Sustainable Cities and Society 94 (2023) 104562

Contents lists available at ScienceDirect

Sustainable Cities and Society

journal homepage: www.elsevier.com/locate/scs

Algorithmic urban planning for smart and sustainable development:

Systematic review of the literature
Tim Heinrich Son a, Zack Weedon a, Tan Yigitcanlar a, *, Thomas Sanchez b, Juan M. Corchado c,
Rashid Mehmood d
a
City 4.0 Lab, School of Architecture and Built Environment, Queensland University of Technology, 2 George Street, Brisbane, QLD 4000, Australia
b
School of Public and International Affairs, Virginia Tech, 900N Glebe Rd. Arlington, VA 22203, United States
c
Bisite Research Group, University of Salamanca, Salamanca 37007, Spain
d
High Performance Computing Center, King Abdulaziz University, Jeddah 21589, Saudi Arabia

A R T I C L E I N F O A B S T R A C T

Keywords: In recent years, artificial intelligence (AI) has been increasingly put into use to address cities’ economic, social,
Artificial intelligence environmental, and governance challenges. Thanks to its advanced capabilities, AI is set to become one of local
Big data governments’ principal means of achieving smart and sustainable development. AI utilisation for urban planning,
Urban planning
nonetheless, is a relatively understudied area of research, particularly in terms of the gap between theory and
Urban development
Sustainable development
practice. This study presents a comprehensive review of the areas of urban planning in which AI technologies are
Smart cities contemplated or applied, and it is analysed how AI technologies support or could potentially support smart and
sustainable development. Regarding the methodological approach, this is a systematic literature review
following the PRISMA protocol. The obtained insights include: (a) Early adopters’ real-world AI applications in
urban planning are paving the way to wider local government AI adoption; (b) Achieving wider AI adoption for
urban planning involves collaboration and partnership between key stakeholders; (c) Big data is an integral
element for effective AI utilisation in urban planning, and; (d) Convergence of artificial and human intelligence is
crucial to address urbanisation issues adequately and to achieve smart and sustainable development. These in­
sights highlight the importance of making planning smarter through advanced data and analytical methods.

1. Introduction the technocentric approach to the notion of the smart city, utilising
advanced technologies in urban planning to achieve smart and sus­
Since the unprecedented rise of global urbanisation, expected to tainable development goals (SDG) (Giuliani et al., 2020; Yigitcanlar
reach about 70% by 2050, cities are facing growing pressure in relation et al., 2020a, 2020b). These goals are presented in Fig. 1.
to economic, social, environmental, and governmental aspects (Perveen Although using technologies in the management of cities is not an
et al., 2017). Presently, the pressures of COVID-19, population increase entirely new concept—going back as far as the late 1950s and early
mismanagement, climate change externalities, environmental degrada­ 1960s—it is only recently that the potential for these technologies is
tion, housing unaffordability, and insecurities associated with the being widely recognised (Huang, 2021). Particularly in urban planning,
water-food-energy nexus have been subjects of intense debate amongst meeting SDGs—especially those goals that are more directly linked with
scholars, urban planners, and policymakers (Berawi, 2019). In response, urban contexts, i.e., SDG 3,6,7,9,11,15 (Zhou et al., 2022)—more effi­
the concept of sustainable development has been pushed to the forefront ciently, is becoming more achievable with the rapid advances in data
of urban policy debate in the hopes of constructing a desirable urban collection (Sanchez et al., 2022). Currently, the implementation of big
future (Yigitcanlar & Teriman, 2015). Sustainable development is data collection sensors that monitor changes in land use, transportation
defined as the balance of sustainable economic growth and ecological patterns, real-estate investments, and energy use, has accelerated the
regeneration. It is the promise of meeting future urban goals without application of disruptive technologies, both new and mature (Sanchez
compromising society’s well-being, quality of life, and environment et al., 2022). These disruptive emerging technologies, including artifi­
(Berawi, 2019). Accordingly, many cities around the globe have taken cial intelligence (AI), internet of things (IoT), machine learning, deep

* Corresponding author.
E-mail address: [email protected] (T. Yigitcanlar).

https://doi.org/10.1016/j.scs.2023.104562
Received 10 December 2022; Received in revised form 23 March 2023; Accepted 28 March 2023
Available online 29 March 2023
2210-6707/© 2023 The Authors. Published by Elsevier Ltd. This is an open access article under the CC BY license (http://creativecommons.org/licenses/by/4.0/).
T.H. Son et al. Sustainable Cities and Society 94 (2023) 104562

Fig. 1. Sustainable development goals (www.un.org/sustainabledevelopment).

learning, artificial neural networks, and 5G/6G, along with big data, prospects of AI applications in sustainable development and urban
have been used to create, expand, and monitor the effectiveness of smart planning are as “diverse as land use, zoning and permitting, environ­
and sustainable development across the globe (Berawi, 2019). As such, mental planning, and transportation” (Sanchez et al., 2022, p.4). More
AI has been acknowledged as a highly promising technology in smart specifically, AI can assist urban planners to provide the best possible and
and sustainable urban development (Sanchez et al., 2022). equitable networks for larger traffic management and public trans­
While most people may encounter the applications of AI every day, portation. It can help urban designers to respond to and design certain
such as their social media feeds, emails, and search engines, the environments, creating more efficient communities. Moreover, AI can

Fig. 2. AI knowledge representation, derived from Corea (2018).

2
T.H. Son et al. Sustainable Cities and Society 94 (2023) 104562

predict and analyse air quality within cities, publishing results for ‘Approaches’ or tools based on: Logic—the knowledge representation
pollution levels, fossil fuel particle density, and future levels, where and problem-solving; Knowledge—the ontology of notions, information,
these provide invaluable input for informing urban administrations and and rule; ‘Probabilistic’ methods—incomplete information and data;
policymakers (Mahendra, 2021). ‘Machine learning’—the ability to learn from historical data; this field
Today, urban big data analytics and data-driven planning technolo­ also contains its sub-category, namely, ‘Deep learning’; ‘Embodied
gies are seen as the foundations of algorithmic urban-planning-based intelligence’—the ability to affect the physical environment; and ‘Search
smart and sustainable development (Lazaroiu et al., 2020; Lazaroiu & and optimisation’—the intelligent search and higher outcomes of solu­
Harrison, 2021; Nica, 2021). When generating smart and sustainable tions. It is from these approaches that the different AI-enabled ‘Tech­
development through urban planning, a comprehensive and critical nology’ is utilised, such as ‘evolutionary algorithms, ambient
understanding of the implemented technologies is paramount. To that computing, distributed AI, autonomous systems, artificial neural net­
end, undertaking a thorough review of the existing knowledge and works, probabilistic programming, decision networks, computer vision,
practice on the topic is essential not only for stocktaking purposes but and natural language processing’ (Corea, 2018).
also as a means of critically examining the risks and limitations, and The variety of AI applications that have evolved in recent years is
future opportunities of AI within the urban planning context. astounding. These include social sensing (Liu et al., 2015), event
Thus, this paper presents a comprehensive review of the urban detection of road traffic from social media data (Alomari et al., 2021a),
planning areas in which AI technologies are contemplated or applied, as AI applications for smart airports and smart districts (Janbi et al., 2020),
well as an analysis of how AI technologies are supporting or could learning public, industry and research perspectives of transportation
potentially support smart and sustainable development. Regarding the and the gaps between them for better planning and policy developments
methodological approach, this systematic literature review has followed (Ahmad et al., 2022), learning a holistic view of dimensions surrounding
the Preferred Reporting Items for Systematic Reviews and Meta Analysis families from research and public perspective with the aim to develop
(PRISMA) protocol. The PRISMA protocol, hence, is adopted to address broad as well as culture-specific policies and solutions (Alqahtani et al.,
the research questions of: (a) In which urban planning areas are AI 2022), co-creating healthcare services from social media (Alahmari
technologies contemplated or applied? (b) How can AI technologies et al., 2022), participatory governance of online and in-class education
support smart and sustainable development planning? Here, we note during the COVID-19 pandemic (Alswedani et al., 2022), learning gov­
that this systematic literature review primarily covers research activities ernment pandemic measures from social media data (Alomari et al.,
as opposed to planning practice, as stated by Geertman (2017) there is a 2021b), governance of AI in energy systems (Alsaigh et al., 2022),
large gap between research and implementation in this area. learning age dynamics in labour markets (Alaql et al., 2022), solar en­
Following this introduction, the remainder of the article is structured ergy forecasting and management (Alkhayat et al., 2022), skin disease
as follows. Section 2 provides a concise literature summary, Section 3 diagnosis using mobile phones and edge devices (Janbi et al., 2022),
introduces the methodological approach, Section 4 reports the results, DNA forensics (Alotaibi et al., 2022), stock trading (Malibari et al.,
followed by a discussion in Section 5 and a conclusion in Section 6. 2022), and many others.
In recent years, AI also started to be used in local government
2. Literature background: artificial intelligence in a nutshell agencies. The utilisation of AI mainly concerned two distinctive but
interconnected areas: (a) Urban service delivery and operations (e.g.,
AI is a widely used term, however, it is not well understood by most traffic control, garbage collection), and; (b) Urban policy, decision-
people, including urban planners (Sanchez et al., 2022). While the making, and planning tasks (e.g., land use planning, development con­
concept of integrating intelligence in machines and systems can be trol) (Allam & Dhunny, 2019; Andrews et al., 2022; Samsurijan et al.,
traced back to the 16th century, it was not until 1956 that the term AI 2022). Since AI’s recent emergence and availability in urban applica­
was officially coined by computer scientists; John McCarthy, Allen tions, smart cities have been utilising the technology to further tackle
Newell, Cliff Shaw, and Herbert Simon (Press, 2016; Anyoha, 2017; urban challenges and sustainable development in smarter ways (Kim
Herath & Mittal, 2022). Back then, AI had yet to be further con­ et al., 2021). Considering the achievement of initiatives, Yigitcanlar
ceptualised to have a standard definition. Some informal definitions et al. (2019a) found that the outcomes of smart city improvements
described AI as the capacity to accomplish goals in a variety of uncertain involve productivity and innovation; liveability and well-being; sus­
environments within highly adaptive, general-purpose systems through tainability and accessibility; and governance and planning.
self-directed learning (Legg & Hutter, 2007). A single definition that Further, these initiatives can be sectioned under the umbrella terms
provides a better, present-day understanding of the AI paradigm de­ of economy, society, environment, and governance (Allam & Dhunny,
scribes AI as the machine mimicry of human cognitive traits and actions 2019). Accordingly, developing smart cities through urban planning
in learning and problem-solving activities such as communication, requires learning different dimensions of perspectives, insights, and ef­
reasoning, knowledge, perception, and planning (Corea, 2018; Yigit­ ficiencies for communities while maintaining attention to the potential
canlar et al., 2021; Frankenfield, 2022). bias and negative impacts (Sanchez et al., 2022). As such, urban plan­
What sets human cognitive abilities apart from AI are the differences ning requires strategic direction at all levels. Therefore, in the context of
in how tasks are performed. While humans automate tasks manually, AI smart cities, urban planning can benefit from the means or incorporation
can reliably and efficiently execute high-volume tasks autonomously of AI into planning practice for smart and sustainable development, solid
(Zhang et al., 2022). Furthermore, AI can automate, repeat, learn, economic growth, and improved quality of life of citizens (Nam &
discover, and adapt large amounts of data (Testi, 2021). Several limi­ Pardo, 2011; Micozzi & Yigitcanlar, 2022).
tations were found in existing computer-based approaches to data Today, AI applications can be seen in various parts of the world.
problem solving (Han & Kim, 1989). Examples such as the ability to Cities in Europe, America, and Asia, particularly in Amsterdam, London,
comprise great amounts of data, graphs, or figures, or analyse, under­ Vienna, Stockholm, Toronto, Singapore, and Hong Kong, have been
stand, or establish relationships (Gauglitz, 2019). In response, AI was utilising AI to achieve sustainable outcomes in their smart city trans­
recommended as the appropriate solution to overcoming formation objectives (Yigitcanlar et al., 2020c; Tekouabou et al., 2022).
problem-solving limitations (Han & Kim, 1989). For example, enhanced analysis capabilities, planning, awareness, and
As illustrated in Fig. 2, AI-enabled technologies are employed to recovery operations in urban disaster management (Mikulsen & Diduck,
address specific problem-solving activities. Large collections of data, 2016). Likewise, urban planning systems have employed, amongst other
commonly referred to as big data, are often obtained by means of IoT- things, “AI in traffic system management, crime detection, air quality
enabled infrastructures (Hajjaji et al., 2021). Utilising these data, AI monitoring, efficient energy management, and water leakage detection
paradigms can be formulated. These paradigms are described as systems” (Jha et al., 2021, p.938).

3
T.H. Son et al. Sustainable Cities and Society 94 (2023) 104562

Proven to be successful to provide solutions to existing issues and Table 1

challenges in more efficient means, certain networks, and algorithms, Selection criteria for publication.
such as random forest, convolutional neural network, and support vector Inclusionary criteria Exclusionary criteria
machine algorithms have been efficiently utilised in the classification
Urban, City, Town or Regional AI not in the Urban, City, Town or Regional
and pattern analysis of urban data. Other approaches such as cellular Planning Planning Field
automata, spatial logic regression, and agent-based modelling were also Artificial Intelligence or AI Not an AI Technology
identified to be best suited to the study of urban population growth, land Deep Learning Irrelevant to the Research Aims and
use changes, and settlement pattern analysis (Gulshad et al., 2022). A Machine Learning Questions
Neural Networks
study by Yigitcanlar et al. (2022a, p.1) disclosed the “AI adoption areas, Sustainable Urban Development
cautionary areas, challenges, effects, impacts, knowledge basis, plans, Smart or Sustainable Cities
preparedness, roadblocks, technologies, deployment timeframes, and Relevant to the Research Aims and
usefulness” in the context of local governments—where most of the Questions
urban planning practice takes place.
It has also been demonstrated that deep learning and computer
understanding of their significance and relevance to the field.
vision can identify global urban issues that encompass land use, build­
Given the context provided above, two primary research questions
ings, and the natural environment (Tekouabou et al., 2022), including
have guided the literature review. First, the study focused on identi­
slums, traffic congestion, and crowd mapping with less effort and data
fying: (a) In which urban planning areas are AI technologies contem­
available (Ibrahim et al., 2021). Moreover, the not-so-far-reaching
plated or applied? Secondly, it concentrated on answering: (b) How can
future of AI and its influence on the challenges of urban environments
AI technologies support smart and sustainable development planning?
are not only exclusive to the aforementioned areas of smart cities but
In this systematic review, the Preferred Reporting Items for Systematic
also cover many multifaceted dimensions (Nikitas et al., 2020). For
Reviews and Meta-Analyses (PRISMA) protocol has been employed (see
instance, a decision/planning support system that evaluates the suit­
Regona et al., 2022). The PRISMA protocol is a widely accepted and
ability of a building or a space by considering a variety of characteristics
recognised framework designed to enhance the transparency, consis­
such as population distribution and composition, transport accessibility,
tency, and comprehensiveness of systematic reviews.
urban form, and availability of infrastructures, was put to the test in the
The protocol encompasses several stages, including defining eligi­
urban environment of Lyon, France. The system was found to be suc­
bility criteria for study selection, developing a thorough literature
cessful (Sideris et al., 2019). Likewise, a case study in China, which
search strategy, selecting relevant studies, extracting data, assessing the
assessed the success of AI, including decision support systems,
risk of bias, and synthesising and analysing the data. Each stage is per­
knowledge-based systems, and artificial neural networks, only called for
formed by at least two independent reviewers’ involvement—i.e., the
refinements and improvements (Feng & Xu, 1999).
first two authors of this paper—to minimise biases and ensure accuracy.
The effectiveness of AI is even reflected in its ability to understand
In cases of disagreement, a third reviewer—i.e., the third author of this
the emotions and reactions of people at any given time and situation.
paper—was consulted. By adhering to the PRISMA protocol, our sys­
According to Kaklauskas et al. (2021, p.3), the development of an “af­
tematic review has maintained a high level of methodological rigour,
fective system for researching emotions in public spaces for urban
which enhances the reliability and validity of the findings. Furthermore,
planning” was based on behavioural economics and “psychology of
this approach facilitates the identification of emerging patterns, trends,
judgement and decision-making and human emotional affective and
and gaps in the existing body of literature, ultimately contributing to a
physiological states”, which assists urban planners in analysing planning
more comprehensive understanding of the research topic. A search for
processes more effectively and in reaching a rational decision. In
relevant publications from Scopus and Web of Science (WoS) databases
another study, conducted by Zou et al. (2019), the use of AI was
was performed, as they are the two leading bibliographic databases for
improved by 50% in the forecasting of populations and in the spatial
scholarly work that contain the body of literature we are focusing on.
modelling of urban environments.
To ensure the diversity and relevance of the documents collected
Moreover, the increase in big data availability in recent years led to
from the databases, various search keywords were combined to refer­
the emergence of new urban data analytics options for planners, such as
ence existing and emerging AI technologies. In September 2022, a
GeoAI which offers advanced computing for scalable processing and
Boolean search query was completed that included the following key­
intelligent analysis of geospatial data (Li, 2020). In sum, assuming urban
words: (TITLE-ABS-KEY("urban planning") OR TITLE-ABS-KEY("city
planners adopt AI broadly, this technology has a significant potential to
planning") OR TITLE-ABS-KEY("town planning") OR TITLE-ABS-KEY
support decision and policymaking as well as fully automate routine and
("regional planning")) AND (TITLE-ABS-KEY("artificial intelligence") OR
menial planning tasks—if the shortcomings of AI are considered and
TITLE-ABS-KEY("AI") OR TITLE-ABS-KEY("machine learning") OR
addressed in each task (Andrews et al., 2022).
TITLE-ABS-KEY("deep learning") OR TITLE-ABS-KEY("neural net­
works")). The publication date range was set to no limit on the start date
3. Methodology and August 2022 as the end date. The search disclosed the date of the
earliest relevant publication was 1986. The search criteria were also
This study presents a review of AI technologies applied or considered limited the records written in English; categorised as either a journal
for urban planning, examining their potential in supporting smart and article, book chapter or conference paper; and the full text had to be
sustainable development. The current body of literature is somewhat available online.
fragmented, primarily because it is still relatively new. Nevertheless, it is The search query produced 5537 records in total from both databases
developing rapidly. The topics are emerging and coalescing, and themes (Scopus = 3430 and WoS =2107). The records were checked for
will become more evident as the research matures over time. Performing duplication, where 4343 records were removed; and then their titles and
a systematic literature review on a new body of research with emerging keywords were checked to see if they were fit. This process has led to the
topics and themes serves several purposes. One of the key reasons is to further removal of 627 records. The remaining list of records (n = 567)
identify the current state of knowledge in the field, which helps re­ was then screened against the inclusion and exclusion criteria by
searchers understand existing discoveries, innovations, and findings. focusing individually on the records’ abstracts (Table 1). This has
This provides a foundation for further investigation and exploration. The resulted in removal of 399 records—leaving 168 records for full-text
objective is to also help establish the context and relevance of the screening. The remaining 168 full-text records were read and their
emerging topics and themes. By situating these new ideas within the compliance with the inclusion and exclusion criteria was checked. This
broader context of related research, researchers can gain a better

