AI-enabled transition to smart European cities

Acta Polytechnica CTU Proceedings 46:85–93, 2024 © 2024 The Author(s). Licensed under a CC-BY 4.0 licence https://doi.org/10.14311/APP.2024.46.0085 Published by the Czech Technical University in Prague AI-ENABLED TRANSITION TO SMART EUROPEAN CITIES Noor Marji∗ , Michal Kohout, Lijun Chen, Gülbahar Emir Isik, Akshatha Ravi Kumar Czech Technical University in Prague, Faculty of Architecture, Thákurova 9, 160 00 Prague 6, Czech Republic ∗ corresponding author: noor.marji@fa.cvut.cz Abstract. Smart cities continue to be discussed throughout Europe as a result of the continent’s rising urbanization and the need for sustainable development. Artificial intelligence (AI) has the potential to significantly promote this shift by assisting cities in becoming more effective, sustainable, and receptive to the requirements of their residents. The goal of this study is to examine the potential and difficulties of AI in urban development and present a framework for incorporating AI into city planning and management in European cities. This is done by analyzing case study examples from European cities and examining primary and secondary data sources, with the aim of providing a comprehensive framework for the sustainable integration of AI systems. This study presents a set of ethical and inclusive AI criteria, such as transparency, inclusion, and accountability, to enable responsible AI research and implementation. It continues by emphasizing the need for efficient AI integration in smart cities and pushing for a holistic AI-enabled transition to inclusive and sustainable smart cities. Keywords: Smart cities, artificial intelligence, Europe, urban planning, transition. 1. Introduction Considering these gaps, this paper examines the Rapid urbanization and rising demands for sustain- potential and difficulties of implementing AI in ur- able development have given smart cities prominence ban development and presents a framework for incor- across Europe [1, 2]. As cities grow larger and more porating AI into city planning and management in complex, innovative solutions must be found to European cities, which is the planned output of this assist cities in using resources more efficiently while research. improving inhabitants’ quality of life and encouraging sustainable development [2]. Artificial intelligence 2. Methodology (AI) has emerged as a possible game-changer in This research incorporates a qualitative data collection this area, providing new avenues for data-driven and analysis methodology and outlines a systematic decision-making, urban system optimization, public approach to investigate the potential and challenges of participation, and citizen engagement [3, 4]. AI in urban development and the subsequent integra- Integrating AI technology in urban planning and tion into smart European cities. It covers literature development could yield several advantages, such as review, case studies, framework development, ethi- enhanced efficiency, lower costs, improved public ser- cal considerations, data analysis, and integration of vices, and citizen involvement [2, 4]. AI’s ability to re- findings to drive informed recommendations. spond to residents’ needs while supporting sustainable (a) Literature Review and Data Collection growth and mitigating environmental impact makes AI particularly effective at guiding smart cities [2, 5]. • Primary and Secondary Sources: review of pri- However, using AI poses certain hurdles such as ethi- mary literature such as academic journals, con- cal concerns regarding algorithm bias as well as data ference papers and reports pertaining to AI ap- privacy considerations [5, 6]. plications in urban development and smart cities The identified knowledge gaps in the field of AI- within European contexts. enabled transition to smart European cities are (b) Case Studies twofold. Firstly, there is a lack of a comprehensive • Criteria for Selection: The cities were selected framework that addresses the sustainable integration based on their use of artificial intelligence in ur- of AI systems into city planning and management, ban development, considering factors like geo- particularly suited for the diverse urbanization chal- graphic location, population size, economic pro- lenges faced nowadays by European cities [2]. Such file and technological infrastructure. a framework should encompass ethical considerations and inclusivity criteria. Secondly, existing literature (c) Framework Development often lacks collective case study analyses from various • Review of Existing Frameworks: Comprehen- European cities, essential for understanding diverse sive evaluation of existing AI-driven urban de- contexts, challenges, and successful AI adoption mod- velopment frameworks in order to identify gaps, els [3, 4]. strengths and weaknesses. 