PreprintPDF Available

Multiscale Modeling in Smart Cities: A Survey on applications, Current Trends, and Challenges

Preprints and early-stage research may not have been peer reviewed yet.

Abstract and Figures

Megacities are complex systems facing the challenges of overpopulation, poor urban design and planning, poor mobility and public transport, poor governance, climate change issues, poor sewerage and water infrastructure, waste and health issues, and unemployment. Smart cities have emerged to address these challenges by making the best use of space and resources for the benefit of citizens. A smart city model views the city as a complex adaptive system consisting of services, resources, and citizens that learn through interaction and change in both the spatial and temporal domains. The characteristics of dynamic development and complexity are key issues for city planners that require a new systematic and modeling approach. Multiscale modeling (MM) is an approach that can be used to better understand complex adaptive systems. The MM aims to solve complex problems at different scales, i.e., micro, meso, and macro, to improve system efficiency and mitigate computational complexity and cost. In this paper, we present an overview of MM in smart cities. First, this study discusses megacities, their current challenges, and their emergence to smart cities. Then, we discuss the need of MM in smart cities and its emerging applications. Finally, the study highlights current challenges and future directions related to MM in smart cities, which provide a roadmap for the optimized operation of smart city systems.
Content may be subject to copyright.
Multiscale Modeling in Smart Cities: A Survey on
Applications, Current Trends, and Challenges
Asif Khana, Sheraz Aslamb,
, Khursheed Aurangzebc, Musaed Alhusseind,
Nadeem Javaide,f
aScience and IT Department, Government of Balochistan, Quetta 87300, Pakistan
bDepartment of Electrical Engineering, Computer Engineering, and Informatics, Cyprus
University of Technology, Limassol 3036, Cyprus
cCollege of Computer and Information Sciences, King Saud University, Riyadh 11543,
Saudi Arabia.(
dDepartment of Computer Engineering, Col lege of Computer and Information Sciences,
King Saud University, Riyadh 11543, Saudi Arabia. (
eDepartment of Computer Science, COMSATS University Islamabad (CUI), Islamabad
44000, Pakistan. (e-mail:
fSchool of Computer Science, University of Technology Sydney (UTS), Ultimo, NSW, 2007,
A smart city model views the city as a complex adaptive system consisting
of services, resources, and citizens that learn through interaction and change in
both the spatial and temporal domains. The characteristics of dynamic evolu-
tion and complexity are key issues for megacity planners and require a new sys-
tematic and modeling approach. Multiscale models involved in smart cities and
megacities have recently become a popular topic because they can understand
complex adaptive systems and efficiently solve complex problems at multiple
scales (i.e., micro, meso, and macro) to improve system efficiency and reduce
computational complexity and cost. However, there are numerous opportuni-
ties to improve this interdisciplinary field considering the lack of applicability
of the multiscale modeling approach in megacities and smart cities, and the
potential of multiscale modeling in various complex systems within smart cities.
Therefore, a review that summarizes the state-of-art researches and opens op-
portunities around the theme of multiscale modeling participating in megacities
Corresponding author
Email address: (Sheraz Aslam)
Preprint submitted to Sustainable Cities and Society November 16, 2021
and smart cities is warranted. This study, therefore, provides a comprehensive
review covering the introduction of megacities, their current challenges, and
their emergence in smart cities. Then, the introduction of the smart city along
with its characteristics and different generations is disclosed. Moreover, we shed
light on multiscale modeling, its categories (i.e., sequential multiscale modeling
and concurrent multiscale modeling), and the need for multiscale modeling in
megacities and smart cities along with its emerging applications. Finally, based
on a literature review, the study highlights the current challenges and future
directions related to multiscale modeling in megacities and smart cities, which
provide a roadmap for the optimized operation of megacities and smart city
Keywords: Multiscale modeling; Multiscale Systems; Megacities;
Smart Cities; Sustainable Cities; Multiscale Modeling
1. Introduction
The population is increasing everyday and more than 50% of the total popu-
lation is living in urban areas [1, 2, 3]. Urbanization is one of the most important
phenomena in today’s society and the rate of urbanization continues to increase;
therefore, the 21st century is often referred to as the century of cities [4, 5, 6].