4
T.H. Son et al. Sustainable Cities and Society 94 (2023) 104562

Fig. 3. Literature selection process.

exercise has removed a further 77 records. This screening reduced the monitoring and development control’.
final selection of publication numbers to 91. Fig. 3 details the literature ‘Urban data analytics and planning decision support’ was a relevant
review selection process. category to 49.5% of the reviewed publications (n = 45). It was followed
The final selection of 91 publications was analysed and clustered into by ‘urban and infrastructure management’ with 20.9% (n = 19), ‘urban
four categories, based on the following urban planning areas in which AI environmental and disaster management’ with 15.4% (n = 14), and
technologies are contemplated or applied: (a) Urban data analytics and ‘urban monitoring and development control’ with 14.2% (n = 13).
planning decision support; (b) Urban and infrastructure management; The most common application type of AI was found in machine
(c) Urban monitoring and development control; and (d) Urban envi­ learning (n = 76), which accounted for 84% of publications. This was
ronmental and disaster management. The list of reviewed publications followed by deep learning (n = 46) with about 51%, and neural net­
and summary of the findings of the review for each paper, is presented in works (n = 31) with 34%—as the reviewed publications reference
Appendix Table A1. Additionally, the PRISMA 2000 checklist is included multiple AI applications, numbers and percentages do not add up to
in Appendix Table B1. 100%.
The top-3 commonly published journal outlets were Environment
4. Results and Planning B (n = 4), Sustainability (n = 4), and Remote Sensing (n =
4). In terms of conference proceedings, IEEE conference series on
4.1. General observations various technology foci were the most popular ones (n = 5), followed by
the International Archives of the Photogrammetry, Remote Sensing and
From the total of 91 selected publications, little over 78% were Spatial Information Sciences (n = 2) and International Conference on
journal articles (n = 71) and almost 22% were book chapters and con­ Computer-Aided Architectural Design Research in Asia (n = 2). Fig. 4
ference papers (n = 20). These publications were clustered into four indicates the number of publications per year. In total, approximately
categories through a full-text analysis of each publication, followed by 77% of the identified publications were published within the past four-
an assignment of the best fitting category. The categorisation has shown year period (2019–2022)—indicating the recent emergence of the
heavy dominance of AI applications in planning concerning the areas of importance of AI for urban planning. The analysis revealed a strong
‘urban data analytics and planning decision support’ and ‘urban relevance of AI in urban planning with the smart city notion (about 78%

5
T.H. Son et al. Sustainable Cities and Society 94 (2023) 104562

Fig. 4. Publications by year.

of reviewed papers concerned the smart city context) in the context of exists a limited, but growing number of examples of the utilisation of AI
advanced technology utilisation to achieve improvements in the per­ technologies in urban service delivery and operations, and urban policy,
formance of cities. decision-making, and planning tasks (Cugurullo, 2020). The expectation
The association of each article with the relevant SDG has shown is that the overall economic productivity of a city will be influenced by
common alignment with goals 3, 6, 7, 9, 11, 15 and 17. Unsurprisingly, AI through contributing to improvement in many areas across the city,
the most common goal was aligned Goal 11 – Sustainable Cities and including health, traffic, planning, air quality, and the like (Jha et al.,
Communities (n = 70). The topics associated to this goal are disaster risk 2021). These overarching uses of AI could support planning decisions.
reduction; sustainable transport; and, sustainable cities and human settle­ The most common areas where AI is used or considered for analytics and
ments. Most of the articles were aligned to the sustainable cities and decision support includes air quality, urban/public health, planning
human settlements topic due to the nature of the research question. The tasks, traffic management, urban policy, and urban service delivery and
second most common goal was aligned to Goal 17 – Partnerships for the operations.
Goals, specifically the Technology topic (n = 52). This SDG is not The results show that many instances of AI have been applied to
commonly associated with urban planning, however, in the context of address urban planning practices in smart cities (Jain, 2011; Cugurullo,
the use of developed and deployed AI across urban planning, technology 2020) and the risks and opportunities related thereto (Gray & Kova­
will be critical in addressing the challenges that growing urbanisation cova, 2021). Innovative solutions that improve the use and analysis of
poses to building sustainable cities. big data in cities could lead urban planners to having a greater under­
An analysis of co-occurrence and frequencies for SDGs and AI tech­ standing of the range of factors involved, that were not perceivable
nologies, in the context of the papers reviewed in this study, is shown in without the use of AI. The desirable engagement or stakeholder activ­
Fig. 5. The node size is the degree (frequency) and the edge shows fre­ ities, associated with the planning decision process, are often not
quency of appearing together. As its evident in Fig. 5, SDGs 3 and 11 are possible to achieve due to time, efficiency and budget limitations. Thus,
the most frequent (Good Health and Well-being, Sustainable Cities and AI can support some of the planning processes to support the planning
Communities), followed by SDGs 1, 6, and 15 (No Poverty, Clean Water decision. AI enabled deep learning, can assess citizens’ urban perception
and Sanitation, Life on Land). On the contrary, SDGs 8 and 10 appear the of streets (Yao et al., 2019) or address the environmental concerns of
least often (Decent Work and Economic Growth, Reduced Inequality), pollution in cities (Ameer & Shah, 2018).
where SDGs 11 and 17 appear together most often (Sustainable Cities This would normally require extensive amounts of resources and
and Communities, Partnerships to Achieve the Goal). Information on the time. Traffic management in smart cities is a complex problem with a
relevance of each reviewed paper and the SDGs is available in Appendix vast array of research and AI applications that are attempting to improve
Table A1. traffic efficiency, congestion, and management. AI has been used to
analyse big data and understand the current transportation use and
traffic forecasting of complex transportation networks (Jha et al., 2021).
4.2. AI for urban data analytics and planning decision support A framework has been developed to measure the social impact of a
development during its design process (Ferreira et al., 2015). Another
Smart city planning, in theory, is focussed predominately on utilising alternative model can provide sustainable planning strategies that
AI technology to improve efficiency, effectiveness, innovation and build respond to social issues in cities (Koehler et al., 2009). Additionally,
a more sustainable and resilient city. Nevertheless, in practice, there

6
T.H. Son et al. Sustainable Cities and Society 94 (2023) 104562

Fig. 5. Co-occurrence and frequencies for SDGs and AI technologies.

deep learning technologies have assisted in the spatial distribution was a common concern that denoted that these AI-supported applica­
patterns of jobs and housing (Yao et al., 2020). All these processes tions may have unethical effects on daily life and urban planning (Engin
require extensive data to be collected, analysed, and processed; with the et al., 2020). Even though the results demonstrate the integration of AI
support of AI the efficiency is far greater. systems can provide better solutions (Chaturvedi & de Vries, 2021), it is
Other more traditional planning decisions can be supported by AI. suggested that greater collaboration between interdisciplinary teams in
The literature highlights that AI, generally expert systems, were used for sharing data to improve outcomes is needed (Bazan-Krzywoszańska
master planning, including the creation of drawings and renderings (Ye et al., 2020).
et al., 2022) and the assessment of the feasibility of the master plan Appendix Table A1 presents a summary—including the aim, rele­
(Polydorides & Petropoulous, 1989). There is great potential in vance, technologies, and findings—of the 45 papers concentrating on AI
streamlining some planning decisions with the support of AI. For for urban data analytics and planning decision support.
example, land use mapping continues to be a human decision-making
involvement as part of land use planning (Balling et al., 1999). How­
ever, numerous examples identify how AI can support land use planning 4.3. AI for urban and infrastructure management
(Chaturvedi & de Vries, 2021; Meeran & Conrad Joyce, 2020; Yuan
et al., 2022). Traffic also affects the adjoining land uses as part of the This domain is concerned with the development, governance, and
greater smart city transport network (Meeran & Conrad Joyce, 2020). management of cities, particularly, with the management of the urban
The reliability and performance of land use planning has a significant environment and infrastructure for more resilient and sustainable smart
impact on the productivity and efficiency of a city. city developments. The literature indicates that as cities continue to
The results in this domain show that there are AI systems that are grow, it is expected that the urban environment and infrastructure
used to assist and inform human decision-making (Tanic, 1986). More within a city will become ever more challenging to plan, design, and
recent examples include the Smart Design Framework where AI is used comprehend (Fang et al., 2022). Alongside this foreseeable challenge,
to support and inform the design process as part of an overall integrated the results largely indicate that the incorporation and implementation of
workflow (Quan et al., 2019). Furthermore, urban planners could direct AI-enabled technologies in urban planning practices can assist in the
and influence how AI is integrated in urban planning processes (Kar­ future management of smart and sustainable urban environments and
vonen et al., 2020) and allow for risk mitigation and ethical consider­ infrastructures (Javed et al., 2022). Achieving not only a smart but also a
ations to be at the forefront of urban planning. While many difficult sustainable city requires the collaboration of multiple stakeholders; the
urban issues and challenges can be solved with the support of AI, there implementation of AI will assist in the implementation of smart city
initiatives that respond to urban and infrastructure challenges (Anthony

7
T.H. Son et al. Sustainable Cities and Society 94 (2023) 104562

Jnr, 2021). 2021) are principal examples of the problems to which AI has provided
As more complex urban issues arise each day, urban planners are solutions. Urban disaster management (Abid et al., 2021), flood resis­
required to respond to them appropriately and efficiently; however, tance (Ye et al., 2021), and flood risk assessment (Pham et al., 2021) for
many of them are impenetrable and thus, it is not always feasible for an urban areas are amongst the other AI-supported applications.
urban planner and/or decision-maker to undertake the problem solving Each paper highlights the extent to which AI influences and supports
themselves. Thereby, the results indicate that this is a key factor asso­ the applications used to plan the natural and physical environment in
ciated with the interest in implementing AI to assist urban planners, the creation of smart cities. For instance, it may be inferred that the
particularly in resolving complex problems in (smart) cities. More spe­ purpose of any AI approach is to enhance the physical and natural
cifically, AI has been demonstrated to support the problem-solving environment for inhabitants to be more comfortable and have increased
associated with urban design and modelling of city infrastructures. accessibility to urban spaces (Bienvenido-Huertas et al., 2020). The re­
amongst these AI technologies are machine learning, deep learning, and sults of the review also show that these urban issues can have significant
conventional neural networks, as well as ad hoc managerial and design impacts on a sizable population, highlighting the value of using
processes, including the automatic and interactive detection, genera­ AI-supported technologies, such as machine learning, for urban mana­
tion, prediction, measurement, information, mapping, and catego­ gerial processes. The studies in this domain disclose that AI can offer
risation of street networks (Boulange et al., 2018; Law et al., 2020; historical and current analyses during crucial times when it comes to
Ibrahim et al., 2021; Fang et al., 2022; Yerram et al., 2022). Notably, environmental concerns (Demmler et al., 2021).
Lee et al. (2022) have demonstrated an implementation of AI which AI outcomes have been shown to be long-lasting, sustainable, and
assessed pedestrian satisfaction with the physical and visual character­ positive for the environment, which has a favourable knock-on effect on
istics of the street infrastructure. These results illustrate how AI is used urban residents’ standard of living. For example, the reviewed papers
to support smart city urban planning procedures, particularly for urban indicated that the utilisation of AI can encompass more than responding
and infrastructure management processes. to environmental disasters and the effects of climate change (Abid et al.,
Another highly elaborate problem cities face is the management of 2021; Milojevic-Dupont & Creutzig, 2021). More specifically,
land use and transportation. As the urban environment of cities changes, AI-supported applications have been introduced to improve strategies
it impacts the management, organisation, and planning for new devel­ that detect, measure, and manage waste volumes in cities, which in turn
opment and related transportation infrastructure (Azad & Wang, 2021). can provide better outcomes at a much lower cost (Conley et al., 2022;
In sum, the results show that AI can effectively learn, predict, estimate, Gupta et al., 2022). Again, these results translate into improvements in
store, manage, and analyse data related to traffic and housing de­ the quality of the natural and physical environment, demonstrating the
velopments (Ito et al., 2020; Azad & Wang, 2021). These approaches to application potential of AI. Because of urban planning processes that
utilising AI lead to improved and more efficient means of enhancing land ensure the management and long-term sustainable development of
use planning, transportation planning, development, and management smart cities, the quality of life of the citizens is improved. Along with
(Kröl, 2016; Kouziokas., 2017; Anastasiou et al., 2019; Aqib et al., these studies, AI has also been used in urban studies and planning
2019). Moreover, these results unveiled the benefits of AI for public research to measure the urban heat island effects—e.g., Yu et al. (2020),
safety. For example, the use of AI in improving transportation man­ Guan et al. (2022).
agement policies such as crime risk prevention strategies in trans­ Appendix Table A1 presents a summary—including the aim, rele­
portation networks (Kouziokas, 2017; Nikitas et al., 2020). vance, technologies, and findings—of the 14 papers concentrating on AI
Ultimately, the results pointed to a common theme, namely, that the for urban environmental and disaster management.
utilisation of AI in urban and infrastructure management can empower
previous and traditional methods, specifically problem-solving-related 4.5. AI for urban monitoring and development control
activities. In turn, smart and sustainable development in the planning
of cities, especially within the domain of urban and infrastructure As smart cities continue to increase in both population and land size,
management, establishes a new change in its fundamental processes the scope and scale of monitoring and the subsequent control of urban
(Nikitas et al., 2020). It should be noted that while the results suggest development will become more demanding. AI technologies have been
positive AI utilisation in urban design and land use/transportation used to address urban monitoring as a tool to support development
management, it is not at all exclusive to it; Utilisation of AI within the control. At a neighbourhood scale, AI has been used to monitor and
urban and infrastructure management domain shows potential in the predict crime hotspots in cities (He & Zheng, 2021), allowing for policies
planning and management of healthcare, education, telecommunica­ and measures to be installed to help foster greater safety in areas of
tion, waste management, food supply, water, industrial and concern. Whereas a more common use of AI is to assist in the monitoring
energy-related infrastructures (Nikitas et al., 2020; Anthony Jnr, 2021). and analysis of land use and urban sprawl (Lan et al., 2021). Also, AI has
Thus, the results highlight AI’s role as an instrument and tool for suc­ been used to monitor and identify where road improvements are needed
cessful and efficient smart and sustainable city development. by means of remote sensing (Exner et al., 2020).
Appendix Table A1 presents a summary—including the aim, rele­ Productive land use throughout cities can support population in­
vance, technologies, and findings—of the 19 papers concentrating on AI crease and drive economic development and infrastructure projects.
for urban and infrastructure management. Land use planning and management is a fundamental aspect of urban
planning and is time-consuming and heavily reliant on data processing
4.4. AI for urban environmental and disaster management and manual expert knowledge. The use of AI in land use planning not
only supports efficiencies in urban planning practice but more impor­
The urban environmental and disaster management domain exam­ tantly helps improve the city. There have been various supervised and
ines the implications of various environmental hazards and emergencies unsupervised AI learning tools that can provide varying classifications
within cities, including weather events, effects of urban development, and analyse land use and some deep learning techniques (Lv et al., 2015)
and the impact of a city’s energy use on the natural and physical envi­ and other multimodal techniques (Srivastava et al., 2019) can automate
ronment. These issues must be addressed in urban planning for a smart the entire process. The analysis of the urban footprint can be difficult
city to maintain sustainable development and environmental quality. and time-consuming, nevertheless, AI can analyse local urban features in
According to the findings, AI has already been used to address many of the planning and design, including building configurations (Shen et al.,
the usual environmental concerns for management. Tree management 2009; Ding et al., 2020). This form of AI provides another layer of
(Timilsina et al., 2019), noise pollution (Mrówcyzyńska et al., 2019), analysis and information that can be used to understand the efficiency of
population growth (Mulligan, 2021), and air pollution (Demller et al., urban areas.