85 N. Marji, M. Kohout, L. Chen et al. Acta Polytechnica CTU Proceedings (d) Ethical Criteria and Guidelines growth, land use planning issues along with various • Exploration of Existing Ethics for AI Implementa- other vital variables [3, 6]. tion, with Emphasis on Transparency, Inclusivity, Predictive modeling is one of the core applications and Accountability. for AI in urban planning, helping forecast population shifts, fluctuations in energy demand, and transporta- • Customization for European Context: Adjusting tion congestion based on historical data [12–15]. These ethical criteria so they fit with European cultural, predictions can then inform planning decisions as well legal, and societal contexts. as investment decisions [2, 3]. Optimizing land use (e) Analysis and Evaluation and transportation systems also uses AI extensively: • Qualitative Analysis: Examining qualitative data AI may discover more effective methods to utilize ur- in the form of case study insights to extract key ban spaces while carrying people and products, while themes and perspectives related to AI-enabled simultaneously improving traffic flow, decreasing con- smart city development. gestion levels, and expanding public transportation services [13–15]. (f) Integration and Recommendations AI can also be leveraged to track and manage ur- • Synthesis of Findings: From our research results, ban systems in real-time. By collecting and analyzing we propose an AI integration framework in Euro- data from sensors and other sources, it provides real- pean cities that prioritizes responsible and inclu- time feedback on operations of urban systems like sive AI practices. energy usage, air quality management, and waste disposal [6]. With this real-time feedback at their dis- 3. Urban planning and sustainable posal, planners, stakeholders, and authorities can iden- cities in Europe tify new development opportunities or alter system performance according to changing requirements [2]. In the 1980s, the Brundtland Report [7] first pro- Lastly, AI may also help encourage citizen partici- posed sustainable development as a concept; defined pation in urban planning through providing access as development that meets present generation needs to data and information with AI assessments of pub- without undermining future ones’ abilities to meet lic input/preferences assessments to facilitate citizen them [7], its concept has since become the cornerstone involvement [16]. for European projects promoting such efforts. The European Union (EU) is a leader in supporting sustainable development through policies and pro- 4.1. Key features of AI-enabled urban grams, like its 2013 Environment Action Program planning which established an environmental action frame- Four key features characterize AI-enabled urban plan- work including sustainable cities as an essential focus ning in smart cities, which are data-driven decision- area [8]. Urban Agenda 2016 also seeks to enhance making and predictive modeling, optimization of land sustainable urban development while simultaneously use and transportation systems, use of sensor data for raising the quality of life across European cities [1]. real-time monitoring and feedback, and integration of Smart cities have recently gained attention as an citizen input and feedback into the planning process, innovative approach to urbanization and sustainable see Figure 1. development issues. A smart city employs technol- ogy and data to enhance inhabitants’ quality of life, 4.1.1. Data-driven decision-making and improve urban services and promote sustainable de- predictive modeling velopment [3, 4]. AI technology plays a pivotal role Urban planners may use AI for data-driven decision- in smart city design by offering new avenues for data- making and predictive modeling. Such evaluation driven decision-making and system optimization [2–5]. can involve collecting massive amounts of information from sensors, social media posts, and public records 4. Artificial Intelligence (AI) in before using an algorithm to find patterns, trends, urban planning and connections human planners might miss; helping AI has the potential to revolutionize urban planning to gain greater insights into complex urban systems by offering new tools for data analysis, modeling, and such as transportation management systems [17, 18]. decision-making. AI utilizes algorithms and machine AI can also predict future urban development trends learning techniques to assess massive datasets in or- by studying historical data and discovering patterns, der to make predictions or suggestions based on pat- providing useful forecasting information such as public terns found within them [9–11]. AI can assess various transit demand or infrastructure expenditure planning data types relevant to urban planning - demographic requirements. Predictive modeling also allows urban data, transportation data, energy consumption infor- planners to anticipate possible problems before they mation as well as environmental considerations [10, 11]. occur - for instance, flood or other natural disaster pre- Through analysis of such data, AI helps urban plan- dictions inform the creation of robust infrastructures ners make more educated decisions related to urban for future development [19]. 