The high pace of urbanization is the emergence of megacities. Around the globe,
there are 30 megacities with populations of around 10 million or more, larger
than several countries [7]. Basically, the megacities act as magnets that at-
tract residents from rural areas to work, seek job opportunities, and improve
their standard of living. They can earn more in cities than in rural and remote
areas. Although people move to cities because of a better standard of living
and many other opportunities, megacities are considered complex systems that
some megacities face challenges such as overpopulation, lack of and inefficient
urban design or planning, inefficient mobility and public transport, poor gov-
ernance, climate change issues, inefficient sewerage and water infrastructure,
waste and health issues, unemployment, unusually fragile infrastructure, the in-
creased crime rate in cities, environmental problems, and fierce competition due
to low resources [8, 9, 10, 11]. The concept of Smart Cities (SCs) is flourishing
in developed and modern countries. The key infrastructure elements present in
SCs are affordable and smart housing, sustainable power supply, efficient public
transport and urban mobility, solid waste management, sanitation, and factors
that contribute to a clean environment for citizens [12, 13]. In addition, robust
information technology and digitization, Internet of Things (IoT), Artificial In-
telligence (AI), and Machine Learning (ML) techniques are used to operate and
execute all the functions of a SC smoothly and efficiently [14].
The use of these components and services provided in a SC to citizens make it
a complex adaptive system [15, 16]. In a complex adaptive system, the elements,
also called agents, are not fixed and they learn and adapt as they interact with
other agents. A complex adaptive system is a framework for studying, learning,
explaining, and understanding the agents of that system [15]. These agents can
range from groups of intelligent cars, humans, and animals to anything that can
generate an emergent pattern and self-organization through correlated feedback.
Therefore, a SC model views the city as a complex adaptive system consisting
of services, resources, and citizens that learn through interaction and change in
spatial and temporal domains [17].
All these features of dynamic development and complexity are key issues for
city planners that require a new modeling and systematic approach. A Multi-
scale Modeling (MM) is an approach that can be used for better understanding
of complex adaptive systems [18]. The MM is a new type of modeling that
uses multiple models at different scales simultaneously to describe a complex
adaptive system [19]. These models at micro, meso, and macro levels focus on
different scales of resolution. Thus, MM in megacities and SCs paves the way
for many emerging intelligent applications that aim to achieve reduced compu-
tational complexity and cost, reliability, sustainability, and many more in SC
subsystems, including smart transportation, smart power system, smart health-
care, smart community, and smart industry. In the literature, there are several
Table 1: Acronyms
Acronyms Definitions
AI Artificial Intelligence
AVs Autonomous Vehicles
CAS Complex Adaptive System
CPS Complex Physical System
CVs Connected Vehicles
IAM Integrated Assessment Model
IAV Influenza A Virus
ICTs Information and Communication
IoT Internet of Things
LSTM Long-Short-Term-Memory
ML Machine Learning
MM Multiscale Modeling
SC Smart City
SCs Smart Cities
LULC Land Use/Land Cover
surveys/reviews on MM, either covering aspects of the MM paradigm, such as
architecture, classifications, principles, and frameworks, or focusing on different
application domains. For example, the authors of [20, 21, 22] reviewed MM for
emergent behavior, complexity, and combinatorial explosion, while the study
[23] presented a survey on MM for complex dynamic issues. The authors of
[24] focused on the applications and benefits of MM in the food drying industry
, while the work [25] presents an overview of MM in behavioral, biomedical,
and biological fields. Furthermore, the study presented in [26] reviews the ap-
plications of MM in food engineering and another paper reviews and presents
challenges related to polymer dynamics [27]. Although megacities, SCs, and
MM have been thoroughly studied in the existing literature, there is no study
that provides a comprehensive overview of the combination of the three research
domains, i.e., megacities, SCs, and MM. Thus, there is an exigent need for a
comprehensive review that provides new researchers and industry with an up-
to-date state of MM in megacities and SCs. In addition, there is a need to
explore the current challenges and opportunities in this multidisciplinary field.
However, a comprehensive survey from the perspective of the MM in megacities
and SCs has yet to be presented in the literature, to the best of our knowledge.
Motivated by the above, this article attempts to review MM in SCs. We
start with the introduction of megacities and their transformation into SCs by
discussing various challenges associated with it. We then examine the emergence
of SCs, their characteristics, and five generations 1.0 to 5.0. We also provide
a detailed overview of MM, its classification, and its emerging applications in
SCs, including urban expansion modeling, social systems modeling, healthcare
systems modeling, and traffic control modeling. Finally, current challenges and
future directions in the field of MM and SCs are presented, including smart
mobility, monitoring of urban infrastructure growth, integration of ML and
MM, uncertainty quantification, and growing and pruning multiscale models.
The remaining paper is organized as follow: Section 2 presents the agenda
and methodology of our survey. In Section 3, megacities, their challenges, and its
transformation towards SCs are unfolded. The next Section introduces MM in
SCs along with its classifications, i.e., sequential and concurrent MM. Section
5 presents some emerging applications of MM in SCs and Section 6 discloses
current challenges and future directions related to MM. At the end, Section 7
concludes this study.
2. Agenda and Methodology of Survey
This section will explore the purpose and review agenda along with the
methodology opted for the survey.
2.1. Purpose and Review Agenda
This literature review aims to collect, analyze, and present various dimen-
sions for the SC by logically classifying the existing relevant body of literature.