8
T.H. Son et al. Sustainable Cities and Society 94 (2023) 104562

As cities continue to grow, horizontally and/or vertically, so does the processes with AI technologies, not only helps in monitoring, analysing,
land area that needs to be managed for optimum performance. Inno­ and evaluating the urban environment but also improves the quality of
vative solutions help achieve faster and better land use analyses in cities, life for citizens and ensures the long-term sustainability of smart cities.
providing urban planners with a greater understanding of the range of One of the main challenges that urban planners face, is monitoring and
factors involved in establishing development controls for urban growth quantifying large data sets to assess specific environmental issues, such
that were not perceivable without the use of AI. The reviewed papers in as the management of negative externalities, natural disasters, and im­
this domain revealed that controlling urban sprawl and speculating provements to infrastructure (Abid et al., 2021).
about the effects can be difficult to predict, nonetheless, AI has been The results strongly point to the existence of a consensus on the
used to simulate urban growth through the analysis of land use patterns importance of urban planners and policymakers capitalising on the op­
and street networks (Shen et al., 2020) and comparative analysis of portunities of emerging AI technologies to enhance the sustainable
similar urban environments (Gharaibeh et al., 2020) could be used to development of cities. The review results reflected literature on the
simulate similar planning policies in other comparative urban contexts. critical and real-time analyses of machine learning techniques. Although
These AI technologies could be used to support development con­ urban environmental issues remain complex, the labour and resources
trols; there have been examples of AI use to measure the changes in can be effectively minimised with these AI technologies. Integrating AI
urban density (Ding et al., 2020) and the impact of urban infrastructure technology is linked to reducing the error rate and improving the overall
on the surrounding urban areas (Feng et al., 2018) and compliance of efficiency of existing systems (Shin et al., 2017). Coinciding with the
urban development with local planning codes (Heikkila & Blewett, research literature, these benefits ultimately lead to improving urban
1992). The integration of the efficiency of AI with the expertise of policy and planning tools in the smart and sustainable development of
humans can lead to great possibilities and innovations in how urban the city environment, improving the quality of life for citizens.
planners tackle complex urban problems. The ease of forecasting and On an interesting note, while the results demonstrate positive sus­
modelling urban solutions by AI, as demonstrated in the literature, can tainable impacts on citizens and the physical environment, further
result in positive solutions assuming that planners will implement these research on the development of AI models and algorithms is recom­
tools judiciously. The effective monitoring and control of urban areas mended, as well as the scope of planning applications. Even though the
can lead to sustainable city development using AI technologies (Kou­ results have demonstrated that public bodies and institutions can
metio Tekouabou et al., 2021). The effective and productive use of land develop adequate and sustainable urban environmental policies using
in cities relies on connections between urban monitoring and effective AI, the analysis of the different factors relative to the urban environ­
development controls to ensure sustainable growth in cities. Along with mental issue is required. This ensures that the results are at their upmost
these studies, Long et al. (2015), Liang et al. (2018), and Kaviari et al. accuracy (Bienvenido-Huertas et al., 2020). Once this is achieved, the
(2019) also provided insights into spatial optimisation, agent-based application of AI technology to environmental policy solutions will have
models, and urban growth boundaries that are amongst the key scalable potential, with the possibility of efficiently reaching urban en­
methods for AI-based urban development control. vironments at the local, state, and national levels (Milojevic-Dupont
Appendix Table A1 presents a summary—including the aim, rele­ et al., 2021).
vance, technologies, and findings—of the 13 papers concentrating on AI Opportunities in the use of AI to support the development, gover­
for urban monitoring and development control. nance, and management of cities are evident from the results of the
review. As the results suggest, AI can assist in tackling future problem-
5. Findings and discussion solving-related urban environment and infrastructure managerial pro­
cesses. Thus, collaboration of relevant stakeholders can be advocated as
5.1. Urban planning areas in which AI technologies are contemplated or a means of pushing the implementation of AI (Nikitas et al., 2020). The
applied results not only illustrate that numerous applications of AI can be
implemented in complex urban problems such as urban design and
With the ever-increasing amount of big data available in cities, there modelling of infrastructure, but with further improvements to the
is a range of urban planning areas in which AI can be applied. Land use technology, a more sustainable smart city can be achieved (Yerram
planning and the subsequent urban developments are the two predom­ et al., 2022). Conversely, as the technology is still regarded as relatively
inant planning areas that could provide the greatest scope to support new, additional research can be beneficial for its accessibility and in­
urban planners as part of their practice. Arguably, the task of classifying crease the likelihood of it being implemented at the different scales of
land use and zoning, and the analysis of proposed developments, is the urban and infrastructure management processes. In other words, a
most significant and time-consuming task that urban planners under­ healthy level of caution is needed before embarking on large-scale AI
take. The findings of this review suggest that big data and AI interven­ adoption and testing the possible externalities and capacity of the
tion could improve efficiency in these tasks (Khediri et al., 2021). The organisation and planners.
application of big data could have a broad reaching, positive impact on Additionally, the application of AI to support urban and infrastruc­
health, traffic, air quality, and the physical environment, creating a ture management processes is gaining application ground to improve the
more sustainable smart city. management, organisation, and planning of land use and transportation
Planning interventions can directly produce urban data that creates (Ito et al., 2020; Azad & Wang, 2021). More specifically, it can be
continued monitoring and control by urban planners. When AI becomes observed that by implementing AI as a support tool to the above­
more commonly used in planning practices, it is expected that these mentioned managerial activities, complex problems that concern many
interventions will create even greater amounts of big data that will stakeholders can be addressed more efficiently, reaching solutions and
require monitoring as part of the urban planning process. Therefore, any decisions sooner. This suggests that AI has the capabilities to generate
area that adopts AI technologies will have a need for a holistic inte­ and adapt to different agents and more importantly, agents involving the
gration of AI as part of the wider urban planning practice, addressing the sustainable development of smart cities. Of course, this leads to the
complexities associated with urban environments (Jha et al., 2021). ethical issues of AI, which have already been raised by the existing lit­
The review results indicate that AI technologies will offer promising erature—How we mitigate the ethical concerns of AI, especially as urban
applications in the future, to deal with the urban environmental and planning activities involve human-centred approaches (Ljubenkov,
disaster management of cities. The range of urban environmental and 2020). Since the technology would presumably be implemented through
disaster matters that AI technologies can be applied to include both the the volition of urban planners, the cognitive bias would need to be
natural and physical environment, supporting urban planning processes, acknowledged and presented. Perhaps, by taking a holistic approach, an
public policy creation, and emergency responses. Integrating planning agreed set of guidelines could be presented amongst the relevant

9
T.H. Son et al. Sustainable Cities and Society 94 (2023) 104562

stakeholders during urban planning practices and developments. This also driven from the complexity of sustainable development (Redclift,
would ensure that the processes remain transparent, legal, fair, and safe 2005; Klopp & Petretta, 2017; Fu et al., 2019) and urban reality (e.g.,
(Ljubenkov, 2020). urban shrinkage/rapid growth, and citizen/stakeholder behaviours,
As AI brings forth a new direction for urban planners and decision- along with socioeconomic, political and governance challenges) and the
makers to improve the processes for the development of a more sus­ difficulty of exploring human-environment interactions (Kashef et al.,
tainable and smarter city, an integrated, collaborative, and holistic 2021). The potential of AI in supporting the mechanism of sustainable
approach is needed. Agents of different disciplines, both public and urban change and human-urban interactions for future sustainable
private, would have to exhibit equal opportunity in fully understanding urban planning is of great importance, and hence should be a future
and utilising the technology. Even then, as there are different scales of development trend in urban AI.
governance and private companies, a major barrier would be that there
may be no real capability to implement innovation in existing processes 5.3. Insights, study limitations, and research directions
(Nili et al., 2022). Nevertheless, the technology exists, and future works
need to shift AI research and development to implementation in This review revealed a series of important insights for urban plan­
real-world urban and infrastructure practices. Not only does this present ning scholars, decision-makers, and practitioners. The first one is that
the opportunity of shaping a more sustainable smart city, but also con­ early adopters’ real-world AI applications in urban planning are paving
fronts the associated risks of AI head on, minimising them in the future. the way to wider local government AI adoption (Nili et al., 2022). This is
to say, in the infancy of AI in urban planning, learning from success and
5.2. How AI technologies are supporting or could support smart and failure examples is a treasured opportunity for local government plan­
sustainable development ning departments as it would lead to improving the prospects and
limiting the constraints of the AI technology and its integration in the
A range of AI technologies currently supports smart and sustainable urban policymaking and planning mechanisms (Wu & Silva, 2010;
development, including traffic system management, crime detection, air Wang et al., 2022).
pollution, energy management and water leakage detection as part of The second key insight is that a major contributor to achieving wider
smart city planning and management (Milton & Roumpani, 2019). The AI adoption for urban planning is to promote collaboration and part­
energy sector uses AI within smart cities most commonly (Nosratabadi nership between key stakeholders. Wider engagement of stake­
et al., 2020) and various technologies support the monitoring, analysis, holders—for example, ranging from local government chief technology
and application of planning processes to mitigate the effects of pollution officers, policymakers, and planners to development companies, service
in urban environments (Mrówczyńska et al., 2019). Efficient public providers, community organisations, and end-users—will help in not
transport networks can be optimised with deep learning applications only effective and efficient but also a responsible use of AI in planning
(Aqib et al., 2019) helping to promote sustainable transport options in (Urban et al., 2021; Deshpande & Sharp, 2022).
smart cities. The next insight relates to problems associated with data. AI is a data
Big data is often used to promote a range of spatial and land use hungry system, and the recent popularisation and availability of big data
modelling to support the preservation of natural environments (Zou is an integral element for effective AI utilisation in urban planning.
et al., 2021). AI can support a range of urban planning practices, helping Many local governments are aware of the necessity of big data for more
promote the integration of sustainable urban technologies, such as the accurate urban analytics and there are attempts to obtain such data from
use of IoT, (Gray et al., 2021) to achieve further efficiencies and support both in-house and outsourced channels (Pencheva et al., 2020; Watson
sustainable city governance and/or management (Lan et al., 2021). Due & Ryan, 2020).
to the scale of IoT data, AI can not only support city development but The fourth insight is that in such a complex system as that of our
also regional development while considering the impact of environ­ cities, tackling complicated planning decisions cannot be done without
mental and sociocultural sustainability issues (Bienvenido-Huertas the aid of advanced technology such as AI. The mistake, however, would
et al., 2020). Regional planning can be supported by machine learning to be seeing these intelligent technologies as systems that could run with
predict future regional growth and development (Mulligan, 2021). little human supervision or oversight. The misuse of AI across the globe
Especially the emerging applications integrating AI with IoT infra­ has demonstrated that the convergence of artificial and human intelli­
structure, in other words artificial intelligence of things (AIoT) (Zhang & gence is crucial to address urbanisation issues adequately and to
Tao, 2020), offer invaluable data-driven knowledge for planning both at achieving smart and sustainable development (Gauglitz, 2019; Araujo
urban and regional scales (Kuguoglu et al., 2021; Yang et al., 2021). et al., 2020; Yigitcanlar et al., 2022b).
Smart cities pose a range of complex problems for urban planners; AI Moreover, while AI has significant potential in urban planning
technologies have demonstrated their potential to promote sustainable practice, gaps and limitations remain within the research. There also
resolutions (Jha et al., 2021). In the future, there will be further op­ remains a significant gap between research and practice. The literature
portunities for more advanced AI technologies to analyse big data and provides a range of frameworks, theories, and applications, however,
assist in solving these complex problems (Ullah et al., 2020) through there is limited evidence of wider implementation of AI across the
better urban design solutions (Quan et al., 2019). This will require sector. Understandably, the fear of AI amongst the public and of its
greater collaboration with all urban planning disciplines (Ye et al., potential impact on the profession of urban planners and their role, is a
2022) and greater community-led collaboration and engagement (Cal­ significant factor in the reduced uptake of AI technology (Yigitcanlar
ixto et al., 2021). et al., 2022c). However, as big data increases and complex urban issues
The impact of autonomous cities has resulted in smart cities being continue to increase, it should be imperative that AI technologies be
shaped by AI technologies (Cugurullo, 2020) due to the ever-increasing adopted sooner than later to ensure successful and productive imple­
use and application of big data in urban planning processes (Jiang, mentation. On the other hand, AI utilisation without considering its
2020). Nevertheless, urban planners can continue to do more with AI risks, such as cybersecurity, privacy, bias, etc., and without skilling up
technologies to better support urban planning practices and the issues urban planners to understand those risks pose a serious threat. There is,
affecting cities now and into the future (Karvonen et al., 2020). There hence, an urgent need for developing innovation adoption processes,
will be many challenges on the way, however, the advancements in AI upskilling planners, and educating planners to become AI literate via
technologies will allow for a revolution in the urban environment and formal university education and/or professional development training.
planning practices (Javed et al., 2022). Unsurprisingly, the literature suggests that to achieve the sustainable
Having said that, the current gap between AI capabilities and their development of the urban environment, the knowledge gap between AI
planning applications to support smart and sustainable development is and its applications needs to be addressed and reduced. An increase in

10
T.H. Son et al. Sustainable Cities and Society 94 (2023) 104562

Fig. 6. Current landscape of algorithmic urban planning to achieve smart and sustainable development.

public awareness could support the best urban planning outcomes. sustainability, resilience, and equity in both the short and long terms.
Nevertheless, the emergence of AI technology in planning practices can Planners must be prepared for the changes this will bring to how the
seem too novel for many planning departments and practitioners. cities of the future are planned, designed, and managed (Sanchez, 2023).
Further practical explanations of AI are needed so that urban planners Nonetheless, for planners to be ready for the AI revolution, the following
and policymakers can efficiently utilise the technology (Ye et al., challenges should be overcome: (a) Fear and uncertainty; (b) The need
2021a). Moreover, one major detected limitation was that AI technol­ for new skills; (c) Changing data needs and considerations; (d) Unclear
ogies require the possession of high-quality data and training (Srivas­ goals; (e) Incorporating transparency and explainability; (f) Encoun­
tava et al., 2019; Exner et al., 2020). Therefore, it can be suggested that tering bias; (g) Addressing ethical issues that may arise with new
to surpass this limitation, the procurement of data relative to the urban methods and data (Sanchez, 2023).
issue need to be either institutionalised or accessible. Consequently, this In that perspective, mapping out the current understanding and
leads to legal implications associated with data protection and big data practice in algorithmic urban planning to achieve smart and sustainable
management. The public also has concerns regarding the data being development is critical. This paper attempts to bridge this gap. Below,
collected and the methods used for collection. Fig. 6 illustrates the existing knowledge and practice landscape of
Likewise, unbiased big data and adequate training with the correct algorithmic urban planning to achieve smart and sustainable develop­
criteria and rules/principles will help decrease the bias and increase the ment. The sizes of the SDG boxes represent the popularity and coverage
efficiency of AI operations. Besides, as cities become more data-rich, they have achieved in our literature analysis, as shown in Fig. 5. Two
better laws and regulations need to be implemented to ensure citizens SDGs (SDG5: Gender Equality and SDG 13: Climate Action) unfortu­
are protected from any privacy breaches. In turn, citizens are incenti­ nately were not covered in the reviewed 91 articles, which should not be
vised in participating and providing relevant data for the development taken as an indication of unimportance. Instead, these two SDGs are of
of urban solutions (Exner et al., 2020). This would lead to the betterment utmost importance in achieving the smart and sustainable development
of sustainable urban planning and AI practice. of cities, given that they represent the inclusion and grassroot and
Besides, AI theory and practice in the context of cities at large, and government action—that are the core foundations of change for good
particularly in urban planning, is still in its infancy, despite the (Yigitcanlar et al., 2019b).
increasing number of successful and in many cases not-so-successful While considering these insights, the following study limitations
examples. The reasons for this include the recent emergence of should be noted: (a) The selection of the search keywords may have led
responsible application perspectives for AI technologies and limited to the omission of some of the literature—possibly due to concentration
application opportunities due to a lack of funds or local technical on Scopus and WoS databases and unconscious selection bias; (b) The
knowledge or regulations (Yigitcanlar et al., 2022a). There is, however, review has not included grey literature, where the most recent tech­
a growing interest in urban policy and practice circles toward AI nology developments and industry practices are reported; (c) Urban
adoption in urban planning and the development of smart and sustain­ planning is a vast domain, hence, the study might have not been able to
able cities. Nevertheless, the current knowledge on how to adopt AI in cover all AI utilisation or adoption areas in the field; (d) Unintentional
urban planning mechanisms is opaque. As also stated by Sanchez bias in analysing the literature and interpreting the study results may
(2023), much of the current planning-related discussion about AI is have led to the omission of the impact of the reported findings; (e) The
related to smart city technologies used to capture and analyse data for review has been undertaken manually—as opposed to using dedicated
optimization processes, such as traffic management and energy pro­ scientometric or bibliometric software or data visualisation softwar­
duction; currently much less attention is being paid to the use of AI in e—which might result in some human error; (f) The study has not
urban planning and associated decision-making activities, including focused on generating an understanding of the research ecosystem—e.
scenario planning and generative designs. g., most active regions, institutions, and researchers/authors in the field,
Next, AI is expected to play an important role in future development and citation, co-authorship and collaboration trends, and so on; (g) In­
activities as well as in the planning and management of cities for sights offered in this paper are driven or generalised from the reviewed

11
 T.H. Son et al.
Table A1
List of reviewed publications and review summaries.
Author Year Title Outlet Aim Relevance Technology Finding Category SDG

Jiang et al. (2022) 2022 A survey on deep Remote Sensing To provide a review of Discusses the various deep Machine Learning, Contribute to the research and Urban data n/a
learning-based change deep-learning change learning structures and the Deep Learning, applicability of deep learning analytics and
detection from high- detection algorithms challenges in change Neural Networks in remote sensing technology. planning
resolution remote sensing using HR remote detection. decision support
images sensing images in the
detection of more
delicate changes.
Lei et al. (2022) 2022 SNLRUX++ for building IEEE Journal of To develop a new deep Explores urban planners’ Machine Learning, Improves the prediction Urban data 17
extraction from high- Selected Topics in learning network for use of building extraction to Deep Learning performance of high- analytics and
resolution remote sensing Applied Earth building extraction. analysis of demographic resolution and high-semantic planning
images Observations and and urban characteristics. feature maps with more stable decision support
Remote Sensing and accelerated network
convergences.
Li et al. (2022) 2022 POI detection of high-rise IEEE Transactions on To propose a new Contributes to the Machine Learning, Found that the proposed Urban data 9, 11,
buildings using remote Geoscience and semantic segmentation development of deep Deep Learning model demonstrated analytics and 17
sensing images: a Remote Sensing model using deep learning utilisation for promising results however planning
semantic segmentation learning for building urban and regional requires a large quantity of decision support
method based on extraction, focusing on planning processes. manually labelled data and
multitasking attention improving the further research is needed.
Res-U-Net extraction of the
boundaries of building
roofs and building
shapes.
Lyu et al. (2023) 2022 IF-City: intelligible fair IEEE Transactions on To propose an AI- Examines the opportunities Machine Learning Demonstrated and evaluated Urban data 1, 10
city planning to measure, Visualization and enabled interactive of utilising AI for assessing the usage and capabilities of analytics and
12

explain and mitigate Computer Graphics visual tool to assist fairness in the urban the proposed method for planning
inequality urban planners in environment. collaborative fair urban decision support
achieving equal access design involving AI and
to amenities that can human planning.
benefit different types of
groups.
Wang and Biljecki 2022 Unsupervised machine Cities To review the Describes how Machine Learning, Uncovered potential research Urban data n/a
(2022) learning in urban studies: applications of unsupervised learning such Deep Learning opportunities in unsupervised analytics and
a systematic review of unsupervised machine as deep learning is applied machine learning within planning
applications learning in urban and researched in urban studies. decision support
studies. urbanisation and regional
studies, built environment,
urban sustainability, and
urban dynamics.
Ye et al. (2022) 2022 MasterplanGAN: Environment and To reduce the time to Identifies an AI system that Machine Learning, Found the system could be Urban data 17
facilitating the smart Planning B: complete master produces master plan Generative useful across number of analytics and