86 vol. 46/2024 AI-enabled transition to smart European cities Figure 1. Key features of AI-enabled urban planning for smart city transitions. Predictive modeling and data-driven decision- the quality of life. making can improve urban planning precision and effectiveness, helping planners use AI to make more 4.1.3. Use of sensor data for real-time informed decisions based on data rather than intu- monitoring and feedback ition or preconceptions. Unfortunately, AI algorithms Utilizing sensor data requires placing sensors through- may reflect any biases present in training data which out a city in order to collect information on air quality, results in decisions being inequitable or unfair [20, 21]; noise levels, temperature levels, and energy use [23]. an over-reliance on predictive modeling may result By collecting and analyzing real-time data, urban in rigidity within the planning process as planners planners can gain insight into the current state of become overly dependent on previous data without cities and make more informed decisions regarding re- taking into account new patterns or situations chang- source allocation and urban growth [24]. Sensors may ing from what was originally planned. be utilized to detect levels of air pollution to guide the deployment of air purifiers or mitigation measures, as 4.1.2. Optimization of land use and well as detect energy waste within buildings to provide transportation systems opportunities for improvement [25]. AI algorithms can analyze traffic data to use its results Sensor data can provide municipal planners with to maximize traffic flow and avoid congestion. Traffic useful data as well as real-time feedback from citi- lights, for example, can be automatically modified zens; providing real-time traffic and public transport based on real-time traffic conditions to reduce emis- updates can aid residents in making decisions about sions and wait times [22]. Furthermore, AI may also their daily commute [26], while air quality monitoring be employed in optimizing public transit systems by allows consumers to reconsider engaging in outdoor ac- analyzing passenger demand information such as route tivities in areas which could harm their health [27]. By performance statistics or vehicle availability data in giving individuals real-time updates of how their city order to enhance service efficiencies while reducing is being developed they may become more involved costs [23, 24]. in the decision-making process as they advocate for Urban planners can make more informed decisions policies reflecting their beliefs and interests [27]. regarding land use and urban development projects by uncovering underutilized sections of cities or locations with significant development potential by studying 4.1.4. Integration of citizen input and data such as population density, land use patterns, and feedback into the planning process zoning rules. Optimized transportation system designs Historically, urban planning was typically carried out have also proven their ability to ease traffic congestion, from the top down with city officials making choices enhance air quality and make cities more livable [22]. without community consultation or involvement. AI Studies have shown that AI-aided planning for trans- technologies provide new avenues of participation for portation networks could lower costs while expanding citizens, enabling engagement through public partic- accessibility for marginalized groups [24–26]. ipation opportunities throughout various planning Optimizing transportation systems, however, could processes. increase reliance on vehicles and individual forms of Digital platforms, online forums, social media plat- transport, contributing to air pollution and climate forms, and mobile applications provide urban planners change. Meanwhile, optimizing land use could result with opportunities to gather citizen opinions on var- in the displacement of communities or the removal of ious urban planning matters such as transportation, green spaces, both of which would negatively impact housing, and public services. Their feedback can then 87 N. Marji, M. Kohout, L. Chen et al. Acta Polytechnica CTU Proceedings Figure 2. Framework for AI integration towards smart city transition. be studied and integrated into planning processes al- learning algorithms, OmaStadi understands citizen lowing urban planners to design more responsive plans inquiries quickly while offering tailored answers – ul- suited to citizen feedback [28, 29]. timately increasing transparency, accountability, as Urban planners can gain real-time insight into peo- well as citizen involvement [28, 29]. ple’s issues and ambitions using machine learning Amsterdam, Netherlands: Amsterdam has imple- algorithms applied to social media data analysis, pro- mented an artificial intelligence-powered energy man- viding real-time insight that allows for enhanced plan- agement system that utilizes data collected via smart ning decisions as well as greater citizen participation. meters and building sensors to optimize energy use Furthermore, immersive planning experiences can be while simultaneously decreasing greenhouse gas emis- created using visualization technologies such as virtual sions. AI algorithms use analysis results in energy- reality or augmented reality technologies for maximum saving actions like changing building temperatures or engagement with the proposed environmental changes. lighting levels resulting in reduced costs and greater One challenge lies in making sure the feedback re- sustainability [29, 31]. flects all aspects of community opinions rather than Copenhagen, Denmark: Copenhagen has recently simply responding to vocal minority interests, while si- implemented an AI-powered trash collection system multaneously using it meaningfully as input into plan- that utilizes sensor data and artificial intelligence ning decisions versus simply giving the appearance of algorithms to optimize waste collection routes and citizen involvement without altering decision-making costs by anticipating waste generation patterns, op- processes. timizing collection schedules, and cutting collection costs [29, 32, 33]. 