This study not only serves the dual purpose of highlighting current problems of
transforming megacities to SCs but also discloses future areas of investigation
via MM. In this study, the authors summarize the diverse conceptualization by
identifying possible gaps. The literature review is based on four sub-areas of in-
terest, including megacities and their challenges, towards SCs and generations,
emerging applications of MM in megacities and SCs, and current challenges and
future directions related to MM in SCs.
The first and most important area of interest is a set of components that are
relevant to the understanding of megacities and their challenges. This appears
as fundamental since it not only gives the readers an overview of megacities but
also their challenges which urged city planners to move towards SCs. In short,
the author’s objective is to highlight and identify the full spectrum of challenges
being faced by megacities planners and also assess how these challenges are
related to or distinct from each other.
The SC uses several integrated components which makes it a complex adap-
tive system. Based on this fact, the author’s second sub-area of interest probes
into the different types of literature focusing on SC and its components. The four
characteristics of SC: sustainability, urbanization, quality of life, and smartness
are explored. The evaluation of SC through five generations is also investigated
to give the readers an updated state-of-the-art data on the SC. Considering the
complexity of components in the SC, the focus of this sub-area is limited to the
SC definitions, characteristics, and generations.
The third area of interest relates to the preliminary on MM and its main
three areas including multiscale analysis, models, and algorithms are discussed.
The authors attempt to define various works based on literature where MM
is used in different categories including sequential and concurrent approaches.
The fourth area of interest attempts to discuss the emerging application of
MM being used in SCs.Thereby, this review is not only intended to give much
information on megacities, SCs, and MM but also various emerging applications
used to deal with the megacities’ problems. In last, the authors identify the
current challenges of MM since modeling at different Spatio-temporal scales has
different complexities and future directions in MM are also discussed.
2.2. Search and Literature Review Methodology
Despite the four authors’ main areas of interest discussed above the primary
objective of the research methodology is to identify, classify, and review the
existing knowledge on megacities, SCs, and MM approaches that are employed
for various problems. The following five steps methodology is being opted for
our comprehensive review:
2.2.1. Search based on keywords:
As a preliminary step, a keyword-based search using Google Scholar is per-
formed. Google Scholar was chosen because it ranks articles using different
factors, including publishers, authors, number of citations, along the published
year. Google Scholar was selected for searching for high-quality papers. The
keywords for which articles were searched include megacities, SCs, MM, mul-
tiscale systems, and MM applications. In this round, we found 300 articles to
consider for further screening.
2.2.2. Articles screening:
In this step, screening of articles were performed that were retrieved during
the first step. The criteria of screening chosen were to focus on the articles
that used multiscale analysis, MM, and multiscale algorithms applications in
megacities problems.
2.2.3. Supplementary articles:
Here supplementary papers were found based on the articles selected in step
2. To be specific, the articles that were cited in the chosen papers and papers
that are citing the selected articles were also screened via similar criteria as
described in the previous step.
2.2.4. Considered for review:
In this step, all papers selected in steps 2 and 3 were considered for review
to disclose their objectives of how MM applications are applied towards various
problems in SCs. Finally, after following selection criteria, this study considers
more than 70 research papers for review process.
2.2.5. Analyzing review results:
In the final step, MM review results were analyzed to find various appli-
cations used to cater SCs problems at various temporal-spatial scales. In this
phase, research gaps and opportunities were also explored for future endeavors
for city planners to use MM approaches for city problems.
3. Megacities and its Challenges: Towards Smart Cities
In this section, we uncover the background of cities and megacities along
with their key challenges, and then the introduction of SCs has been unfolded.
3.1. Preliminary on Cities and Megacities
The city has significant importance for growth as it carries most of the human
activities, i.e., cultural, economic, and social. In other words, it is the engine
and crystallization point of cultural and social change. Today, urbanization is
accepted as an important phenomenon of society and economy around the globe
[28, 29]. A city can be defined as:
Definition 1. A city is an urban area that performs all its functions and pro-
vides facilities to its residents without using Information and Communication
Technologies (ICTs), and other digital technologies [30].
Fig. 1 presents the trends of urban, rural, and total population around the globe.
In addition, this figure also shows the urbanization trend (red curve), which is
continuously growing. It can also be seen from the figure that in 1950, the
population in rural areas was more than twice the size of the urban population;
however, due to the increasing pace of urbanization, the urban population has
been increasing since 2005 and will be more than twice the size of the rural
population in 2050 [28, 31]. The cities are facing many problems due to lack
of efficient technologies and other factors. The key problems in cities include,
mobility, health, housing, environmental, and security issues.
The high pace of urbanization is the emergence of megacities, where more than
10 million people live in a single megacity [32]. According to [7], there are 30
megacities in the world with combined residents of more than 300 million, and
they are greater in size than many countries. Therefore, in this study, we use
the definition of megacities as presented in [33, 34, 35].