Sustainable Cities and Society 94 (2023) 104562

rendering of urban master planning drawings and renderings automatically. adversarial design and planning planning
plans via generative renderings. networks, neural disciplines with further decision support
adversarial networks networks development.
Yu et al. (2022) 2022 A combined Remote Sensing To propose a deep Investigates how AI can Machine Learning, Found the proposed model Urban data 11, 17
convolutional neural learning model for improve urban land use Deep Learning outperformed other models, analytics and
network for urban land- complex and diverse classification. illustrating its effectiveness planning
use classification with GIS urban land-use and feasibility in land-use decision support
data classification. classification.
Yuan et al. (2022) 2022 Fine-grained IEEE Journal of To propose a deep Describes deep learning Machine Learning, Found that the deep learning- Urban data 11, 17
classification of urban Selected Topics in learning-based approaches capability in Deep Learning based classification algorithm analytics and
functional zones and Applied Earth approach for fine- classifying multispectral performs better than planning
landscape pattern Observations and grained urban and hyperspectral data. traditional supervised decision support
analysis using Remote Sensing functional zones
(continued on next page)
 T.H. Son et al.
Table A1 (continued )
Author Year Title Outlet Aim Relevance Technology Finding Category SDG

hyperspectral satellite mapping, enabling methods, as well as retain as

imagery: a case study of better landscape pattern much information as possible.
Wuhan analysis.
Calixto et al. (2021) 2021 A layered approach for Association for To evaluate how the Examines how AI can be a Machine Learning Found that AI needs to Urban data 11, 17
the data-driven design of Computer-Aided design of AI can be tool to support in the design improve their analytics and
smart cities Architectural Design improved to assist in and planning of a smart comprehensiveness to allow planning
Research in Asia planning smart cities city. for wider implementation decision support
from the bottom-up. across disciplines.
Chaturvedi and de 2021 Machine learning Urban Science To review which AI Identifies planning Machine Learning Noted various AI applications Urban data 11
Vries (2021) algorithms for urban land techniques are suitable problems, datasets, and and the potential for analytics and
use planning: a review for land use and spatial suitable AI techniques. integrating systems together planning
analyse. to improve outcomes. decision support
Khediri et al. (2021) 2021 improving intelligent land to assess the identifies a range of ai- multiple ai provide a broad overview of urban data 11, 17
decision making in urban opportunities of ai to based tools that support how big data is being utilised analytics and
planning: using machine use big data to design planning processes to as a part of urban planning planning
learning algorithms and plan cities. address urban problems. practices. decision support
Gray and Kovacova 2021 Internet of things sensors geopolitics, history, to identify the explores the opportunities machine learning noted the integration of urban data 9,11,
(2021) and digital urban and international opportunities, risks, and of smart cities integrating sustainable urban analytics and 17
governance in data- relations priorities of urban sustainable urban technologies, particularly iot, planning
driven smart sustainable technologies, including technologies. is beneficial to the sustainable decision support
cities machine learning-based urban governance networks.
analytics and smart
software systems, in
smart cities.
Jha et al. (2021) 2021 A review of AI for urban International Journal To develop AI that can Reviews where AI is lacking Machine Learning, Found that a range of Urban data 17
planning: towards of Business Analytics assist in resolving and proposes a set of AI that Naïve Bayes technologies provided analytics and
building sustainable challenging urban can support, decision Classifier suitable outcomes, however, planning
13

smart cities problems. making that is informed by greater integration between decision support
the physical and social the systems is needed to
aspects of the city. improve efficiency.
Kamrowska-Załuska 2021 Impact of AI-based tools Information To develop a land-cover Proposes a hybrid network, Machine Learning, Provides further Urban data 9, 11,
(2021) and urban big data semantic segmentation utilising two deep learning Deep Learning, understanding of efficient analytics and 17
analytics on the design model using a hybrid of models, for land cover Convolutional utilisation of deep learning planning
and planning of cities deep learning semantic segmentation Neural Networks techniques to provide decision support
techniques. using high-spatial accurate segmentation maps
resolution satellite images. compared to other methods.
Podrasa et al. (2021) 2021 Machine learning for land ISPRS International To propose a city-scale Provides insight to the Machine Learning, Found that the approach has Urban data n/a
use scenarios and urban Journal of Geo- approach to façade opportunities of AI for Deep Learning satisfied accuracy for building analytics and
design information colour measurement urban colour identification façade segmentation, colour planning
using AI methods. and urban renewal deviation and overall decision support
information. accuracy for building
functional classification

Sustainable Cities and Society 94 (2023) 104562

compared to other methods.
Wan and Shi (2021) 2021 Research on urban Security and Explores the feasibility Describes the use of AI in Machine Learning, Found than the AI-enabled Urban data 11, 17
renewal public space Communication in machine learning, urban design. Deep Learning methods are not accurate analytics and
design based on Networks deep learning, and other enough due to limited planning
convolutional neural AI-enabled image computer hardware and decision support
network model processing for urban algorithm.
design.
Wang and Lee (2021) 2021 Regional population Remote Sensing To assess AI capabilities Provides insight on deep Machine Learning, Explored the potential of deep Urban data 9, 11,
forecast and analysis in semantic learning approaches for Deep Learning learning-based approaches on analytics and 17
based on machine segmentation of light segmentation tasks. point clouds for urban planning
learning strategy detection and ranging scenarios, demonstrating the decision support
point clouds. promising future of AI in
(continued on next page)
 T.H. Son et al.
Table A1 (continued )
Author Year Title Outlet Aim Relevance Technology Finding Category SDG

providing more accurate and

efficient results.
Zhang et al. (2021) 2021 Development of a city- International To propose AI-enabled Describes a new type of Machine Learning, Addressed the challenging Urban data 11, 17
scale approach for façade Archives of the models in capturing and convolutional neural Convolutional task of providing automatic analytics and
colour measurement with Photogrammetry, aggregating urban networks model that can Neural Networks, building extraction from high- planning
building functional Remote Sensing and multi-scale contextual produce high-precision Deep Learning resolution remote sensing decision support
classification using deep Spatial Information information on urban building segmentation from imagery with the use of a new
learning and street view Sciences building. high resolution sensing convolutional neural network
images images and preserve segmentation model, which
detailed spatial information can greatly benefit urban
of small and complex planning, navigation, and
building images. disaster management.
Zou et al. (2021) 2021 Towards urban scene Heliyon To enhance the Integrates the use of Machine Learning, Contributed to the Urban data 11,
semantic segmentation temporal and spatial artificial neural network Artificial Neural improvement in AI-enabled analytics and 15, 17
with deep learning from modelling by training into cellular automata Network model predictions, assisting in planning
lidar point clouds: a case AI-enabled models in Markov chain to ensure the preservation of decision support
study in Baden- predicting land-use higher accurate estimates of ecosystems and sustainable
Württemberg, Germany changes. land-use changes. development of urban areas.
Bazan-Krzywoszańska 2020 City as a system Urban and Regional To utilise AI to assist in Demonstrates how AI and Deep Learning Found that increased Urban data 11, 16
et al. (2020) supported by artificial Planning effective urban data can be used in Interdisciplinary analytics and
intelligence management. planning decision-making. collaboration and sharing of planning
data can assist in urban decision support
planning practice.
Cugurullo (2020) 2020 Urban artificial Frontiers in To investigate how AI is Identifies a range of AI Multiple AI Defined autonomous cities Urban data 7, 11
intelligence: from Sustainable Cities shaping autonomous practices used in urban and the evolution AI will have analytics and
automation to autonomy cities. planning in smart cities. in these developing cities. planning
in the smart city decision support
14

Engin et al. (2020) 2020 Data-driven urban Journal of Urban To review how AI is Reviews real-time and Machine Learning Noted that further Urban data n/a
management: mapping Management being utilised in urban evidence-based analyse to investigation is needed to analytics and
the landscape management and complex urban problems. consider the challenges and planning
planning practices. ethical implications of AI decision support
based urban management.
Jiang (2020) 2020 Urban planning reform Journal of Physics To review how AI is Examines a range of Multiple AI Noted that AI will continue to Urban data 11
trend based on artificial supporting urban technologies that will impact urban planning as analytics and
intelligence planning reform in support urban planning more big data is accessed and planning
China. practices into the future. utilised in practice. decision support
Karvonen et al. (2020) 2020 Urban planning and the Urban Planning To review how urban Identifies smart Machine Learning Found that urban planners Urban data 9, 11,
smart city: projects, planners are planning urbanisation practices in have an opportunity to direct analytics and 16
practices, and politics smart cities. smart cities. technological innovation in planning
relation to urban governance decision support
of smart cities.
Meeran and Conrad 2020 Machine learning for Remote Sensing of To propose an AI- Proposes a deep learning- Machine Learning, Proposed multimodal model Urban data 17

Sustainable Cities and Society 94 (2023) 104562

Joyce (2020) comparative urban Environment enabled model in based approach to automate Deep Learning, performed more accurately analytics and
planning at scale: an automated land use land use mapping to Convolutional than other models by a large planning
aviation case study mapping prediction. minimise time and labour. Neural Networks margin, demonstrating AI’s decision support
capabilities and opportunities
in urban planning.
Schmidt and Kada 2020 Object detection of aerial IEEE Access To propose a deep Describes the assessment Machine Learning, Found that the deep learning Urban data 17
(2020) image using masque- learning method to capabilities of Deep Learning method assesses analytics and
region convolutional assess the quality of data OpenStreetMap through AI- OpenStreetMap accurately planning
neural network (masque in OpenStreetMap. enabled methods. and effectively and provides decision support
R-CNN) insight for urban planning
opportunities where it lacks
the data.
(continued on next page)
 T.H. Son et al.
Table A1 (continued )
Author Year Title Outlet Aim Relevance Technology Finding Category SDG

Nice et al. (2020) 2020 The “Paris-end” of town? Urban Science To propose a deep Improves current deep Machine Learning, Found that the proposed deep Urban data 17
Deriving urban typologies learning neural network learning methods in Deep Learning learning method performed analytics and
using three imagery types method to bridge the building extraction with a better in terms of training and planning
gap of efficiently proposed deep learning response time and quality decision support
rationalising the method. value compared to other deep
different level features learning methods.
in building extraction.
Qian et al. (2020) 2020 Coupling cellular Environment and To demonstrate how AI Identifies how AI can Machine Learning, Found the accuracy of Urban data 11
automata with area Planning B can be used to predict support early stages of Generative generative adversarial analytics and
partitioning and building configurations. design with potential Adversarial Network network will further support planning
spatiotemporal solutions. urban planning solutions. decision support
convolution for dynamic
land use change
simulation
Ullah et al. (2020) 2020 Applications of artificial Computer To review how AI has Identifies how types of AI Machine Learning & Found that with further Urban data 7, 11,
intelligence and machine Communications been used in smart cities have addressed many urban Deep Reinforcement technological advancements analytics and 15
learning in smart cities over time. planning problems. Learning the use of big data will have planning
greater applications to decision support
support smart cities.
Yao et al. (2020) 2020 Delineating mixed urban Complexity To propose a deep Provides insights into the Machine & Deep Found that the method was Urban data 8, 17
“jobs-housing” patterns at learning model to use of AI in spatial analysis. Learning, able to understand complex analytics and
a fine scale by using high determine the spatial urban and spatial structures. planning
spatial resolution remote- distribution pattern of decision support
sensing imagery jobs and housing.
Milton and Roumpani 2019 Accelerating urban International To review which AI is Identifies the use of AI to Multiple AI Noted a range of AI that have Urban data 11
(2019) modelling algorithms Conference on being used in support traffic system been used across smart cities analytics and
with artificial intelligence Inventive developing smart cities. management, crime domains. planning
15

Computation detection, air pollution, decision support

Technologies energy management and
water leakage detection as
part of smart city planning
and management.
Quan et al. (2019) 2019 Artificial intelligence- Environment and To demonstrate how Demonstrates how the Machine Learning, Noted better urban design Urban data 17
aided design: smart Planning B smart design framework integration of AI can assist Genetic Algorithm solutions by using AI as a analytics and
design for sustainable city can be used to design in human decision making supporting aid to the design planning
development sustainable urban that is not possible with process. decision support
development. knowledge-based AI.
Yao et al. (2019) 2019 A human-machine International Journal To propose an AI- Develops a deep learning, Machine Learning, Explored the potential of AI, Urban data 11, 17
adversarial scoring of Geographical enabled urban street view imagery and Deep Learning and although the proposed analytics and
framework for urban Information Science perception assessment iterative feedback method is found to be feasible, planning
perception assessment model. framework that efficiently efficient, and highly accurate, decision support
using street-view images and cost-effectively assess many limitations found need

Sustainable Cities and Society 94 (2023) 104562

urban perceptions. to be further researched to
ensure reliability and
accuracy.
Ameer and Shah 2018 Exploiting big data IEEE Conference on To manage and analyse Identifies the potential for IoT, Machine Found that machine learning Urban data 13
(2018) analytics for smart urban Vehicular Technology pollution in smart cities AI to produce real-time Learning, Neural can assign the relevant air analytics and
planning using IoT sensors data. processing with reduced Network quality index classification by planning
errors. unsupervised learning. decision support
Yang et al. (2018) 2018 Building extraction in Remote Sensing To develop an object Explores machine learning- Machine Learning, Found that the proposed Urban data 17
very high-resolution detection method of based artificial neural Deep Learning, method provides significant analytics and
imagery by dense- aerial image by using networks for machine Neural Network accuracy. This was benefitted planning
attention networks region convolutional object detection. by the increased image decision support
neural network. training and epoch time.
(continued on next page)
 T.H. Son et al.
Table A1 (continued )
Author Year Title Outlet Aim Relevance Technology Finding Category SDG

Zhang et al. (2018) 2018 Measuring human Landscape and Urban To propose a deep Describes how deep Machine Learning, Found that the proposed Urban data 11, 17
perceptions of a large- Planning learning-based learning approaches can Deep Learning model can assist in predicting, analytics and
scale urban region using approach in predicting measure perceptions of a analysing, and determining planning
machine learning human perception of place in large urban scales. perceptions of places. decision support
street view images.
Ferreira et al. (2015) 2015 Urbane: a 3D framework IEEE Conference on To propose a framework Identifies the social impacts Machine Learning Revealed that the framework Urban data 16
to support data driven Visual Analytics that analyses the social development will have on could be applied to other analytics and
decision making in urban Science and impact of a urban areas. cities with the assistance of planning
development Technology development during the readily available big data. decision support
design and decision-
making process.
Jain (2011) 2011 A review study on urban International Journal To review how AI has Identifies how AI is being Machine Learning, Noted that the different AI Urban data 11
planning & artificial of Soft Computing and been used in urban land used in urban planning Expert systems, techniques had varying results analytics and
intelligence Engineering dynamics. practices in smart cities. decision support and efficiencies. planning
systems & integrated decision support
systems
Koehler et al. (2009) 2009 Computer-based methods International To develop a model that Examines AI to analyse Embodied, Agent Found that a range of AI Urban data 11
for a socially sustainable Conference on provides sustainable spatial structures, based modelling applications can provide analytics and
urban and regional Computational planning strategies that residential segregation and insight into the social planning
planning Science and Its respond to social issues interactions between the sustainability of a city. decision support
Applications within cities. built environment and
socio-spatial stakeholders.
Turkienicz et al. (2008) 2008 Turkienicz, B., Gonçalves, International To develop an AI- Demonstrates the Machine Learning, Explored the capabilities of AI Urban data 11, 17
B. B., & Grazziotin, P. Conference on Asian enabled query engine to architecture, construction, Deep Learning, in assisting urban planning analytics and
(2008). CityZoom: a Language Processing facilitate urban and opportunities of an AI- Convolutional processes to sustainable planning
visualization tool for the planning process. enabled query engine, Neural Networks growth, such as informed decision support
assessment of planning opening future promises to decisions from large
16

regulations. International which planners can quantities of data.

Journal of Architectural capitalise the technology.
Computing, 6(1), 79-95.
Witlox (2005) 2005 Expert systems in land- Expert Systems with To review expert Identifies which types of AI Knowledge Base, Noted improvements to the Urban data 11
use planning: an overview Applications systems in urban have been used in urban Expert systems, identified AI, however, since analytics and
planning practices. planning practices. decision support publication large number of planning
systems & integrated these limitations have been decision support
systems addressed by advancements in
technology.
Balling et al. (1999) 1999 Multiobjective urban Transportation To use AI to understand Explores different AI Machine Learning, Found that both systems Urban data 9, 11,
planning using genetic Research Procedia current transportation applications efficiency in Genetic Algorithm & provided good outcomes, analytics and 17
algorithm trends and improve the analysing complex Artificial Immune however, a combined system planning
network transportation networks. System would allow for both systems’ decision support
benefits to be recognised.
Feng and Xu (1999) 1999 Hybrid artificial Expert Systems with For AI to suggest Identifies how AI can be Knowledge Base, Noted that integrating the Urban data 11, 17

Sustainable Cities and Society 94 (2023) 104562

intelligence approach to Applications alternative used when trying to solve Fuzzy Logic & Neural three technologies resulted in analytics and
urban planning developments based on complex urban planning Networks solving complex problems. planning
real time data. problems of smart cities. decision support
Polydorides and 1989 An expert system for the Ekistics To demonstrate how AI Identifies the practicability Knowledge Base, Noted that the AI is being Urban data 11, 17
Petropoulos (1989) evaluation of urban plans can evaluate a master of a master plan before it is Expert System more broadly tested in a analytics and
plan based on implemented to support current context and further planning
consistency, efficiency, sustainable development. evidence will be forthcoming. decision support
and feasibility of the
plan.
Tanic (1986) 1986 Urban planning and Computers, To propose an AI system Explores how AI can be used Knowledge Base, Presented an early form of Urban data 11, 17
artificial intelligence: the Environment and that will support urban in planning practice. Expert System knowledge-based AI. analytics and
URBYS system Urban Systems planning knowledge- planning
decision support
(continued on next page)
 T.H. Son et al.
Table A1 (continued )
Author Year Title Outlet Aim Relevance Technology Finding Category SDG

based decision-making
processes.
Fang et al. (2022) 2022 Incorporating planning Journal of Urban To propose the use of Develops an AI-enabled tool Machine Learning, Discussed the opportunities in Urban and 11, 17
intelligence into deep Technology deep neural networks that provides more realistic Conventional Neural the systematic and intuitive infrastructure
learning: a planning and planning guidance predictions of street Network, Deep integration of combining deep management
support tool for street to automate a context- configurations. Learning learning algorithms and
network design aware, learning-based, subjective planning
and user-guided street- knowledge.
network generator.
Javed et al. (2022) 2022 Future smart cities: Cities To identify the latest Discusses how the future Machine Learning, Described the requirements, Urban and 17
requirements, emerging technological technologies for smart cities Deep Learning, strategies, and challenges for infrastructure
technologies, advancements of AI for are going to revolutionise Neural Networks the implementation in future management
applications, challenges, future smart city the urban environment, and smart city technologies.
and future aspects application frameworks. the requirements needed in
the application and
implementation of
strategies.
Lee et al. (2022) 2022 A machine learning and Sustainability To understand the Identified pedestrian Machine Learning, Suggested a methodology Urban and 11
computer vision study of urban design of satisfaction with machine Deep Learning, SHAP using computer vision infrastructure
the environmental pedestrian-friendly learning techniques, techniques, such as machine management
characteristics of cities with the AI particularly the physical learning, to identify
streetscapes that affect analyses of the visual and visual characteristics of pedestrian satisfaction and
pedestrian satisfaction streetscape and the urban street landscapes. what the urban planning of
inclusion of urban streets should consider.
theory.
Yerram et al. (2022) 2022 Extraction and Journal of Imaging To propose a method Gives insight into the Machine Learning, Found that the model Urban and 9, 11,
calculation of roadway deep learning method to efficiency of AI-enabled Deep Learning improves existing models in infrastructure 17
17

area from satellite images calculate the area of models compared to extracting and calculating management
using improved deep roads covered in traditional methods. roadway areas from satellite
learning model and post- satellite images. images.
processing
Anthony Jnr (2021) 2021 A case-based reasoning AI & Society To achieve a sustainable Identifies the potential for Knowledge-Based, Advocated that in mitigating Urban and 1, 2, 3,
recommender system for society with urban planners/decision- Case-Based present challenges of infrastructure 4, 6, 7,
sustainable smart city stakeholders makers of the adoption in Reasoning urbanisation and improving management 11,
development strategically deciding supported AI-integrated smart city’s healthcare, 15, 16
smart city initiatives best practice transportation, education,
with the support of a recommendation and energy sectors, AI-supported
recommender system retention of smart city systems are applicable and
that promotes smart city initiatives. Providing should be adopted.
planning. insight on alternative
strategies to future urban
issues and better