5. Case studies: AI in European These case studies illustrate how AI can be utilized smart cities to promote sustainable urban development and en- hance citizen services in smart cities. However, it’s A number of European cities – Barcelona, Amster- crucial that AI solutions be created and used responsi- dam, and Copenhagen among them – have launched bly and inclusively to address potential concerns while smart city programs covering topics as varied as energy upholding ethical AI principles. efficiency, transportation needs management, waste removal management, and citizen involvement [4]. Barcelona, Spain: Barcelona has successfully im- 6. Framework for AI integration plemented an AI-powered traffic control system that towards smart cities uses real-time data to optimize traffic flow and alle- To support the responsible development and applica- viate congestion. This system collects information tion of AI in smart cities, a framework for integrating from traffic sensors, GPS devices, and public transit AI into municipal planning and management is nec- systems using AI algorithms that predict patterns to essary, see Figure 2. The following is a suggested alter lights accordingly; as a result of which travel framework by the authors for ethical and inclusive AI times and air quality have both seen significant im- guidelines, based on the previously conducted analysis provements [30]. and review: Helsinki, Finland: “OmaStadi” is an AI-powered Identify Opportunities: The initial step should in- virtual assistant designed to increase citizen participa- volve identifying areas where AI may help enhance tion in city planning and decision-making processes. urban systems and citizen services. For this to work Utilizing natural language processing and machine effectively, collaboration among municipal officials, 88 vol. 46/2024 AI-enabled transition to smart European cities citizens, technology specialists, and experts must oc- serious ramifications for underrepresented groups as it cur to find opportunities that fit with city values and will exacerbate existing imbalances [21, 35]. Studies interests. have demonstrated how using biased datasets with Assess Risks: Once opportunities have been iden- AI algorithms can lead to discriminatory outcomes tified, the next step should be evaluating any risks across several areas, including criminal justice, em- posed by AI deployment. Threats to data privacy, ployment practices, and lending practices [21, 35]. It ethics, and social justice need to be assessed before is imperative that training data is carefully curated AI implementation begins; residents and stakeholders and assessed to reduce biases and promote fairness should be included as much as possible during this within deployed AI systems [36, 37]. AI’s use in life- process to address any potential concerns. altering decisions such as housing, employment, and Develop Guidelines for Ethical and Inclusive AI : social services raises ethical concerns. AI algorithms Once potential risks have been assessed, guidelines may make judgments on age, color, gender, or socioe- must be created that address ethical and inclusive conomic status that lead to discrimination or unfair AI rules. Recommendations such as those set forth treatment [38]. Therefore it is crucial that these al- by international bodies (for example, the European gorithms’ decision-making processes remain visible, Commission Ethics Recommendations for Trustworthy responsible, and equitable [39]. AI) should serve as inspiration when crafting such Data privacy is also an ethical concern of AI sys- guidelines; principles like transparency, accountability, tems. AI relies heavily on collecting, using, and storing and inclusivity should all be included within such personal data; any changes could have profound ram- recommendations. ifications on individuals’ rights to privacy [40]. In Implement AI Solutions: Once ethical and inclusive order to uphold individuals’ privacy rights effectively, AI principles have been established, the next step specific criteria for data collection, utilization, and should be implementing solutions that comply with storage must be developed [41, 42]. them. This involves selecting appropriate algorithms, AI system development presents many risks related data sources, and governance frameworks to guarantee to algorithmic bias, which may arise at various points reliable, transparent, and accountable AI services. during AI’s development process – for instance during Evaluate and Adapt: Finally, AI solutions must data collection, algorithm design, or model training be constantly evaluated and modified so they remain processes. Training data used for an AI system’s productive while adhering to ethical and inclusive construction should ideally not contain bias if possi- AI norms. This involves monitoring their effects on ble [35]. Keeping in mind that implicit bias can only citizens and civic processes as well as, when needed, be avoided with sufficient variety and representation revising algorithms or governance frameworks accord- among datasets for the training of these models. ingly. To address these concerns, ethical and legal frame- Cities can ensure AI development and application works regulating AI systems used for urban planning take an inclusive, responsible