent design of public space, mobility, and transportation

Modern urban life is very demanding in terms of economic and physical resources. Moreover, more crowded and complex city life increases the time pressure on the urban citizen. It has been shown that the primary source of the increased stress in everyday life is the mismatch between these demands and the individual's available resources. The European Environment Agency describes this as “urban stress” (EEA, 2020). Although some stress can be beneficial for increased alertness, athletic ability, or better focusing during an exam, it is well known that long-term urban stress can be harmful to the human body. Emotion regulation techniques such as meditation, yoga, and walking outdoors may help to reduce stress. However, urban lifestyles such as long working hours and commutes prevent the most well-known emotion regulation techniques.

Today, designers need a better understanding of the psychological and emotional impacts and threats of changing physical and social landscapes. Neuroaesthetics, which studies the neural basis of aesthetic experiences, can play a crucial role in designing public spaces that elicit positive emotional responses. Neuroarchitecture aims to create buildings that are capable of adapting to human emotions and cognitive states by modifying light and sound to influence sensory experiences (Makanadar, 2024). Incorporating natural elements such as green spaces and natural lighting can enhance mood, promoting a sense of calm and wellbeing. In contrast, exposure to unnatural light sources can adversely impact mental health, intensifying feelings of anxiety and depression. Understanding and regulating these emotional shifts can be enriched by applying neurotechnology. Consequently, two significant challenges emerge here: First, sensing stress, fatigue, anger, and other potentially harmful emotional or physical states, and second, facilitating emotion regulation in urban environments.

A third challenge relates to another pivotal component of smart cities: the sensors that can be statically deployed throughout urban areas. They enable monitoring of key parameters about the city, such as noise, air pollution, traffic levels, or availability of parking places. Sensors are mostly used as data loggers, and raw data from the nodes can be transmitted, mostly through wireless interfaces, to a server or cloud where further data processing occurs. Deploying the sensors around a city and in large quantities presents the challenge of installing a complex and costly infrastructure. In addition, questions such as how to power the devices and maintain/replace them in case of failures arise. Cities already face the challenge of replacing old infrastructures, like water distribution and sewage systems, and making the introduction of new structures is even more daunting. Thus, easy-to-deploy and easy-to-maintain sensing systems are required.

Finally, as the social economy develops, there is a rapid increase in both the global population and private automobiles. The increasing demand for mobility burdens the existing transportation infrastructure, resulting in increased road accidents, congestion, and inefficient use of energy resources, consequently causing environmental, economic, and societal impairments. Rather than building more roads, there is a need for more efficient use of currently available means of transportation through the development of intelligent transportation and logistics in addition to regulating traffic flow, throughput, and safety.

3.1.2 Technologies and solutions 3.1.2.1 On physiological sensing-based emotion regulation with wearables across smart cities

Since smart watches, smart wristbands, and other smart wearables such as rings have become pervasive and their sensing modalities have increased, sensing emotional states has become feasible using wearables (Can et al., 2019). There are observable physiological changes in the human body, such as a reaction to stress. Heart Rate Variation (HRV) and Electro-Dermal Activity (EDA) are well-known stress indicators. Most wearables have Photoplethysmography (PPG) sensors that can measure HRV, which changes quickly in a stressful situation. In addition, some wearables have Galvanic Skin Response sensors for detecting EDA. Although an EDA reaction is slower than an HRV response, it has been shown that multimodality improves the emotion/fatigue sensing capability (Can and Ersoy, 2023).

The ubiquity of wearables is not the only reason for the improvements in emotion sensing. The sensors create multiple streams of time series data. Recent advances in machine learning (ML) techniques, such as deep learning and faster ML hardware, have enabled more accurate models for emotion sensing. If the data is not extensive, an acceptable sensing accuracy can be achieved using classical ML techniques such as support vector machines and decision trees. With the availability of larger physiological data from wearable sensors, better overall emotion-sensing performance can be obtained using deep learning techniques. However, there are increased privacy concerns about health-related data. New techniques, such as federated learning also facilitate privacy-aware emotion sensing with better accuracy (Can and Ersoy, 2021). Then, the issue becomes acting to regulate emotions after detecting the emotional states while living in urban environments, such as working in the office or commuting. One EU-funded H2020 research project targeting these issues was Affectech: Personal Technologies for Affective Health (https://cordis.europa.eu/project/id/722022). In that project, interactive tools and wearable actuators with vibrotactile and temperature-changing elements were recommended for emotion regulation in urban environments. In that context, there is much room for improvement in urban spaces. Nudges or subtle environmental cues that encourage positive behavior can also be incorporated into the design of public spaces to promote healthy habits, such as walking or socializing. In addition, smart layouts and sound systems can further enhance the safety and accessibility of public spaces.