Sustainable Cities and Society 94 (2023) 104562

understanding the
dimensions of a smart city.
Azad and Wang (2021) 2021 Land use change ontology ISPRS International To use distinct Utilises deep learning Machine Learning, Contributed to AI’s potential Urban and 11, 17
and traffic prediction Journal of Geo- expressions of how land methods and effective data Deep Learning in urban planning, specifically infrastructure
through recurrent neural information use changes relate to the mining computing to land use transportation management
networks: a case study in integration of temporal predict traffic flow and land planning, by demonstrating
Calgary, Canada land use information, use change impacts. that the deep learning
utilising deep neural approach predicts more
network to predict accurately and better than
traffic. other existing models.
Ibrahim et al. (2021) 2021 URBAN-i: from urban Environment and To introduce AI-enabled Develops a framework that Machine Learning, Provide urban planners a Urban and 1, 9,
scenes to mapping slums, Planning B frameworks in utilises a convolutional Deep Learning, better understanding of urban infrastructure 11, 17
transport modes, and neural network, resulting in structures including the management
(continued on next page)
 T.H. Son et al.
Table A1 (continued )
Author Year Title Outlet Aim Relevance Technology Finding Category SDG

pedestrians in cities using advancing urban a reliable a reliable Convolutional pedestrian, transport modes,
deep learning and modelling. framework. Neural Networks and settlement conditions,
computer vision categorising them as either
planned or unplanned
settlements.
Ito et al. (2020) 2020 A method for estimating The International To investigate transport, Identifies smart vehicles Machine Learning, Found AI has a transformative Urban and 9, 11
the number of households Archives of the AI and the smart city and the interaction with AI, Deep Learning ability but needs to operate infrastructure
in a region from the Photogrammetry, and their effect on urban IoT and Industry 4.0 as part within a responsible, user- management
number of buildings Remote Sensing and areas. of a smart city. centred, and sustainable
estimated by deep Spatial Information framework.
learning with the Sciences
adjustment of its number
using ancillary datasets:
case study in Djakarta
Law et al. (2020) 2020 Street-frontage-net: urban International Journal Develop further past Proposes a new and Machine Learning, Contributes to more efficient Urban and 11, 17
image classification using of Geographical experiments of a improved convolutional Deep Learning, and alternative methods using infrastructure
deep convolutional neural Information Science convolutional neural neural network model in Convolutional deep learning methods in management
networks network that evaluates successfully evaluating Neural Networks geographic information and
street frontage quality. street frontage as either urban design.
being active or blank.
Ljubenkov et al. (2020) 2020 Optimizing bike sharing Science of the Total To propose an AI- Compares and tests the Machine Learning, Found that the model Urban and 17
system flows using graph Environment enabled land use change proposed method Deep Learning, achieved good performance infrastructure
mining, convolutional model for decision traditional methods and Convolutional however due to identified management
and recurrent neural support and urban demonstrates the improved Neural Networks limitations, the proposed
networks planning. approach. model is not recommended for
the projection of longer
periods of time.
18

Nikitas et al. (2020) 2020 Artificial intelligence, Journal of Sensors To propose a Demonstrates the improved Machine Learning, Demonstrates the increasing Urban and n/a
transport, and the smart classification approach deep learning classification Deep Learning opportunities of AI by infrastructure
city: definitions and to an existing deep approach to other AI- demonstrating a new management
dimensions of a new learning model for enabled models, such as classification model against
mobility era detailed urban SVM and conventional old AI-enabled models.
mapping. neural networks, by
producing unvaried
mapping results and
preservation of details.
Anastasiou et al. 2019 ADMSv2: a modern International Journal To develop software Explores how CityZoom Probabilistic Noted that further Urban and 11
(2019) architecture for of Architectural that combines multiple effectively visualises urban Methods, Decision development could include infrastructure
transportation data Computing software applications areas as part of the planning support system greater AI functionality to management
management and analysis and further integration process. become an autonomous
to use AI to extend system.
functionality.

Sustainable Cities and Society 94 (2023) 104562

Aqib et al. (2019) 2019 Rapid transit systems: Sustainability To optimise the Identifies the challenges of Machine Learning, Contributed to the novel deep Urban and 9, 11
smarter urban planning spatiotemporal holistically analysing and Deep Learning, GIS, learning models, algorithms, infrastructure
using big data, in-memory planning of urban predicting urban metro Remote sensing, Big application, analysis methods, management
computing, deep transportation systems with a computing Data, Neural and software tools in the
learning, and GPUs networks, specifically, platform. Networks analysis of metro systems.
train-based rapid transit
systems.
Gora (2019) 2019 Designing urban areas Journal of Urban Use of algorithms to Examines how land use Machine Learning, Noted the plans still require Urban and 11, 17
using traffic simulations, Planning and map and assess land use maps can be assessed Genetic Algorithm decision theory when infrastructure
artificial intelligence and Development plans. against a range of objectives choosing a plan. management
acquiring feedback from to assist planners in
stakeholders decision-making.
(continued on next page)
 T.H. Son et al.
Table A1 (continued )
Author Year Title Outlet Aim Relevance Technology Finding Category SDG

Xie et al. (2019) 2019 OpenStreetMap data IEEE European To identify and predict Utilises both convolutional Machine Learning, Findings benefit urban Urban and 3
quality assessment via Technology and dynamic patterns of and artificial recurrent Deep Learning, planners and bike-sharing infrastructure
deep learning and remote Engineering bikes using neural neural network to find and Convolutional and companies by optimising the management
sensing imagery Management Summit networks. predict bike flows. Artificial Recurrent manual transportation of
Neural Network bikes by saving time and cost
with AI-enabled analysing
systems.
Boulange et al. (2018) 2018 Improving planning Journal of Transport To develop, apply, and Examines how AI can be Machine Learning, Indicated that the AI could be Urban and 3, 11
analysis and decision Geography trial a functional and used to support the Walkability PSS extended to health and infrastructure
making: the development practical walkability planning process. environmental issues to management
and application of a planning support system support real time decision
walkability planning with local spatial making.
support system planners.
Kouziokas (2017) 2017 The application of Conference on To use AI to identify Identifies potential crime Machine Learning, Indicated the model provided Urban and 11
artificial intelligence in Sustainable Urban crime risks in the risks allowing for greater Neural Networks high prediction accuracy that infrastructure
public administration for Mobility transportation network. transportation management could be used to develop a management
forecasting high crime policies to be introduced safe transportation network.
risk transportation areas and build safe communities.
in urban environment
Shin et al. (2017) 2017 Towards a deep learning Transportation To propose AI that will Investigates a Traffic Machine Learning, Noted that further testing is Urban and 9, 11
powered query engine for Research Procedia support traffic Simulation Framework to Neural Networks required but has potential for infrastructure
urban planning management and support traffic management greater efficiency for management
associated land use and its impact on the planners.
implications. associated land use.
Kröl (2016) 2016 The application of the International To develop an urban Identifies AI modelling that Machine Learning, Found the AI could be Urban and 11, 17
artificial intelligence Conference on modelling and can provide planners Deep Learning applicable to other planning infrastructure
methods for planning of Geographical visualisation tool to different scenarios when practices including zoning management
19

the development of the Information Systems reflect real time developing new city and trip calculations.
transportation network Theory, Applications planning interventions. interventions.
and Management
Conley et al. (2022) 2022 Using a deep learning Computers, To use vehicle mounted Generates insight on the Machine Learning, Improves the narrative in Urban 3, 6,
model to quantify trash Environment and cameras and deep opportunities of deep Deep Learning, deep learning-based environmental 11,
accumulation for cleaner Urban Systems convolutional neural learning trash detection Convolutional monitoring approaches by and disaster 14, 17
urban stormwater networks to monitor, models to efficiently Neural Networks revealing that the approach management
prioritise, and measure monitor trash that goes in provides greater data
trash inputs to local waterways and the ocean. collection, understanding of
waterways and the urban trash sources, and
ocean, in the stormwater regulatory
effectiveness of its requirements compared to
reduction. visual assessments.
Gupta et al. (2022) 2022 A deep learning approach Complex & Intelligent To propose a deep Utilises image classification Machine Learning, Contributed to the problem Urban 3, 6,
based hardware solution Systems learning-based through different Convoluting Neural solving in urban issues, such environmental 11,

Sustainable Cities and Society 94 (2023) 104562

to categorise garbage in hardware solution for convolutional neural Networks as garbage detection in and disaster 12, 17
environment quicker and efficient network systems in unhealthy environments. management
garbage classification. segregating garbage
between biodegradable and
non-biodegradable objects
at a base level.
Abid et al. (2021) 2021 Toward an integrated Sustainability To understand how AI is Explores how AI is used to Machine Learning, Found that AI has great Urban 11
disaster management used as part of the manage and respond to Deep Learning, GIS, potential to respond to environmental
approach: how artificial disaster management disasters that affect urban Remote sensing disasters, however, and disaster
intelligence can boost process of mitigation, areas. effectiveness is based on data management
disaster management preparedness, response, management and team
and recovery. success.
(continued on next page)
 T.H. Son et al.
Table A1 (continued )
Author Year Title Outlet Aim Relevance Technology Finding Category SDG

Demmler et al. (2021) 2021 A novel approach of PLOS One To utilise AI to provide Discusses the use of AI for Machine Learning, Provided an insight for policy Urban 3, 11,
creating sustainable urban planning air pollution optimisation. Deep Learning, makers and developers in environmental 13, 15
urban planning solutions solutions that addresses Neural Networks opportunities of effectively and disaster
that optimise the local air urban air pollution analysing the dynamics of management
quality and issues. population characteristics and
environmental equity in urban disparities to minimise
Helsinki, Finland: the air pollutants, accessibility,
CouSCOUS study protocol transportation, and
environmental impacts,
resulting into better city
planning models.
Khan et al. (2021) 2021 Deep hybrid network for Journal of King Saud To explore the Identifies the challenges, Machine Learning, Found several areas in Urban 11, 17
land cover semantic University Computer application of machine opportunities, and future Deep Learning, machine learning models can environmental
segmentation in high- and Information learning in mitigating research directions of Neural Networks be applied in urban planning, and disaster
spatial resolution satellite Sciences challenges of modern machine learning methods including but not limited to, management
images urban planning. for intelligence urban the prediction of urban
planning applications. sustainability issues, the
smart, digital and connect
creation of cities, the
development of sustainable
urban forms, and the use of
urban space.
Milojevic-Dupont and 2021 Machine learning for Sustainable Cities and To use an AI framework Identifies AI that can inform Machine Learning, Low-carbon urban Urban 9, 11,
Creutzig (2021) geographically Society that provides policy the improvement of urban Remote Sensing infrastructure would greatly environmental 15
differentiated climate solutions to urban areas, infrastructure. benefit cities; however, the and disaster
change mitigation in streets, buildings, and use of AI and well-designed management
urban areas household contexts. policy is needed to resolve this
20

issue.
Mulligan (2021) 2021 Computationally Entropy To propose a machine Describes the use of Machine Learning, Found that the approach Urban 11, 17
networked urbanism and learning based method machine learning to close Deep Learning objectively observes features environmental
advanced sustainability to analyse and forecast any gaps including bias. of importance in present and and disaster
analytics in internet of population growth. future regional growth and management
things-enabled smart city provides an objective
governance reference for urban and
regional planning.
Pham et al. (2021) 2021 Flood risk assessment Journal of Hydrology To develop a flood risk Explores the AI or Machine Machine Learning, Found that the approach is Urban 11,
using hybrid artificial assessment map using Learning method Deep Learning, suitable and reliable in flood environmental 15, 17
intelligence models Multi Criteria Decision integration with MCDA to Neural Network risk mapping as it reduces and disaster
integrated with multi- Analysis method, with minimise human bias, time series meteorological, management
criteria decision analysis integration of AI- estimating the relevant and streamflow data.
in Quang Nam Province, enabled models. criteria in the decision- However, this approach
Vietnam making processes for flood cannot analyse frequencies of

Sustainable Cities and Society 94 (2023) 104562

risk assessment and flooding in the mapping and
frameworks. was only tested in one
location.
Ye et al. (2021) 2021 Towards an AI-driven Computational Urban To demonstrate how AI Identifies a framework that Machine Learning, Created a query tool that Urban 11,
framework for multi-scale Science can facilitate flood bridges the gap between the Convolutional could be useful for multiple environmental 15, 17
urban flood resilience resistance when planning and design Neural Networks disciplines in the creation of and disaster
planning and design planning and designing disciplines when resilient cities. management
in coastal areas. developing new coastal
urban areas.
Bienvenido-Huertas 2020 Comparison of artificial Sustainable Cities and To monitor Explores how public bodies Machine Learning, Suggested a methodology Urban 11,
et al. (2020) intelligence algorithms to Society sustainability indicators can manage environmental Regression using M5P and multilayer environmental 16, 17
estimate sustainability of a region so that and sociocultural issues algorithms, multiple perceptions provided the best and disaster
indicators policies can be linear regressions, indication of sustainability. management
(continued on next page)
 T.H. Son et al.
Table A1 (continued )
Author Year Title Outlet Aim Relevance Technology Finding Category SDG

established to advance pertaining to sustainability multilayer

sustainable through the assistance of AI. perceptron
development.
Nosratabadi et al. 2020 State of the art survey of International To investigate the Identifies the most used AI Machine & Deep Noted that the energy sector Urban 7, 11
(2020) deep learning and Conference on Global different types of applications in smart cities. Learning, Artificial used AI the most often and AI environmental
machine learning models Research and machine and deep neural networks, of varying applications have and disaster
for smart cities and urban Education learning used in current support vector been used across a range of management
sustainability urban planning machines, domains of a smart city.
practices in smart cities. Ensembles,
Bayesians, hybrids,
and neuro-fuzzy,
deep learning
Mrówczyńska et al. 2019 The use of artificial Sustainability To use AI to track noise Identifies noise pollution by Machine Learning, Found numerous planning Urban 11
(2019) intelligence as a tool pollution and apply it to using AI to plan for a Support Vector conventions that could be environmental
supporting sustainable local planning policies sustainable urban city Machines implemented in cities to and disaster
development local policy reduce noise pollution by management
using AI
Timilsina et al. (2019) 2019 Mapping urban trees ISPRS Annuals of the To review remote Identifies how AI Machine Learning, Found that a combination of Urban 11
within cadastral parcels Photogrammetry, sensing AI’s technology can be used to Convolutional CNN and OBIA methods environmental
using an object-based Remote Sensing and effectiveness of urban assist in urban management Neural Networks provided more accurate and disaster
convolutional neural Spatial Information tree management. to support sustainability of outcomes. management
network Sciences cities.
Miao et al. (2017) 2017 City afforestation: SIGSPATIAL To evaluate the Identifies how AI can help Machine Learning, Found that this AI could be Urban 3, 9,
abstracting the urban International effectiveness of an AI improve traffic forecasting Deep Learning transferrable to the environmental 11, 13
geometries into tree Conference on based transportation and bus arrival time management of air quality and disaster
structures for urban fabric Advances in management system. estimation through real- and weather. management
optimization Geographic time and historical data.
21

Information Systems
He and Zheng (2021) 2021 Prediction of crime rate in Engineering To use AI to identified Identifies areas of concern Machine Learning, Noted that the information Urban 11
urban neighborhoods Applications of crime hotspots in cities. and introduces measure for Generative identified could be used in monitoring and
based on machine Artificial Intelligence the prevention of crime in Adversarial Network developing strategies to development
learning urban areas. support master planning control
interventions.
Koumetio Tekouabou 2021 Reviewing the application Technology To build a healthy and Proposes a deep learning Machine Learning, Proposed model efficiently Urban 11, 15
et al. (2021) of machine learning Forecasting and effective regional and neural network Deep Learning, analyses the urban sprawl monitoring and
methods to model urban Social Change economy with green, algorithm to analyse urban Neural Networks environment, contributing development
form indicators in healthy, and sustainable sprawl levels. significant value to urban and control
planning decision support urban development. regional value, specifically, in
systems: potential, issues urban sprawl analysis
and challenges systems.
Lan et al. (2021) 2021 Constructing urban Geopolitics, History, To provide an Identifies the benefits of Machine Learning, Provide insight in the smart Urban 11, 17
sprawl measurement and International understanding of how deep learning techniques in Deep Learning sustainable city governance monitoring and

Sustainable Cities and Society 94 (2023) 104562

system of the Yangtze Relations smart sustainable city smart city paradigms. and management of the development
River economic belt zone governance and development in AI-enabled control
for healthier lives and management optimise technologies.
social changes in IoT with AI
sustainable cities technologies.
Luca et al. (2021) 2021 A survey on deep learning ACM Computing To explore the Examines perspectives on Machine Learning, Found that net-location and Urban 9, 17
for human mobility Surveys predictive and deep learning approaches to Deep Learning crowd flow predictions, and Monitoring and
generative tasks of human mobility. trajectory and flow generation Development
human mobility and the can be improved with deep Control
significance of deep learning approaches
learning model compared with traditional
approaches. models.
(continued on next page)
 T.H. Son et al.
Table A1 (continued )
Author Year Title Outlet Aim Relevance Technology Finding Category SDG