approach which improves development and deployment must be set in place. citizen services and quality of life while contributing This involves setting ethical rules relating to AI cre- to sustainable development by following such frame- ation and use, conducting audits for prejudice or dis- works. crimination issues within AI systems, as well as hon- oring individuals’ privacy rights [41]. Furthermore, 7. Challenges, limitations, and greater collaboration among developers, urban plan- further research ners, affected communities, and affected individuals must occur to guarantee the ethical implementation 7.1. Challenges and considerations for of AI technologies [43, 44]. AI-enabled urban planning AI is an exciting prospect when applied to urban 7.1.2. Data privacy and security planning; however, its use also presents significant Since AI-enabled urban planning heavily relies on data, hurdles and potential pitfalls which must be considered data privacy, and security are of critical importance. carefully before embarking upon its application. Below Cities collect massive amounts of data from sources we explore several of these considerations in more including sensors, social media accounts, and mobile depth for AI-enabled urban planning: phones which contain personal data including location details, health details, and biometric indicators [42] 7.1.1. Ethical considerations and that could compromise people’s privacy rights if used algorithmic bias for AI planning applications without consent. With AI’s growing use in urban planning comes an in- General Data Protection Regulation (GDPR), in- creased need to consider ethical concerns and algorith- troduced by the European Union in 2018, requires mic bias. AI systems can only ever be as objective organizations to gain consent before collecting or us- as the data on which they are trained; if such data ing personal data, identify its goals, and implement contains prejudices and inequities, then the AI re- security measures aimed at protecting individual data produces or amplifies them [20, 34]. This could have rights while simultaneously imposing sanctions for 89 N. Marji, M. Kohout, L. Chen et al. Acta Polytechnica CTU Proceedings noncompliance [43]. When processing personal data experiences, skillsets, perspectives, and interests that in AI-enabled urban planning applications, compli- contribute towards crafting more holistic and equi- ance with GDPR standards is of utmost importance in table AI-powered urban planning approaches [3, 49]. order to preserve individuals’ rights [44–48]. Gaining Technical language barriers between various fields individuals’ consent before collecting or using their make multidisciplinary collaboration challenging. Ur- data, installing necessary safeguards against illegal ban planners might lack experience with AI while data access or usage, and giving individuals the option of scientists could struggle with the social and cultural access or removal is essential for ethical AI integra- context of urban planning. To overcome this barrier, tion [44–50]. organizations must promote communication and col- As part of GDPR compliance, cybersecurity risks laboration as much as possible, in addition to trans- associated with AI-enabled urban planning must also lating technical terms so nontechnical stakeholders be assessed. Intrusions and ransomware attacks can can understand [24, 30]. Moreover, interdisciplinary have severe repercussions for individuals as well as teams for AI-enabled urban planning projects must cities alike; cities should employ safeguards like en- be formed where representatives from several depart- cryption, multi-factor authentication, vulnerability ments or agencies come together as one team [46, 47]. assessments, penetration testing, as well as incident This would aid in overcoming the fragmented decision- response processes in order to keep data protected making trees and unnecessary bureaucratic elements against cyber threats [41]. typically involved in large-scale masterplanning. 7.1.3. Need for community engagement and Knowledge exchange is key for successfully applying participation AI in urban planning, with best practices, case studies, and research findings shared among municipalities In AI-enabled urban planning, community engagement and regions to reduce duplication. Cities that share and participation are key for successful implementa- their knowledge can avoid reinventing the wheel while tion. AI can assist urban planners in processing and learning from the successes and failures of other cities analyzing large amounts of data but cannot replace the human element in urban planning. Community – an approach encouraged by platforms such as the participation enables residents to provide input and European Union’s Smart Cities Marketplace, where feedback into planning decisions, ensuring that plan- cities share and collaborate across boundaries [50]. ning decisions meet residents’ needs and values [44]. Public meetings, internet forums, surveys, and semi- 7.2. Research limitations nars are among many forms of community engagement This study presents a conceptual framework for the strategies that allow citizens to express concerns, pose incorporation of artificial intelligence (AI) in urban questions to planners, and receive replies [45]. planning and deployment of smart cities using the Community participation presents many unique