3.1.2.2 On crowd-sensing with mobile, wearable, and iot devices across smart cities

Crowd-sourced or participatory sensing enables collecting ambient personal data, particularly from the devices carried/worn/owned by people. Smartphones, wearables, and IoT devices are the ideal platforms for sensing applications with the integrated rich set of sensors, their ubiquity, ease of use, support for mobility, and wireless interfaces (Ìncel and Özgövde, 2018). Consider the sensors available in today's smartphones: microphone, camera, GPS, light, accelerometers, gyroscope, compass, GPS, pressure, temperature, magnetic field, humidity, and heart rate, among others. The sensor set can be extended by those not integrated into the device but can be connected via wireless interfaces, such as gas and occupancy sensors. They can also communicate and complement existing, static sensing infrastructures. These devices gather ambient information and collect data about the context, such as activities, location, emotional states, and health status, of the users carrying them. Users can also be sensors by reporting their observations (Berntzen et al., 2016) through app interfaces. Programming such devices is easy, and the applications can be delivered to large populations worldwide through app stores. Hence, they enable global mobile sensor networks.

Typical applications of crowd-sourced systems in smart cities include environmental monitoring, traffic and transportation monitoring, and monitoring road conditions such as detecting road obstacles. Normally, these devices are personal, and the sensing applications are for personal purposes, such as step counters, fitness tracking, and wellbeing monitoring. However, context recognition of crowds and communities rather than individuals enables a new set of application domains in urban planning and transportation. By monitoring what is happening in urban environments, for example, it will be possible to discern those regions' typical or routine actions: which regions are used and for which activities? By analyzing common activities and transportation modes in urban settings, suitable zones and times for cycling, for instance, can be marked on maps. Alternatively, extraordinary situations can be monitored, and possible emergency or disaster situations can be detected: what if many users start running in an area where people usually sit or walk? In addition, people's transportation modes, such as cycling, car/bus, and train/metro, can be tracked, thus laying the groundwork for applications such as extracting transportation-type maps of cities.

3.1.2.3 On cooperative intelligent transportation systems

In traffic, automated vehicles need to interact with other vehicles, bicycles, pedestrians, and other road users, as well as with IoT services (including via Road-Side-Units), over Dedicated Short-Range Communication or 5G-V2X networking (Zhang et al., 2023). Information exchange between Connected Automated Vehicles through Vehicle-to-Vehicle and Vehicle-to-Infrastructure wireless communication enables automated vehicles to cooperate and interact within their (socio)-cyber-physical environments. Connected automated vehicles, forming the so-called Internet of Vehicles, are predicted to transform transportation and urban life (Loke, 2019). Other automated vehicles, such as self-driving wheelchairs and self-driving motorcycles, are also being developed. In addition, cooperative Intelligent Transport Systems (e.g., cooperative driving) is an active area of research.

As the logistics industry struggles with the demand for faster, more energy-efficient, and autonomous delivery methods, the focus on last-mile delivery has intensified. Cooperative delivery mission planning, and the utilization of Unmanned Aerial Vehicles (UAVs) and Automated Ground Vehicles (AGVs) emerge as a promising alternative to conventional methods for last-mile delivery. The synchronization of UAVs with AGVs provides the advantage of delivering hard-to-reach places. These heterogeneous mobile platforms also offer a range of strategic alternatives in terms of operational time, delivery time, and fuel consumption. Moreover, since the UAVs cruise through the air, the delivery is independent of roads, traffic lights, speed limits, and pedestrian crossings. This aerial freedom is particularly beneficial in urban settings where numerous external factors can impede ground deliveries.

Meanwhile, in today's traffic, the limitations of human perception of traffic conditions and response times define the boundaries of safe inter-vehicle distances. Erroneous human driving behaviors can trigger traffic flow instabilities, which result in the so-called shockwaves. For instance, in dense traffic, an overreaction by one driver to a momentary disturbance—like a preceding vehicle's slight deceleration—can set off a ripple effect. Such disruptions might bring the traffic to a full stop, kilometers away from the original source, causing traffic jams for no apparent reason. In this respect, addressing these disturbances across the vehicle string, an aspect encapsulated by string stability, is an essential requirement for vehicle platooning. Fortunately, wireless information exchange between vehicles provides the means for overcoming sensory limitations of human or Adaptive Cruise Control (ACC) operated vehicles and, therefore, can significantly enhance the traffic flow, especially on highways.

Cooperative Adaptive Cruise Control (CACC) is an advanced system designed to regulate inter-vehicle distances, enhancing traffic flow stability and throughput. As an extension of ACCs, CACC achieves superior performance by leveraging wireless information exchange between vehicles through Dedicated Short-Range Communication