Ding et al. (2020) 2020 P-linknet: linknet with International To test the effectiveness Proposes a new method for Machine Learning, Demonstrated the practicing Urban 11, 17
spatial pyramid pooling Archives of the of deep learning using estimating the density Deep Learning use potential in spatial monitoring and
for high-resolution Photogrammetry, satellite/aerial images within a 100-metre grid cell analysis. Although the paper development
satellite imagery Remote Sensing and in spatial analysis. radius using deep learning suggests that the deep control
Spatial Information and adjustments from learning within this model
Sciences statistical data. needs to be improved.
Gharaibeh et al. (2020) 2020 Improving land-use Urban Science To demonstrate how AI Explores how AI can be used Machine Learning, Found that the AI successfully Urban 11
change modelling by can compare urban as part of planning practice Neural networks compares cities and can monitoring and
integrating ANN with typologies of different to inform processes. provide context to a local area development
cellular automata-Markov cities. by referencing similar control
chain model planning policy in similar
contexts when planning urban
areas.
Shen et al. (2020) 2020 Machine learning assisted International To use AI to simulate Investigates how AI can Machine Learning, Found the AI simulation was Urban 11
urban filling Conference on urban growth. support urban growth Cellular automata comparable to real urban monitoring and
Computer-Aided planning by considering spaces and has the potential to development
Architectural Design land use patterns and street assist in future urban growth control
Research in Asia networks. planning.
Srivastava et al. (2019) 2019 Understanding urban International To demonstrate how Identifies how AI can be Machine Learning, Found that supervised Urban 11, 17
land use from the above Conference on Urban supervised and part of the planning process Neural Networks learning while time monitoring and
and ground perspectives: Development, unsupervised learning and not just analysis of the consuming provides design development
a deep learning, Regional Planning can complete land-use design phase of urban solutions, whereas control
multimodal solution and Information scenarios. development. unsupervised learning
Society provided analysis of land use.
Feng et al. (2018) 2018 Urban zoning using Association for To understand the Identifies and categorises Machine Learning, Found that using AI across a Urban 11, 17
higher-order Markov Computer Aided impact of airports on the airports based on satellite Neural Networks continent provided greater monitoring and
random fields on multi- Design in surrounding urban imagery to understand the information that what would development
22

view imagery data Architecture context. impact to local urban areas have been possible at a human control
Conference and informs planners of the scale.
required actions needed to
resolve issues.
Lv et al. (2015) 2015 Urban land use and land European Conference To develop AI model for Explores multiple imagery Machine Learning, Noted the framework Urban 11, 17
cover classification using on Computer Vision urban zoning based on datasets that support land Markov Random completed land use zoning monitoring and
remotely sensed SAR data multiple imagery use zoning practices. Fields practices automatically and development
through deep belief datasets. reduces the time needed for control
networks urban planners to undertake
the process.
Shen et al. (2009) 2009 Geosimulation model International To develop a prototype Identifies how AI can Machine Learning, Found the prototype to be Urban 11, 17
using geographic Conference on AI that can synthesise analyse street networks and Cognitive Design useful but there is a large monitoring and
automata for simulating Computers in Urban urban features. land use to inform planning Computing amount of work required to development
land-use patterns in urban Planning and Urban practices. optimise the application. control
partitions Management

Sustainable Cities and Society 94 (2023) 104562

Heikkila and Blewett 1992 Using expert systems to Journal of the To assess how expert Examines how AI can be Logic & Knowledge Noted that the system could Urban 11, 17
(1992) check compliance with American Planning systems can be used in effectively used in statutory Based, Expert System be modified to meet each local monitoring and
municipal building codes Association local governments to planning to support government’s preferred development
assess developments enforcement of planning decision-making guidelines control
adherence to building policies. and staff implications.
codes.
Exner et al. (2020) 2020 Monitoring street Real Corp 2020: To use AI to monitor Identifies road Machine Learning, Noted the aspirational goals of Urban 11, 17
infrastructures with Shaping Urban road infrastructure. improvements without the IoT & Remote the framework are reliant on monitoring and
artificial intelligence Change Livable City need for physical inspection Sensing community participation development
Regions for the 21st using AI. which at the time of control
Century publication might not be
possible. However, there is
opportunity for
(continued on next page)
T.H. Son et al. Sustainable Cities and Society 94 (2023) 104562

literature along with supplementary literature cited/quoted; but the

authors extensive experience in fields related to this study might have

SDG
resulted in sharing some biased views.
Lastly, further research is needed to understand how AI is being used
in current planning practices and the opportunity for broader dissemi­
nation of information across the profession. Hence, it will be possible to
Category
configure the most suitable and responsible ways for its adoption in
planning. Current research is focused primarily on frameworks and
theories that have only been trialled in limited or defined urban con­
texts. Real-world applications are needed to support the introduction of
achievable in the long term.

AI properly to the urban planning practice (Kirwan & Zhiyong, 2020; As

framework in stages to be

& Basu, 2022; Matsuo & Iwamitsu, 2022).

implementation of the

6. Conclusion

As cities become more digitally transformed and continue to face

Finding

massive economic, social, environmental, and governance complexities,

there is a need for AI to support urban planning practices. Likewise, as
stated by Popelka et al. (2023, p.13), “the real impact of AI in cities is not
on the technology but on its implementation in urban planning and
design. It is in the plan-making process of cities that AI, e.g., in the form
of machine learning, has its major impact”. The study reported in this
Technology

paper suggests a range of current and emerging technologies that could

be utilised to meet these challenges, and advocates that responsible
algorithmic urban planning could support smart and sustainable
development. Nonetheless, interaction and collaboration between re­
searchers, planners, organisation, and communities—the key urban
decision stakeholders—are needed for a broader and more holistic
adoption of AI in our cities. Privacy, biases, and inequality continue to
be complex issues that need thoughtful consideration in the best way to
minimise these risks when implementing AI technologies for urban
Relevance

planning.
Moreover, the perceptions of both planners and the community will
have a significant role in how these challenges can be overcome, and
how adequate and responsible AI technology solutions could be adop­
ted. As the literature suggests, AI has been and can be used in a range of
urban planning tasks to address complex urban development challenges
and meet the needs and priorities of local communities. There is great
potential for AI to assist in improving the overall safety experience,
liveability, and sustainability of cities and their inhabitants in smarter
Aim

ways. Nevertheless, we also need to be aware of the shortcomings of AI

and find ways to tackle them. Finally, urban planners must understand
the benefits of AI methods that they will be using and effectively
communicate this to apprehensive stakeholders.
Furthermore, existing and emerging applications of AI in urban
planning could be useful to potentially pave the way to wider AI
Outlet

adoption. This is particularly likely in the case that present and future
capabilities and capacities of AI cater for urban planners’ needs. This
paper is an attempt to contribute to this issue. Likewise, we not only
believe in coupling different AI algorithms/applications—such as AI and
IoT—, but also in coupling human and artificial intelligences which have
the potential to help solve our complex urban planning problems.
Moreover, it is critical to start developing and actioning strategies for
responsible AI adoption in planning. In this perspective, as Sanchez
(2023) puts forward, a useful way to think about where and how AI
Title

might fit into everyday sustainable planning practice is to consider

strategic points of intervention that represent the work that planners do:
Year

(a) Community visioning; (b) Plan making; (c) Standards, policies, and
incentives; (d) Development work, and; (e) Public investments. Planners
can think about what tasks comprise the work of each of these points and
Table A1 (continued )

from there explore how they might be able to use AI to automate any of
those tasks. At the same time, planners can consider how data that is
generated through their planning work can be optimally collected and
configured to power AI analyses for better data-driven and
Author

evidence-based decision making for achieving sustainable outcomes.

We conclude the article with a quote from Professor Regina Barzilay

23
T.H. Son et al. Sustainable Cities and Society 94 (2023) 104562

Table B1 Table B1 (continued )

PRISMA 2000 checklist (http://prisma-statement.org/documents/PRISMA Section and Topic Item Checklist item Location where item is
_2020_checklist.pdf). # reported
Section and Topic Item Checklist item Location where item is 10b List and define all other n/a
# reported variables for which data
Title were sought (e.g.,
Title 1 Identify the report as a Title participant and
systematic review. intervention
Abstract characteristics, funding
Abstract 2 See the PRISMA 2020 for Abstract sources). Describe any
Abstracts checklist. assumptions made about
Introduction any missing or unclear
Rationale 3 Describe the rationale for Introduction information.
the review in the context Study risk of bias 11 Specify the methods used Methodology, and
of existing knowledge. assessment to assess risk of bias in the Findings and
Objectives 4 Provide an explicit Introduction included studies, Discussion
statement of the objective including details of the
(s) or question(s) the tool(s) used, how many
review addresses. reviewers assessed each
Methods study and whether they
Eligibility criteria 5 Specify the inclusion and Methodology worked independently,
exclusion criteria for the and if applicable, details
review and how studies of automation tools used
were grouped for the in the process.
syntheses. Effect measures 12 Specify for each outcome n/a
Information 6 Specify all databases, Methodology the effect measure(s) (e.
sources registers, websites, g., risk ratio, mean
organisations, reference difference) used in the
lists and other sources synthesis or presentation
searched or consulted to of results.
identify studies. Specify Synthesis methods 13a Describe the processes Methodology
the date when each source used to decide which
was last searched or studies were eligible for
consulted. each synthesis (e.g.,
Search strategy 7 Present the full search Methodology tabulating the study
strategies for all intervention
databases, registers and characteristics and
websites, including any comparing against the
filters and limits used. planned groups for each
Selection process 8 Specify the methods used Methodology synthesis (item #5)).
to decide whether a study 13b Describe any methods n/a
met the inclusion criteria required to prepare the
of the review, including data for presentation or
how many reviewers synthesis, such as
screened each record and handling of missing
each report retrieved, summary statistics, or
whether they worked data conversions.
independently, and if 13c Describe any methods Results
applicable, details of used to tabulate or
automation tools used in visually display results of
the process. individual studies and
Data collection 9 Specify the methods used Methodology syntheses.
process to collect data from 13d Describe any methods n/a
reports, including how used to synthesize results
many reviewers collected and provide a rationale
data from each report, for the choice(s). If meta-
whether they worked analysis was performed,
independently, any describe the model(s),
processes for obtaining or method(s) to identify the
confirming data from presence and extent of
study investigators, and if statistical heterogeneity,
applicable, details of and software package(s)
automation tools used in used.
the process. 13e Describe any methods n/a
Data items 10a List and define all Methodology used to explore possible
outcomes for which data causes of heterogeneity
were sought. Specify amongst study results (e.
whether all results that g. subgroup analysis,
were compatible with meta-regression).
each outcome domain in 13f Describe any sensitivity n/a
each study were sought analyses conducted to
(e.g., for all measures, assess robustness of the
time points, analyses), synthesized results.
and if not, the methods Reporting bias 14 Describe any methods n/a
used to decide which assessment used to assess risk of bias
results to collect. due to missing results in a
synthesis (arising from
reporting biases).
(continued on next page)

24
T.H. Son et al. Sustainable Cities and Society 94 (2023) 104562

Table B1 (continued ) Table B1 (continued )

Section and Topic Item Checklist item Location where item is Section and Topic Item Checklist item Location where item is
# reported # reported

Certainty 15 Describe any methods n/a Discussion 23a Provide a general Findings and
assessment used to assess certainty interpretation of the Discussion
(or confidence) in the results in the context of
body of evidence for an other evidence.
outcome. 23b Discuss any limitations of Findings and
Results the evidence included in Discussion
Study selection 16a Describe the results of the Methodology the review.
search and selection 23c Discuss any limitations of Findings and
process, from the number the review processes used. Discussion
of records identified in the 23d Discuss implications of Findings and
search to the number of the results for practice, Discussion, and
studies included in the policy, and future Conclusion
review, ideally using a research.
flow diagram. Other information
16b Cite studies that might n/a Registration and 24a Provide registration n/a
appear to meet the protocol information for the
inclusion criteria, but review, including register
which were excluded, and name and registration
explain why they were number, or state that the
excluded. review was not registered.
Study 17 Cite each included study Appendix Table A1 24b Indicate where the review n/a
characteristics and present its protocol can be accessed,
characteristics. or state that a protocol
Risk of bias in 18 Present assessments of n/a was not prepared.
studies risk of bias for each 24c Describe and explain any n/a
included study. amendments to
Results of 19 For all outcomes, present, Results information provided at
individual for each study: (a) registration or in the
studies summary statistics for protocol.
each group (where Support 25 Describe sources of Acknowledgements
appropriate) and (b) an financial or non-financial
effect estimate and its support for the review,
precision (e.g. and the role of the funders
confidence/credible or sponsors in the review.
interval), ideally using Competing 26 Declare any competing Acknowledgements
structured tables or plots. interests interests of review
Results of 20a For each synthesis, briefly Results authors.
syntheses summarise the Availability of 27 Report which of the Acknowledgements
characteristics and risk of data, code and following are publicly
bias amongst contributing other materials available and where they
studies. can be found: template
20b Present results of all Results data collection forms;
statistical syntheses data extracted from
conducted. If meta- included studies; data
analysis was done, used for all analyses;
present for each the analytic code; any other
summary estimate and its materials used in the
precision (e.g., review.
confidence/credible
interval) and measures of
statistical heterogeneity. of MIT’s Computer Science and Artificial Intelligence Laboratory
If comparing groups, (Sacan, 2022, p.16): “When the technology shifted from steam power to
describe the direction of
the effect.
electricity, the first attempts to bring electricity to industry were not
20c Present results of all Results very successful because people were just trying to copy steam machines.
investigations of possible I think something similar is now going on with AI. We need to figure out
causes of heterogeneity how to integrate it into many different areas: not only in healthcare, but
amongst study results.
also in education, in the design of materials, in urban planning, and so
20d Present results of all n/a
sensitivity analyses on. Of course, there is more to be done on the technological side,
conducted to assess the including making better algorithms, but we are bringing this technology
robustness of the into highly regulated environments, and we have not really looked at
synthesized results. how to do that yet”.
Reporting biases 21 Present assessments of n/a
risk of bias due to missing
results (arising from
reporting biases) for each
Declaration of Competing Interest
synthesis assessed.
Certainty of 22 Present assessments of n/a The authors declare that they have no known competing financial
evidence certainty (or confidence) interests or personal relationships that could have appeared to influence
in the body of evidence
the work reported in this paper.
for each outcome
assessed.
Discussion Data availability

Data will be made available on request.