European context as an example. However, it must be challenges to all members of a given community, partic- acknowledged that AI use for urban planning remains ularly people with impairments and limited language in its infancy, with challenges yet to be overcome, and proficiency. Reaching out to underrepresented popula- ethical and social considerations to be carefully taken tions such as low-income households or minorities that into consideration during the deployment of solutions might otherwise not participate is equally vital [46]. incorporating this technology. Integrating community input effectively into plan- While this paper explores the potential advantages ning decisions presents another challenge, necessitat- and challenges of artificial intelligence in sustainable ing a two-way dialogue between citizens and plan- urban planning in Europe, its findings may not ap- ners requiring active listening from both parties and ply equally elsewhere or under different conditions. a clear explanation of how input from this source Furthermore, while the examples provide indications has been considered [47]. When properly conducted, of potential AI applications within urban planning community engagement and participation can lead to and transition to smart city solutions; they do not more inclusive and informed planning decisions while encompass all possibilities and specific AI technolo- building trust between citizens and planners through gies should be studied in further detail, preferably in greater transparency and accountability in planning defined urban contexts. processes. Moreover, it reveals community concerns and priorities which might not be evident through This research proposes ethical and inclusive AI de- data analysis alone [48]. velopment and implementation criteria, although these should not be seen as exhaustive; their implementa- 7.1.4. Interdisciplinary collaboration and tion should be regularly assessed and evaluated as knowledge sharing AI technology rapidly advances and ethical concerns To effectively incorporate AI in urban planning for arise along with it. transition to smart European cities, urban planners, Finally, this research does not assess the financial policymakers, data scientists, social scientists, and and technical feasibility of using AI for urban planning. community members must come together across dis- Policymakers, urban planners, and stakeholders must ciplines to share knowledge. Each brings different carefully examine both costs and resources necessary 90 vol. 46/2024 AI-enabled transition to smart European cities for adopting this form of planning, as well as the and security, the need for community engagement and technical talent required for success. participation, and interdisciplinary collaboration and information exchange. 7.3. Recommendations for further Responding to the identified gaps in research and research in order to promote responsible and inclusive AI de- While AI may offer many benefits in urban planning, velopment and implementation, this article presents it’s essential to study alternative techniques as well. a framework for implementing AI technologies into Participatory planning approaches, which integrate transition to smart cities, which is the output of this citizen participation and feedback directly, may pro- research, that aims to facilitate responsible AI research duce different outcomes and perspectives than AI- and deployment. Policymakers, urban planners, and enabled planning approaches. stakeholders are encouraged to explore AI as a way Long-term Impacts Assessment: Many potential of creating more livable cities that foster equality and advantages of AI for urban planning may not become sustainability while remaining cognizant of limitations immediately evident and require further examination; associated with this study and further investigation future studies may focus on exploring long-term eco- on its efficacy as an urban planning strategy tool. nomic, environmental, and social impact associated Future research should examine in more detail the with AI-enabled urban planning solutions. capabilities of specific AI tools in defined urban con- AI Implementation Evaluation: As AI technology texts, which could provide a basis for examining the becomes an integral part of urban planning processes, implementation of emerging technological tools in its efficacy needs to be assessed carefully. Further Europe and elsewhere. Furthermore, exploring its ca- research efforts might focus on measuring whether pacity for solving global concerns like climate change, AI-enabled planning strategies meet their intended pandemics, and natural disasters. When used appro- goals effectively. Moreover, future studies should take priately with ethical considerations in mind, AI can steps to develop more detailed recommendations to serve as an indispensable resource for creating more address the ethical and societal repercussions of using resilient, inclusive cities in Europe and elsewhere. AI in various stages of urban planning. Future research should also assess how AI could be Acknowledgements applied in urban planning across different regions and This research was supported by grant: SGS23/081/OHK1/ countries in order to gain a fuller picture of its deploy- 1T/15 by the Faculty of Architecture, Czech Technical ment and efficacy in each setting, with a specific focus University in Prague. on tools such as computer vision and machine learning capabilities of AI. Continued research of AI-enabled References urban planning is necessary in order to guarantee its [1] European Commission. Urban agenda for the EU, appropriate and efficient application, in the hopes 2016. [2023-05-01]. of creating more liveable, equitable, and sustainable https://ec.europa.eu/futurium/en/urban-agenda cities. [2] S. E. Bibri. The IoT for smart sustainable cities of the future: An analytical framework for sensor-based big 8. Conclusion data applications for environmental sustainability. Sustainable Cities and Society 38:230–253, 2018. This article examined the potential and difficulties of https://doi.org/10.1016/j.scs.2017.12.034 AI in urban development and presented a framework for incorporating AI into city planning and manage- [3] A. Caragliu, C. Del Bo, P. Nijkamp. Smart cities in ment in European cities. Cities could achieve in- Europe. Journal of Urban Technology 18(2):65–82, 2011. https://doi.org/10.1080/10630732.2011.601117 creased efficiency, sustainability, and livability with AI being integrated into urban planning processes – [4] European Commission. Ethics guidelines for such as data-driven decision-making, predictive mod- trustworthy AI, 2019. [2023-05-01]. eling, optimizing land use and transportation as well https://ec.europa.eu/digital-single-market/en/ as real-time monitoring using sensor data and involv- news/ethics-guidelines-trustworthy-ai ing citizens in the planning process are some aspects [5] A. Luusua, J. Ylipulli, M. Foth, A. Aurigi. Urban AI: that have been presented herein – showing it has understanding the emerging role of artificial intelligence great promise as an instrument of building resilient in smart cities. AI & SOCIETY 38(3):1039–1044, 2022. European cities. https://doi.org/10.1007/s00146-022-01537-5 This paper also explored several AI applications for [6] R. Kitchin. Thinking critically about and researching sustainable urban planning in Europe, such as green algorithms. Information, Communication & Society infrastructure and urban biodiversity, energy-efficient 20(1):14–29, 2016. building design and retrofitting, smart transportation https://doi.org/10.1080/1369118x.2016.1154087 systems and mobility planning, and resilient urban [7] United Nations. Report of the World Commission on design and disaster mitigation. However, AI adoption Environment and Development: Our common future. presents its own set of unique challenges, such as [2023-05-01]. https://www.un-documents.net/our- ethical concerns and algorithmic bias, data privacy common-future.pdf 91 N. Marji, M. Kohout, L. Chen et al. Acta Polytechnica CTU Proceedings [8] European Commission. 2013 7th environment action [22] M. Akhtar, S. Moridpour. A review of traffic program, 2013. [2023-05-01]. https://ec.europa.eu/ congestion prediction using artificial intelligence. environment/policies/strategies/7eap_en.htm Journal of Advanced Transportation 2021:8878011, [9] I. H. Sarker. AI-based modeling: Techniques, 2021. https://doi.org/10.1155/2021/8878011 applications and research issues towards automation, [23] V. Marinakis. Big data for energy management and intelligent and smart systems. SN Computer Science energy-efficient buildings. Energies 13(7):1555, 2020. 3(2):158, 2022. https://doi.org/10.3390/en13071555 https://doi.org/10.1007/s42979-022-01043-x [24] S. A. A. Bokhari, S. Myeong. Use of artificial [10] S. C. K. Tékouabou, J. Chenal, R. Azmi, et al. intelligence in smart cities for smart decision-making: Identifying and classifying urban data sources for A social innovation perspective. Sustainability 14(2):620, machine learning-based sustainable urban planning and 2022. https://doi.org/10.3390/su14020620 decision support systems development. Data 7(12):170, [25] G. Castelli, G. Tognola, E. F. Campana, et al. Urban 2022. https://doi.org/10.3390/data7120170 intelligence: a modular, fully integrated, and evolving [11] S. Shams Amiri, S. Mottahedi, E. R. Lee, S. Hoque. model for cities digital twinning. In 2019 IEEE 16th Peeking inside the black-box: Explainable machine International Conference on Smart Cities: Improving learning applied to household transportation energy Quality of Life Using ICT & IoT and AI consumption. Computers, Environment and Urban (HONET-ICT). IEEE, 2019. Systems 88:101647, 2021. https: https://doi.org/10.1109/honet.2019.8907962 //doi.org/10.1016/j.compenvurbsys.2021.101647 [26] Y. Zhang, S. Balochian, P. Agarwal, et al. Artificial [12] Y.-M. Guo, Z.-L. Huang, J. Guo, et al. Bibliometric intelligence and its applications. Mathematical Problems analysis on smart cities research. Sustainability in Engineering 2014:840491, 2014. 