25
T.H. Son et al. Sustainable Cities and Society 94 (2023) 104562

Acknowledgments Azad, A., & Wang, X. (2021). Land use change ontology and traffic prediction through
recurrent neural networks: A case study in Calgary, Canada. ISPRS international
Journal of Geo-information, 10(6), 358. https://doi.org/10.3390/ijgi10060358
This research was funded by the Australian Research Council Dis­ Balling, R., Taber, J., Brown, M., & Day, K. (1999). Multiobjective urban planning using
covery Grant Scheme, grant number DP220101255. The authors thank genetic algorithm. Journal of Urban Planning and Development, 125(2), 86–99. doi:10/
the editor and anonymous referees for their constructive comments. The bnm55h.
Bazan-Krzywoszańska, A., Lach, R., & Mrówczyńska, M. (2020). City as a system
authors declare no conflict of interest. The sources of the data used in supported by artificial intelligence. Urban and Regional Planning, 5(2), 32–39.
this paper are listed in Appendix Table A1 and available upon request. https://doi.org/10.11648/j.urp.20200502.11
Berawi, M. (2019). The role of industry 4.0 in achieving sustainable development goals.
International Journal of Technology, 10(4), 644–647. https://ijtech.eng.ui.ac.id/arti
Appendix A. ▒ cle/view/3341.
Bienvenido-Huertas, D., Farinha, F., Oliveira, M., Silva, E., & Lança, R. (2020).
Table A1 Comparison of artificial intelligence algorithms to estimate sustainability indicators.
Sustainable Cities and Society, 63, Article 102430. https://doi.org/10.1016/j.
scs.2020.102430
Appendix B. ▒ Boulange, C., Pettit, C., Gunn, L., Giles-Corti, B., & Badland, H. (2018). Improving
planning analysis and decision making: The development and application of a
walkability planning support system. Journal of Transport Geography, 69, 129–137.
Table B1 https://doi.org/10.1016/j.jtrangeo.2018.04.017
Calixto, V., Canuto, R., Noronha, M., Afrooz, A., Gu, N., & Celani, G. (2021). A layered
References approach for the data-driven design of smart cities. In International conference for the
association for computer-aided architectural design research in Asia.
Chaturvedi, V., & de Vries, W. (2021). Machine learning algorithms for urban land use
Abid, S., Sulaiman, N., Chan, S., Nazir, U., Abid, M., Han, H., et al. (2021). Toward an
planning: A review. Urban Science, 5(3), 68. https://doi.org/10.3390/
integrated disaster management approach: How artificial intelligence can boost
urbansci5030068
disaster management. Sustainability, 13(22), 12560. https://doi.org/10.3390/
Conley, G., Zinn, S. C., Hanson, T., Mcdonald, K., Beck, N., & Wen, H. (2022). Using a
su132212560
deep learning model to quantify trash accumulation for cleaner urban stormwater.
Ahmad, I., Alqurashi, F., Abozinadah, E., & Mehmood, R. (2022). Deep journalism and
Computers, Environment and Urban Systems,, 93, Article 101752. https://doi.org/
DeepJournal V1.0: A data-driven deep learning approach to discover parameters for
10.1016/j.compenvurbsys.2021.101752
transportation. Sustainability, 14(9), 5711. https://doi.org/10.3390/SU14095711
Corea, F. (2018). AI knowledge map: How to classify AI technologies. https://www.fo
Alahmari, N., Alswedani, S., Alzahrani, A., Katib, I., Albeshri, A., & Mehmood, R. (2022).
rbes.com/sites/cognitiveworld/2018/08/22/ai-knowledge-map-how-to-classify-ai
Musawah: A data-driven ai approach and tool to co-create healthcare services with a
-technologies.
case study on cancer disease in Saudi Arabia. Sustainability, 14(6), 3313. https://doi.
Cugurullo, F. (2020). Urban artificial intelligence: From automation to autonomy in the
org/10.3390/SU14063313
smart city. Frontiers in Sustainable Cities, 2, 38. doi:10/gh7wxj.
Alaql, A., AlQurashi, F., & Mehmood, R. (2022). Data-driven deep journalism to discover
Demmler, J., Gosztonyi, A., Du, Y., Leinonen, M., Ruotsalainen, L., Järvi, L., et al. (2021).
age dynamics in multi-generational labour markets from LinkedIn media. Preprints.
A novel approach of creating sustainable urban planning solutions that optimise the
doi:10.20944/preprints202210.0472.v1.
local air quality and environmental equity in Helsinki, Finland: The CouSCOUS study
Alkhayat, G., Hasan, S., & Mehmood, R. (2022). SENERGY: A novel deep learning-based
protocol. PloS One, 16(12), Article e0260009. https://doi.org/10.1371/journal.
auto-selective approach and tool for solar energy forecasting. Energies, 15(18), 6659.
pone.0260009
https://doi.org/10.3390/EN15186659
Deshpande, A., & Sharp, H. (2022). Responsible AI systems: Who are the stakeholders?.
Allam, Z., & Dhunny, Z. (2019). On big data, artificial intelligence and smart cities. Cities
In In: AAAI/ACM conference on AI, ethics, and society.
(London, England), 89, 80–91. https://doi.org/10.1016/j.cities.2019.01.03
Ding, Y., Wu, M., Xu, Y., & Duan, S. (2020). P-linknet: Linknet with spatial pyramid
Alomari, E., Katib, I., Albeshri, A., & Mehmood, R. (2021a). COVID-19: Detecting
pooling for high-resolution satellite imagery. International Archives of the
government pandemic measures and public concerns from Twitter Arabic data using
Photogrammetry. Remote Sensing & Spatial Information Sciences, 43(B3), 35–40.
distributed machine learning. International Journal of Environmental Research and
Engin, Z., van Dijk, J., Lan, T., Longley, P., Treleaven, P., Batty, M., et al. (2020). Data-
Public Health, 18(1), 282. https://doi.org/10.3390/ijerph18010282
driven urban management: Mapping the landscape. Journal of Urban Management, 9
Alomari, E., Katib, I., Albeshri, A., Yigitcanlar, T., & Mehmood, R. (2021b). Iktishaf+: A
(2), 140–150. doi:10/ghwvrh.
big data tool with automatic labelling for road traffic social sensing and event
Exner, J., Nalbach, O., & Werth, D. (2020). Monitoring street infrastructures with
detection using distributed machine learning. Sensors, 21(9), 2993. https://doi.org/
artificial intelligence. In Real Corp 2020: Shaping urban change livable city regions for
10.3390/s21092993
the 21st century.
Alotaibi, H., Alsolami, F., Abozinadah, E., & Mehmood, R. (2022). TAWSEEM: A deep-
Fang, Z., Jin, Y., & Yang, T (2022). Incorporating planning intelligence into deep
learning-based tool for estimating the number of unknown contributors in DNA
learning: A planning support tool for street network design. Journal of Urban
profiling. Electronics, 11(4), 548. https://doi.org/10.3390/ELECTRONICS11040548
Technology, 29(2), 99–114. https://doi.org/10.1080/10630732.2021.2001713
Alqahtani, E., Janbi, N., Sharaf, S., & Mehmood, R. (2022). Smart homes and families to
Feng, S., & Xu, L. (1999). An intelligent decision support system for fuzzy comprehensive
enable sustainable societies: A data-driven approach for multi-perspective parameter
evaluation of urban development. Expert Systems with Applications, 16(1), 21–32.
discovery using Bert modelling. Sustainability, 14(20), 13534. https://doi.org/
https://doi.org/10.1016/S0957-4174(98)00028-1
10.3390/SU142013534
Feng, T., Truong, Q., Nguyen, D., Koh, J., Yu, L., Binder, A., et al. (2018). Urban zoning
Alsaigh, R., Mehmood, R., & Katib, I. (2022). AI explainability and governance in smart
using higher-order Markov random fields on multi-view imagery data. Computer
energy systems: A review. arXiv, 2211.00069. doi:10.48550/arxiv.2211.00069.
vision – ECCV 2018 (pp. 627–644). Singapore: Springer.
Alswedani, S., Mehmood, R., & Katib, I. (2022). Sustainable participatory governance:
Ferreira, N., Lage, M., Doraiswamy, H., Vo, H., Wilson, L., Werner, H., et al. (2015).
Data-driven discovery of parameters for planning online and in-class education in
Urbane: A 3D framework to support data driven decision making in urban
Saudi Arabia During COVID-19. Frontiers in Sustainable Cities. https://doi.org/
development. In IEEE conference on visual analytics science and technology.
10.3389/FRSC.2022.871171
Frankenfield, J. (2022). What is artificial intelligence (AI)? https://www.investopedia.
Ameer, S., & Shah, M. (2018). Exploiting big data analytics for smart urban planning. In
com/terms/a/artificial-intelligence-ai.asp.
IEEE conference on vehicular technology.
Fu, B., Wang, S., Zhang, J., Hou, Z., & Li, J. (2019). Unravelling the complexity in
Anastasiou, C., Lin, J., He, C., Chiang, Y., & Shahabi, C. (2019). ADMSv2: A modern
achieving the 17 sustainable-development goals. National Science Review, 6(3),
architecture for transportation data management and analysis. In 21st SIGSPATIAL
386–388. https://doi.org/10.1093/nsr/nwz038
international conference on advances in geographic information systems.
Gauglitz, G. (2019). Artificial vs. human intelligence in analytics: Do computers
Andrews, C., Cooke, K., Gomez, A., Hurtado, P., Sanchez, T., Shah, S. et al. (2022). AI in
outperform analytical chemists? Analytical and Bioanalytical Chemistry, 411(22),
planning opportunities and challenges and how to prepare. https://www.planning.
5631–5632. https://doi.org/10.1007/s00216-019-01972-2
org/publications/document/9255930.
Geertman, S., Allan, A., Pettit, C., & Stillwell, J. (2017). Planning support science for
Anthony Jnr, B. (2021). A case-based reasoning recommender system for sustainable
smarter urban futures. New York: Springer.
smart city development. AI & Society, 36(1), 159–183. https://doi.org/10.1007/
Gharaibeh, A., Shaamala, A., Obeidat, R., & Al-Kofahi, S. (2020). Improving land-use
s00146-020-00984-2
change modellinging by integrating ANN with Cellular Automata-Markov Chain
Anyoha, R. (2017). The history of artificial intelligence. https://sitn.hms.harvard.edu/fla
model. Heliyon, 6(9). https://doi.org/10.1016/j.heliyon.2020.e05092. e05092-
sh/2017/history-artificial-intelligence.
e05092.
Aqib, M., Mehmood, R., Alzahrani, A., Katib, I., Albeshri, A., & Altowaijri, S. (2019).
Giuliani, G., Mazzetti, P., Santoro, M., Nativi, S., Van Bemmelen, J., Colangeli, G., et al.
Rapid transit systems: Smarter urban planning using big data, in-memory
(2020). Knowledge generation using satellite earth observations to support
computing, deep learning, and GPUs. Sustainability, 11(10), 2736. https://doi.org/
sustainable development goals (SDG): A use case on Land degradation. International
10.3390/su11102736
Journal of Applied Earth Observation and Geoinformation, 88, Article 102068. https://
Araujo, T., Helberger, N., Kruikemeier, S., & De Vreese, C. (2020). In AI we trust?
doi.org/10.1016/j.jag.2020.102068
Perceptions about automated decision-making by artificial intelligence. AI & Society,
Gora, P. (2019). Designing urban areas using traffic simulations, artificial intelligence
35(3), 611–623.
and acquiring feedback from stakeholders. Transportation Research Procedia, 41,
As, I., & Basu, P. (2022). Artificial intelligence in urban planning and design: Technologies,
532–534. doi:10/gnm5ds.
implementation, and impacts. New York: Elsevier.

26
T.H. Son et al. Sustainable Cities and Society 94 (2023) 104562

Gray, M., & Kovacova, M. (2021). Internet of things sensors and digital urban governance framework. Smart cities for technological and social innovation (pp. 9–28). London:
in data-driven smart sustainable cities. Geopolitics, History, and International Academic Press.
Relations, 13(2), 107–120. https://doi.org/10.22381/GHIR13220218 Kirwan, C., & Zhiyong, F. (2020). Smart cities and artificial intelligence: Convergent systems
Guan, Q., Yao, Y., Ma, T., Hong, Y., Bie, Y., & Lyu, J. (2022). Under the dome: A 3D urban for planning, design, and operations. New York: Elsevier.
texture model and its relationship with urban land surface temperature. Annals of the Klopp, J., & Petretta, D. (2017). The urban sustainable development goal: Indicators,
American Association of Geographers, 112(5), 1369–1389. https://doi.org/10.1080/ complexity and the politics of measuring cities. Cities (London, England), 63, 92–97.
24694452.2021.1972790 https://doi.org/10.1016/j.cities.2016.12.019
Gulshad, K., Wang, Y., Li, N., Wang, J., & Yu, Q. (2022). Likelihood of transformation to Koehler, H., Koenig, R., Kalisch, D., & Steinhoefel, J. (2009). Computer-based methods
green infrastructure using ensemble machine learning techniques in Jinan, China. for a socially sustainable urban and regional planning. Revista Internacional
Land, 11(3), 317. https://doi.org/10.3390/land11030317 Sostenibilidad, Tecnología y Humanismo,, 4, 115–124.
Gupta, T., Joshi, R., Mukhopadhyay, D., Sachdeva, K., Jain, N., Virmani, D., et al. (2022). Koumetio Tekouabou, S., Diop, E., Azmi, R., Jaligot, R., & Chenal, J. (2021). Reviewing
A deep learning approach based hardware solution to categorise garbage in the application of machine learning methods to model urban form indicators in
environment. Complex & Intelligent Systems, 8(2), 1129–1152. https://doi.org/ planning decision support systems: Potential, issues and challenges. Journal of King
10.1007/s40747-021-00529-0 Saud University – Computer and Information Sciences,, 34(8), 5943–5967. https://doi.
Hajjaji, Y., Boulila, W., Farah, I. R., Romdhani, I., & Hussain, A. (2021). Big data and IoT- org/10.1016/j.jksuci.2021.08.007
based applications in smart environments: A systematic review. Computer Science Kouziokas, G. (2017). The application of artificial intelligence in public administration
Review, 39, Article 100318. https://doi.org/10.1016/j.cosrev.2020.100318 for forecasting high crime risk transportation areas in urban environment.
Han, S., & Kim, T. (1989). Can expert systems help with planning? Journal of the Transportation Research Procedia, 24, 467–473. https://doi.org/10.1016/j.
American Planning Association, 55(3), 296–308. https://doi.org/10.1080/ trpro.2017.05.083
01944368908975417 Kröl, A. (2016). The application of the artificial intelligence methods for planning of the
He, J., & Zheng, H. (2021). Prediction of crime rate in urban neighborhoods based on development of the transportation network. Transportation Research Procedia, 14,
machine learning. Engineering Applications of Artificial Intelligence, 106, Article 4532–4541. doi:10/gmvhf5.
104460. doi:10/gnbkq6. Kuguoglu, B., van der Voort, H., & Janssen, M. (2021). The giant leap for smart cities:
Heikkila, E., & Blewett, E. (1992). Using expert systems to check compliance with Scaling up smart city artificial intelligence of things (AIOT) initiatives. Sustainability,
municipal building codes. Journal of the American Planning Association, 58(1), 72–80. 13(21), 12295. https://doi.org/10.3390/su132112295
doi:10/b5nbcz. Lăzăroiu, G., & Harrison, A. (2021). Internet of things sensing infrastructures and data-
Herath, H., & Mittal, M. (2022). Adoption of artificial intelligence in smart cities: A driven planning technologies in smart sustainable city governance and management.
comprehensive review. International Journal of Information Management Data Insights, Geopolitics, History & International Relations, 13(2), 23–36. https://doi.org/
2(1), Article 100076. https://doi.org/10.1016/j.jjimei.2022.100076 10.22381/GHIR13220212
Huang, Z. (2021). AI in urban planning: 3 ways it will strengthen how we plan for the Lăzăroiu, G., Ionescu, L., Andronie, M., & Dijmărescu, I. (2020). Sustainability
future. https://www.ura.gov.sg/Corporate/Resources/Ideas-and-Trends/AI-in management and performance in the urban corporate economy: A systematic
-Urban-Planning. literature review. Sustainability, 12(18), 7705. https://doi.org/10.3390/su12187705
Ibrahim, M., Haworth, J., & Cheng, T. (2021). URBAN-i: From urban scenes to mapping Lan, H., Zheng, P., & Li, Z. (2021). Constructing urban sprawl measurement system of the
slums, transport modes, and pedestrians in cities using deep learning and computer Yangtze River economic belt zone for healthier lives and social changes in
vision. Environment and Planning B, 48(1), 76–93. https://doi.org/10.1177/ sustainable cities. Technological Forecasting & Social Change, 165, Article 120569.
2399808319846517 https://doi.org/10.1016/j.techfore.2021.120569
Ito, R., Hara, K., Shimazaki, Y., Mori, N., Kani, Y., & Okabe, A. (2020). A method for Law, S., Seresinhe, C., Shen, Y., & Gutierrez-Roig, M. (2020). Street-Frontage-Net: Urban
estimating the number of households in a region from the number of buildings image classification using deep convolutional neural networks. International Journal
estimated by deep learning with the adjustment of its number using ancillary of Geographical Information Science, 34(4), 681–707. https://doi.org/10.1080/
datasets: Case study in Djakarta. The International Archives of the Photogrammetry, 13658816.2018.1555832
Remote Sensing and Spatial Information Sciences, XLIII-B4-2020, 673–678, doi:10 Lee, J., Kim, D., & Park, J. (2022). A machine learning and computer vision study of the
.5194/isprs-archives-XLIII-B4-2020-673-2020. environmental characteristics of streetscapes that affect pedestrian satisfaction.
Jain, K. (2011). A review study on urban planning & artificial intelligence. Artificial Sustainability, 14(9), 5730. https://doi.org/10.3390/su14095730
Intelligence, 1(5), 4. Legg, S., & Hutter, M. (2007). A collection of definitions of intelligence. Advances in
Janbi, N., Katib, I., Albeshri, A., & Mehmood, R. (2020). Distributed artificial artificial general intelligence (p. 17.24). Amsterdam: IOS Press.
intelligence-as-a-service (DAIaaS) for smarter IoE and 6G environments. Sensors, 20 Lei, Y., Yu, J., Chan, S., Wu, W., & Liu, X. (2022). SNLRUX++ for building extraction
(20), 5796. https://doi.org/10.3390/s20205796 from high-resolution remote sensing images. IEEE Journal of Selected Topics in Applied
Janbi, N., Mehmood, R., Katib, I., Albeshri, A., Corchado, J., & Yigitcanlar, T. (2022). Earth Observations and Remote Sensing, 15, 409–421. https://doi.org/10.1109/
IMTIDAD: A reference architecture and a case study on developing distributed AI JSTARS.2021.3135705
Services for skin disease diagnosis over cloud, fog and edge. Sensors, 22(5), 1854. Li, B., Gao, J., Chen, S., Lim, S., & Jiang, H. (2022). POI detection of high-rise buildings
https://doi.org/10.3390/S22051854 using remote sensing images: A semantic segmentation method based on multitask
Javed, A., Shahzad, F., Rehman, S., Zikria, Y., Razzak, I., Jalil, Z., et al. (2022). Future attention Res-U-Net. IEEE Transactions on Geoscience and Remote Sensing, 60, 1–16.
smart cities: Requirements, emerging technologies, applications, challenges, and https://doi.org/10.1109/TGRS.2022.3174399
future aspects. Cities (London, England), 129, Article 103794. https://doi.org/ Li, W. (2020). GeoAI: Where machine learning and big data converge in GIScience.
10.1016/j.cities.2022.103794 Journal of Spatial Information Science, (20), 71–77.
Jha, A., Ghimire, A., Thapa, S., Jha, A., & Raj, R. (2021). A review of AI for urban Liang, X., Liu, X., Li, X., Chen, Y., Tian, H., & Yao, Y. (2018). Delineating multi-scenario
planning: Towards building sustainable smart cities. In International conference on urban growth boundaries with a CA-based FLUS model and morphological method.
inventive computation technologies. Landscape and Urban Planning, 177, 47–63. https://doi.org/10.1016/j.
Jiang, H., Peng, M., Zhong, Y., Xie, H., Hao, Z., Lin, J., & Hu, X. (2022). A survey on deep landurbplan.2018.04.016
learning-based change detection from high-resolution remote sensing images. Liu, Y., Liu, X., Gao, S., Gong, L., Kang, C., Zhi, Y., et al. (2015). Social sensing: A new
Remote Sensing, 14(7), 1552. approach to understanding our socioeconomic environments. Annals of the
Jiang, M. (2020). Urban planning reform trend based on artificial intelligence. In In Association of American Geographers, 105(3), 512–530. https://doi.org/10.1080/
journal of physics: Conference series. 00045608.2015.1018773
Kaklauskas, A., Bardauskiene, D., Cerkauskiene, R., Ubarte, I., Raslanas, S., Radvile, E., Ljubenkov, D., Kon, F., & Ratti, C. (2020). Optimizing bike sharing system flows using
et al. (2021). Emotions analysis in public spaces for urban planning. Land use policy, graph mining, convolutional and recurrent neural networks. In 2020 IEEE European
107, Article 105458. https://doi.org/10.1016/j.landusepol.2021.105458 technology and engineering management summit.
Kamrowska-Załuska, D. (2021). Impact of ai-based tools and urban big data analytics on Long, Y., Han, H., Tu, Y., & Shu, X. (2015). Evaluating the effectiveness of urban growth
the design and planning of cities. Land, 10(11), 1209. boundaries using human mobility and activity records. Cities (London, England), 46,
Karvonen, A., Cook, M., & Haarstad, H. (2020). Urban planning and the smart city: 76–84. https://doi.org/10.1016/j.cities.2015.05.001
Projects, practices and politics. Urban Planning, 5(1), 65–68. https://doi.org/ Luca, M., Barlacchi, G., Lepri, B., & Pappalardo, L. (2021). A survey on deep learning for
10.17645/up.v5i1.2936 human mobility. ACM Computing Surveys, 55(1), 1–44. https://doi.org/10.1145/
Kashef, M., Visvizi, A., & Troisi, O. (2021). Smart city as a smart service system: Human- 3485125
computer interaction and smart city surveillance systems. Computers in Human Lv, Q., Dou, Y., Niu, X., Xu, J., Xu, J., & Xia, F. (2015). Urban land use and land cover
Behavior, 124, Article 106923. https://doi.org/10.1016/j.chb.2021.106923 classification using remotely sensed SAR data through deep belief networks. Journal
Kaviari, F., Mesgari, M. S., Seidi, E., & Motieyan, H. (2019). Simulation of urban growth of Sensors, 2015, 1–10. https://doi.org/10.1155/2015/538063
using agent-based odelling and game theory with different temporal resolutions. Lyu, Y., Lu, H., Lee, M. K., Schmitt, G., & Lim, B. Y. (2023). IF-city: Intelligible fair city
Cities (London, England), 95, Article 102387. https://doi.org/10.1016/j. planning to measure, explain and mitigate inequality. IEEE Transactions on
cities.2019.06.018 Visualization and Computer Graphics. https://doi.org/10.1109/TVCG.2023.3239909
Khan, S. D., Alarabi, L., & Basalamah, S. (2021). Deep hybrid network for land cover Mahendra, S. (2021). Artificial intelligence and urban design. https://www.aiplusinfo.
semantic segmentation in high-spatial resolution satellite images. Information, 12(6), com/blog/artificial-intelligence-and-urban-design.
230. Malibari, N., Katib, I., & Mehmood, R. (2022). Smart robotic strategies and advice for
Khediri, A., Laouar, M. R., & Eom, S. B. (2021). Improving intelligent decision making in stock trading using deep transformer reinforcement learning. Applied Sciences, 12
urban planning: using machine learning algorithms. International Journal of Business (24), 12526. https://doi.org/10.3390/APP122412526, 12526, 12.
Analytics, 8(3), 40–58. Matsuo, T., & Iwamitsu, S. (2022). Sustainable city planning and public administration
Kim, H., Sabri, S., & Kent, A. (2021). Smart cities as a platform for technological and assisted by green AI: Attendant legal challenges under Japanese law. Transforming
social innovation in productivity, sustainability, and livability: A conceptual