11(13):3606, 2019. https://doi.org/10.1155/2014/840491 https://doi.org/10.3390/su11133606 [27] X. Zhang, K. Shu, S. Rajkumar, V. Sivakumar. [13] M. Batty, K. W. Axhausen, F. Giannotti, et al. Research on deep integration of application of artificial Smart cities of the future. The European Physical intelligence in environmental monitoring system and Journal Special Topics 214(1):481–518, 2012. real economy. Environmental Impact Assessment https://doi.org/10.1140/epjst/e2012-01703-3 Review 86:106499, 2021. [14] R. Abduljabbar, H. Dia, S. Liyanage, S. A. Bagloee. https://doi.org/10.1016/j.eiar.2020.106499 Applications of artificial intelligence in transport: [28] B. Shin, M. Rask, P. Tuominen. Learning through An overview. Sustainability 11(1):189, 2019. online participation: A longitudinal analysis of https://doi.org/10.3390/su11010189 participatory budgeting using big data indicators. [15] Y. Gu, Y. Wang, S. Dong. Public traffic congestion Information Polity 27(4):517–538, 2022. estimation using an artificial neural network. ISPRS https://doi.org/10.3233/ip-211551 International Journal of Geo-Information 9(3):152, [29] AI4Cities. AI accelerating cities’ transition to carbon 2020. https://doi.org/10.3390/ijgi9030152 neutrality, 2022. [2023-05-01]. [16] A. Urban, D. Hick, J. R. Noennig, D. Kammer. With https://innovation-procurement.org/ai4cities/ a little help from AI: Pros and cons of AI in urban [30] A. Ortega-Fernández, R. Martín-Rojas, V. J. García- planning and participation. International Journal of Morales. Artificial intelligence in the urban environment: Urban Planning and Smart Cities 2(2):19–33, 2021. Smart cities as models for developing innovation and https://doi.org/10.4018/ijupsc.2021070102 sustainability. Sustainability 12(19):7860, 2020. [17] X. Xiang, Q. Li, S. Khan, O. I. Khalaf. Urban water https://doi.org/10.3390/su12197860 resource management for sustainable environment [31] J. Gusc, P. Bosma, S. Jarka, A. Biernat-Jarka. The planning using artificial intelligence techniques. big data, artificial intelligence, and blockchain in true Environmental Impact Assessment Review 86:106515, cost accounting for energy transition in Europe. 2021. https://doi.org/10.1016/j.eiar.2020.106515 Energies 15(3):1089, 2022. [18] M. Zekić-Sušac, S. Mitrović, A. Has. Machine learning https://doi.org/10.3390/en15031089 based system for managing energy efficiency of public [32] P. K. Gupta, V. Shree, L. Hiremath, S. Rajendran. sector as an approach towards smart cities. International The Use of Modern Technology in Smart Waste Journal of Information Management 58:102074, 2021. Management and Recycling: Artificial Intelligence and https://doi.org/10.1016/j.ijinfomgt.2020.102074 Machine Learning, pp. 173–188. Springer International [19] W. Sun, P. Bocchini, B. D. Davison. Applications of Publishing, 2019. artificial intelligence for disaster management. Natural https://doi.org/10.1007/978-3-030-12500-4_11 Hazards 103(3):2631–2689, 2020. [33] J. M. Gutierrez, M. Jensen, M. Henius, T. Riaz. Smart https://doi.org/10.1007/s11069-020-04124-3 waste collection system based on location intelligence. [20] C. O’Neil. Weapons of Math Destruction. Penguin, Procedia Computer Science 61:120–127, 2015. UK, 2016. https://doi.org/10.1016/j.procs.2015.09.170 [21] J. Angwin, J. Larson, S. Mattu, L. Kirchner. [34] R. Benjamin. Race after technology: Abolitionist Machine bias. In Ethics of data and analytics, pp. tools for the new jim code. Social Forces 98(4):1–3, 254–264. Auerbach Publications, 2022. 2019. https://doi.org/10.1093/sf/soz162 92 vol. 46/2024 AI-enabled transition to smart European cities [35] J. Buolamwini, T. Gebru. Gender shades: [43] Protection Regulation. Regulation (EU) 2016/679 of Intersectional accuracy disparities in commercial gender the European Parliament and of the Council, 2016. classification. In Conference on fairness, accountability [44] L. Sandercock. Out of the closet: The importance of and transparency, pp. 77–91. PMLR, 2018. stories and storytelling in planning practice. Planning [36] K. Holstein, J. Wortman Vaughan, H. Daumé, et al. Theory & Practice 4(1):11–28, 2003. Improving fairness in machine learning systems: What https://doi.org/10.1080/1464935032000057209 do industry practitioners need? In Proceedings of the [45] S. R. Arnstein. A ladder of citizen participation. 2019 CHI Conference on Human Factors in Computing Journal of the American Institute of Planners Systems, CHI ’19, pp. 1–16. ACM, 2019. 35(4):216–224, 1969. https://doi.org/10.1145/3290605.3300830 https://doi.org/10.1080/01944366908977225 [37] S. Barocas, A. D. Selbst. Big data’s disparate impact. California law review pp. 671–732, 2016. [46] G. Brown, M. Kyttä. Key issues and research priorities for public participation GIS (PPGIS): [38] V. Eubanks. Automating inequality: How high-tech A synthesis based on empirical research. Applied tools profile, police, and punish the poor. St. Martin’s Geography 46:122–136, 2014. Press, 2018. https://doi.org/10.1016/j.apgeog.2013.11.004 [39] J. Burrell. How the machine ‘thinks’: Understanding opacity in machine learning algorithms. Big Data & [47] J. Forester. Planning in the face of power. Journal of Society 3(1):205395171562251, 2016. the American Planning Association 48(1):67–80, 1982. https://doi.org/10.1177/2053951715622512 https://doi.org/10.1080/01944368208976167 [40] L. Floridi. AI ethics: An ethical framework for AI and [48] S. Campbell, S. S. Fainstein. Introduction: The data use. Philosophy & Technology pp. 639–662, 2019. structure and debates of planning theory. Readings in [41] A. Jobin, M. Ienca, E. Vayena. The global landscape planning theory pp. 1–16, 2003. of AI ethics guidelines. Nature Machine Intelligence [49] M. Batty. The new science of cities. MIT Press, 2013. 1(9):389–399, 2019. [50] European Commission. Smart cities marketplace. https://doi.org/10.1038/s42256-019-0088-2 [2023-05-01]. [42] R. Kitchin. The data revolution: Big data, open https://smart-cities-marketplace.ec.europa.eu/ data, data infrastructures and their consequences. The Data Revolution pp. 1–240, 2014. 93