27
T.H. Son et al. Sustainable Cities and Society 94 (2023) 104562

Government: People, Process and Policy, 16(3), 334–346. https://doi.org/10.1108/TG- Schmidt, V., & Kada, M. (2020). Object detection of aerial image using mask-region
06-2021-0109 convolutional neural network (mask R-CNN). In , (Vol. 500, No. 1,. IOP conference
Meeran, A., & Conrad Joyce, S. (2020). Machine learning for comparative urban series: earth and environmental science. IOP Publishing.
planning at scale: An aviation case study. In Annual conference of the association of Shen, Z., Kawakami, M., & Kawamura, I. (2009). Geosimulation model using geographic
computer aided design in architecture. automata for simulating land-use patterns in urban partitions. Environment and
Miao, Y., Koenig, R., & Knecht, K. (2017). City afforestation: Abstracting the urban Planning B, 36(5), 802–823. doi:10/fbn2dz.
geometries into tree structures for urban fabric optimization. Weda51_35118. Shen, J., Liu, C., Ren, Y., & Zheng, H. (2020). Machine learning assisted urban filling. In
Micozzi, N., & Yigitcanlar, T. (2022). Understanding smart city policy: Insights from the CAADRIA conference.
strategy documents of 52 local governments. Sustainability, 14(16), 10164. https:// Shin, T. Y., Zihong, Y., Siong, N. W., Yangfan, Z., & Phangt, V. (2017). Towards a deep
doi.org/10.3390/su141610164 learning powered query engine for urban planning. In Proceedings of the international
Mikulsen, M., & Diduck, A. (2016). Towards an integrated approach to disaster conference on Asian language processing (pp. 99–102). IEEE.
management and food safety governance. International Journal of Disaster Risk Shin, T. Y., Zihong, Y., Siong, N. W., Yangfan, Z., & Phangt, V. (2017). December).
Reduction, 15, 116–124. https://doi.org/10.1016/j.ijdrr.2016.01.003 Towards a deep learning powered query engine for urban planning. In Proceedings of
Milojevic-Dupont, N., & Creutzig, F. (2021). Machine learning for geographically the international conference on Asian language processing (pp. 99–102). IEEE.
differentiated climate change mitigation in urban areas. Sustainable Cities and Society, Sideris, N., Bardis, G., Voulodimos, A., Miaoulis, G., & Ghazanfarpour, D. (2019). Using
64, Article 102526. doi:10/gnbkq3. random forests on real-world city data for urban planning in a visual semantic
Milton, R., & Roumpani, F. (2019). Accelerating urban modelling algorithms with decision support system. Sensors, 19(10), 2266. https://doi.org/10.3390/s19102266
artificial intelligence. In 5th International conference on geographical information Srivastava, S., Vargas-Muñoz, J., & Tuia, D. (2019). Understanding urban land-use from
systems theory, applications and management. the above and ground perspectives: A deep learning, multimodal solution. Remote
Mrówczyńska, M., Sztubecka, M., Skiba, M., Bazan-Krzywoszańska, A., & Bejga, P. Sensing of Environment, 228, 129–143. https://doi.org/10.1016/j.rse.2019.04.014
(2019). The use of artificial intelligence as a tool supporting sustainable Tanic, E. (1986). Urban planning and artificial intelligence: The URBYS system.
development local policy. Sustainability, 11(15), 4199. doi:10/gmdgv2. Computer, Environment and Urban Systems,, 10, 135–146. doi:10/bdvts4.
Mulligan, K. (2021). Computationally networked urbanism and advanced sustainability Tekouabou, S., Diop, E., Azmi, R., & Chenal, J. (2022). Artificial intelligence based
analytics in internet of things-enabled smart city governance. Geopolitics, History and methods for smart and sustainable urban planning: A systematic survey. Archives of
International Relations, 13(2), 121–134. https://doi.org/10.22381/GHIR13220219 Computational Methods in Engineering. https://doi.org/10.1007/s11831-022-09844-2
Nam, T., & Pardo, T. (2011). Conceptualizing smart city with dimensions of technology, Testi, I. (2021). Computer vision for urban regeneration. https://medium.com/urban-a
people, and institutions. In Annual international digital government research conference. i/computer-vision-for-urban-regeneration-dca4594f6317.
Nica, E. (2021). Urban big data analytics and sustainable governance networks in Timilsina, S., Sharma, S., & Aryal, J. (2019). Mapping urban trees within cadastral
integrated smart city planning and management. Geopolitics, History, and parcels using an object-based convolutional neural network. ISPRS Annals of the
International Relations, 13(2), 93–106. https://doi.org/10.22381/GHIR13220217 Photogrammetry, Remote Sensing and Spatial Information Sciences, 4, 111–117. doi:10/
Nice, K. A., Thompson, J., Wijnands, J. S., Aschwanden, G. D., & Stevenson, M. (2020). gnbkrb.
The “Paris-End” of town? Deriving urban typologies using three imagery types. Turkienicz, B., Gonçalves, B. B., & Grazziotin, P. (2008). CityZoom: A visualization tool
Urban Science, 4(2), 27. for the assessment of planning regulations. International Journal of Architectural
Nikitas, A., Michalakopoulou, K., Njoya, E., & Karampatzakis, D. (2020). Artificial Computing, 6(1), 79–95.
intelligence, transport and the smart city: Definitions and dimensions of a new Ullah, Z., Al-Turjman, F., Mostarda, L., & Gagliardi, R. (2020). Applications of artificial
mobility era. Sustainability, 12(7), 2789. https://doi.org/10.3390/su12072789 intelligence and machine learning in smart cities. Computer Communications, 154,
Nili, A., Desouza, K., & Yigitcanlar, T. (2022). What can the public sector teach us about 313–323. doi:10/ggwf8b.
deploying artificial intelligence technologies? IEEE Software, 39(6), 58–63. https:// Urban, A., Hick, D., Noennig, J., & Kammer, D. (2021). With a little help from AI: Pros
doi.org/10.1109/MS.2022.3193154 and cons of AI in urban planning and participation. International Journal of Urban
Nosratabadi, S., Mosavi, A., Keivani, R., Ardabili, S., & Aram, F. (2020). State of the art Planning and Smart Cities, 2(2), 19–33. https://doi.org/10.4018/
survey of deep learning and machine learning models for smart cities and urban IJUPSC.2021070102
sustainability. Engineering for Sustainable Future. Wan, J., & Shi, H. (2021). Research on urban renewal public space design based on
Pencheva, I., Esteve, M., & Mikhaylov, S. (2020). Big data and AI: A transformational convolutional neural network model. Security and Communication Networks, 2021,
shift for government: So, what next for research? Public Policy and Administration, 35 9504188.
(1), 24–44. https://doi.org/10.1177/0952076718780537 Wang, J., & Biljecki, F. (2022). Unsupervised machine learning in urban studies: A
Perveen, S., Kamruzzaman, M., & Yigitcanlar, T. (2017). Developing policy scenarios for systematic review of applications. Cities, 129, 103925.
sustainable urban growth management: A Delphi approach. Sustainability, 9(10), Wang, C. Y., & Lee, S. J. (2021). Regional population forecast and analysis based on
1787. https://doi.org/10.3390/su9101787 machine learning strategy. Entropy, 23(6), 656.
Pham, B., Luu, C., Phong, T., Nguyen, H., Le, H., Tran, T., et al. (2021). Flood risk Wang, Y., Zhang, N., & Zhao, X. (2022). Understanding the determinants in the different
assessment using hybrid artificial intelligence models integrated with multi-criteria government AI adoption stages: Evidence of local government chatbots in China.
decision analysis in Quang Nam Province, Vietnam. Journal of Hydrology, 592, Social Science Computer Review, 40(2), 534–554. https://doi.org/10.1177/
Article 125815. https://doi.org/10.1016/j.jhydrol.2020.125815 0894439320980132
Podrasa, D., Zeile, P., & Neppl, M. (2021). Machine Learning for Land Use Scenarios and Watson, R., & Ryan, P. (2020). Big data analytics in Australian local government. Smart
Urban Design. In International conference on urban development, regional planning and Cities, 3(3), 657–675. https://doi.org/10.3390/smartcities3030034
information society. Witlox, F. (2005). Expert systems in land-use planning: An overview. Expert Systems with
Polydorides, N., & Petropoulos, P. (1989). An expert system for the evaluation of urban Applications, 29(2), 437–445. doi:10/fmp745.
plans. Ekistics; reviews on the problems and science of human settlements, 56(338/339), Wu, N., & Silva, E. (2010). Artificial intelligence solutions for urban land dynamics: A
342–347. review. Journal of Planning Literature, 24(3), 246–265.
Popelka, S., Zertuche, L., & Beroche, H. (2023). Urban AI guide. https://urbanai. Xie, X., Zhou, Y., Xu, Y., Hu, Y., & Wu, C. (2019). Openstreetmap data quality assessment
fr/our-works/urban-ai-guide. via deep learning and remote sensing imagery. IEEE Access, 7, 176884–176895.
Press, G. (2016). A very short history of artificial intelligence (AI). https://www.forbes. Yang, H., Wu, P., Yao, X., Wu, Y., Wang, B., & Xu, Y. (2018). Building extraction in very
com/sites/gilpress/2016/12/30/a-very-short-history-of-artificial-intelligence-ai. high resolution imagery by dense-attention networks. Remote Sensing, 10(11), 1768.
Qian, Y., Xing, W., Guan, X., Yang, T., & Wu, H. (2020). Coupling cellular automata with Yang, X., Xu, Y., Kuang, L., Wang, Z., Gao, H., & Wang, X. (2021). An information fusion
area partitioning and spatiotemporal convolution for dynamic land use change approach to intelligent traffic signal control using the joint methods of multiagent
simulation. Science of the Total Environment, 722. https://doi.org/10.1016/j. reinforcement learning and artificial intelligence of things. IEEE Transactions on
scitotenv.2020.137738, 137738-137738. Intelligent Transportation Systems, 23(7), 9335–9345. https://doi.org/10.1109/
Quan, S., Park, J., Economou, A., & Lee, S. (2019). Artificial intelligence-aided design: TITS.2021.3105426
Smart design for sustainable city development. Environment and Planning B, 46(8), Yao, Y., Liang, Z., Yuan, Z., Liu, P., Bie, Y., Zhang, J., et al. (2019). A human-machine
1581–1599. doi:10/ggrx7n. adversarial scoring framework for urban perception assessment using street-view
Redclift, M. (2005). Sustainable development (1987-2005): An oxymoron comes of age. images. International Journal of Geographical Information Science, 33(12), 2363–2384.
Sustainable development, 13(4), 212–227. doi:10.1002/sd.281. https://doi.org/10.1080/13658816.2019.1643024
Regona, M., Yigitcanlar, T., Xia, B., & Li, R. (2022). Opportunities and adoption Yao, Y., Qian, C., Hong, Y., Guan, Q., Chen, J., Dai, L., et al. (2020). Delineating mixed
challenges of AI in the construction industry: A PRISMA review. Journal of Open urban “jobs-housing” patterns at a fine scale by using high spatial resolution remote-
Innovation: Technology, Market, and Complexity, 8(1), 45. https://doi.org/10.3390/ sensing imagery. Complexity, 2020, Article 8018629. https://doi.org/10.1155/
joitmc8010045 2020/8018629
Sacan, E. (2022). Inclusive artificial intelligence. https://www.fontys.nl/actueel/asset Ye, X., Wang, S., Lu, Z., Song, Y., & Yu, S. (2021). Towards an AI-driven framework for
/1142970/erdincsacan-inclusieveartificialintelligence. multi-scale urban flood resilience planning and design. Computational Urban Science,
Samsurijan, M., Ebekozien, A., Azazi, N., Shaed, M., & Badaruddin, R. (2022). Artificial 1(1), 11. doi:10/gnbkrf.
intelligence in urban services in Malaysia: A review. PSU Research Review. https:// Ye, X., Du, J., & Ye, Y. (2022). MasterplanGAN: Facilitating the smart rendering of urban
doi.org/10.1108/PRR-07-2021-0034 master plans via generative adversarial networks. Environment and Planning B, 49(3),
Sanchez, T., Shumway, H., Gordner, T., & Lim, T. (2022). The prospects of artificial 794–814. doi:10/gnm5dn.
intelligence in urban planning. International Journal of Urban Sciences. https://doi. Yerram, V., Takeshita, H., Iwahori, Y., Hayashi, Y., Bhuyan, M. K., Fukui, S., et al.
org/10.1080/12265934.2022.2102538 (2022). Extraction and calculation of roadway area from satellite images using
Sanchez, T. (2023). Planning with artificial intelligence in urban planning. American improved deep learning model and post-processing. Journal of Imaging, 8(5), 124.
Planning Association. https://doi.org/10.3390/jimaging8050124

28
T.H. Son et al. Sustainable Cities and Society 94 (2023) 104562

Yigitcanlar, T., & Teriman, S. (2015). Rethinking sustainable urban development: Yu, S., Chen, Z., Yu, B., Wang, L., Wu, B., Wu, J., et al. (2020). Exploring the relationship
Towards an integrated planning and development process. International Journal of between 2D/3D landscape pattern and land surface temperature based on
Environmental Science and Technology, 12(1), 341–352. https://doi.org/10.1007/ explainable eXtreme Gradient Boosting tree: A case study of Shanghai, China. Science
s13762-013-0491-x of the Total Environment, 725, Article 138229. https://doi.org/10.1016/j.
Yigitcanlar, T., Han, H., Kamruzzaman, M., Ioppolo, G., & Sabatini-Marques, J. (2019a). scitotenv.2020.138229
The making of smart cities: Are Songdo, Masdar, Amsterdam, San Francisco and Yu, J., Zeng, P., Yu, Y., Yu, H., Huang, L., & Zhou, D. (2022). A combined convolutional
Brisbane the best we could build? Land use policy, 88, Article 104187. https://doi. neural network for urban land-use classification with GIS data. Remote Sensing, 14
org/10.1016/j.landusepol.2019.104187 (5), 1128.
Yigitcanlar, T., Foth, M., & Kamruzzaman, M. (2019b). Towards post-anthropocentric Yuan, J., Wang, S., Wu, C., & Xu, Y. (2022). Fine-grained classification of urban
cities: Reconceptualizing smart cities to evade urban ecocide. Journal of Urban functional zones and landscape pattern analysis using hyperspectral satellite
Technology, 26(2), 147–152. https://doi.org/10.1080/10630732.2018.1524249 imagery: A case study of Wuhan. IEEE Journal of Selected Topics in Applied Earth
Yigitcanlar, T., Butler, L., Windle, E., Desouza, K., Mehmood, R., & Corchado, J. (2020a). Observations and Remote Sensing, 15, 3972–3991. https://doi.org/10.1109/
Can building “artificially intelligent cities” safeguard humanity from natural JSTARS.2022.3174412
disasters, pandemics, and other catastrophes? An urban scholar’s perspective. Zhang, J., & Tao, D. (2020). Empowering things with intelligence: A survey of the
Sensors, 20(10), 2988. https://doi.org/10.3390/s20102988 progress, challenges, and opportunities in artificial intelligence of things. IEEE
Yigitcanlar, T., Desouza, K., Butler, L., & Roozkhosh, F. (2020b). Contributions and risks Internet of Things Journal, 8(10), 7789–7817. https://doi.org/10.1109/
of artificial intelligence (AI) in building smarter cities: Insights from a systematic JIOT.2020.3039359, 0.
review of the literature. Energies, 13(6), 1473. https://doi.org/10.3390/en13061473 Zhang, F., Zhou, B., Liu, L., Liu, Y., Fung, H., Lin, H., et al. (2018). Measuring human
Yigitcanlar, T., Kankanamge, N., Regona, M., Ruiz Maldonado, A., Rowan, B., Ryu, A., perceptions of a large-scale urban region using machine learning. Landscape and
et al. (2020c). Artificial intelligence technologies and related urban planning and Urban Planning, 180, 148–160. https://doi.org/10.1016/j.landurbplan.2018.08.020
development concepts: How are they perceived and utilized in Australia? Journal of Zhang, D., Pee, L., Pan, S., & Liu, W. (2022). Orchestrating artificial intelligence for
Open Innovation: Technology, Market, and Complexity, 6(4), 187. https://doi.org/ urban sustainability. Government Information Quarterly, 39(4), Article 101720.
10.3390/joitmc6040187 https://doi.org/10.1016/j.giq.2022.101720
Yigitcanlar, T., Corchado, J., Mehmood, R., Li, R., Mossberger, K., & Desouza, K. (2021). Zhang, J., Fukuda, T., & Yabuki, N. (2021). Development of a city-scale approach for
Responsible urban innovation with local government artificial intelligence (AI): A façade color measurement with building functional classification using deep learning
conceptual framework and research agenda. Journal of Open Innovation: Technology, and street view images. ISPRS International Journal of Geo-Information, 10(8), 551.
Market, and Complexity, 7(1), 71. https://doi.org/10.3390/joitmc7010071 Zhou, M., Ma, Y., Tu, J., & Wang, M. (2022). SDG-oriented multi-scenario sustainable
Yigitcanlar, T., Agdas, D., & Degirmenci, K. (2022a). Artificial intelligence in local land-use simulation under the background of urban expansion. Environmental Science
governments: Perceptions of city managers on prospects, constraints and choices. AI and Pollution Research, 29(48), 72797–72818. https://doi.org/10.1007/s11356-022-
& Society. https://doi.org/10.1007/s00146-022-01450-x 20904-9
Yigitcanlar, T., Li, R., Inkinen, T., & Paz, A. (2022b). Public perceptions on application Zou, Y., Zhang, S., & Min, Y. (2019). Exploring urban population forecasting and spatial
areas and adoption challenges of AI in urban services. Emerging Sciences Journal, 6 distribution modeling with artificial intelligence technology. Computer Modeling in
(6), 1199–1236. https://doi.org/10.28991/ESJ-2022-06-06-01 Engineering & Sciences, 119(2), 295–310. https://doi.org/10.32604/
Yigitcanlar, T., Degirmenci, K., & Inkinen, T. (2022c). Drivers behind the public cmes.2019.03873
perception of artificial intelligence: Insights from major Australian cities. AI & Zou, Y., Weinacker, H., & Koch, B. (2021). Towards urban scene semantic segmentation
Society. https://doi.org/10.1007/s00146-022-01566-0 with deep learning from lidar point clouds: A case study in Baden-Württemberg,
Germany. Remote Sensing, 13(16), 3220. https://doi.org/10.3390/rs13163220

29