1950 55 60 65 70 75 80 85 90 95 2000 05 10 15 20 25 30 35 40 45
Population (Thousands)
Urbanization rate (%)
Urban population
Rural population
Total population
Urbanization rate
Figure 1: Evolving urban and rural populations along with the growth of urbanization around
the globe [28, 31]
Definition 2. Some megacities are dense centers of population, economic ac-
tivities, and pollutant emissions, and at the same time areas where effective
pollution control strategies could maximize benefits [33, 34, 35].
Basically, the megacities act as magnets that attract residents from rural areas
to work, seek job opportunities, and improve their standard of living [36]. They
can earn more in cities than in rural and remote areas. Although people move
to cities because of a better standard of living and many other opportunities,
there are many problems raised by rapid urbanization, such as deterioration of
the ecological environment, health problems, traffic congestion, unusually fragile
infrastructure, the increased crime rate in cities, environmental problems, fierce
competition due to low resources, living conditions of migrant workers, etc.
[8, 9]. Some of the major challenges of megacities are discussed below. Here,
it is important to note that the main challenges related to megacities could be
due to various lacks/ situations (e.g., the political situation as well as the lack
of integrated governance in the urban structure), not directly relevant to the
poor systems and facilities.
3.1.1. Mobility Issue
Megacities face severe mobility problems due to inefficient transportation
systems [37]. Especially in Asia, where the population is high relative to other
contents, congestion is common in the megacities. Due to lack of planning and
inefficient control of transportation systems, it has become a Herculean task
for ordinary citizens to drive a few miles. In [38], the authors demonstrated a
14.3-30.4% increase in carbon emissions during peak hours using an open-access
congestion index in the road network of Shenzhen, China. In a case study of
Seoul, a study focused on development around transportation hubs through the
provision of improved public facilities, pedestrian improvements, and proper
implementation of multi-level planning to achieve desired outcomes between
metropolitan and stakeholders [10]
3.1.2. Health Issue
Chaotic traffic and environmental problems cause air, noise, and other pol-
lution that can lead to mass health challenges. These challenges include respi-
ratory problems, high blood pressure, heart disease, etc. The paper presented
in [11] had given a composite index based on three community aspects: health,
safety, and environment. The result showed that the health-environment had
the highest score for the municipality of Tehran, Iran. According to Kelly
and Fussell, air pollutants in megacities have a significant negative impact on
public health and are not only difficult but also expensive to control. The four
main sources contributing to human health problems are biomass emissions from
urban households, air pollution from megacities, desert storms, and wildfires,
which need to be managed by promoting SCs [39].
3.1.3. Housing Issue
The housing projects that are being carried out in the megacities are not
able to meet the growing needs of the population demands. Moreover, the illegal
construction of shopping malls and hospitals in busy areas in most developing
countries violates urban planning rules and building codes. A study focused on
the housing problems in China’s megacities, which give a theoretical framework
for rental housing and land supply [40]. Housing problems are not limited to land
supply, but also include water issues [41], energy shortages [42], and sanitation
3.1.4. Environmental Issue
Lack of planning and uncontrolled urbanization and settlement of megacities
have led to severe environmental problems. Megacities have become the most
polluted cities with piles of garbage piling up in every nook and corner. In
developing countries, due to the lack of a sewage system, a light downpour
causes sewage to erupt from everywhere and the streets become the sight of a
flood. [28] noted in their review article that air pollution is one of the biggest
problems of megacities in China. The authors in [44], focused on environmental
crimes related to carbon and other emissions in the use of energy from renewable
and non-renewable energy sources.
3.1.5. Security Issue
The law and order situation is also a major problem in megacities. Incidents
of cell phone robbery and robbery with the use of firearms are rapidly increasing
in the outskirts of cities in developing countries. [45] applied a population-based
ant colony algorithm to identify ecological security patterns based on corridors
and restoration of 34 key points during a case study of Beijing city. In addition
to physical security, as mentioned in [46], there are other challenges, including
water security [47], food security [48], and natural disasters such as flooding [49].
It is pertinent to mention here that the problems of megacities are inherently
neglected, interconnected, and complex. Therefore, the vision of the multiscal-
ing approach can play an important role in the transformation towards SCs,
which uses a more efficient approach to culminate all these problems.
3.2. Introduction to Smart City
A SC integrates several smart components in a unified way to facilitate their
citizens in every filed of life. Citizens’ satisfaction in terms of providing smart
services is the key hallmark of smart cites. There are several components that
build SC architecture, i.e., smart healthcare, smart homes, smart security, smart
community, smart industry, smart energy, smart transportation, and smart ed-
ucation system, as presented in Fig. 2. These components are what make cities
smart and efficient.
The concept of connecting several devices through state-of-the-art network
became very praising with the advent and advancements of smart devices. The
IoT emerged from the development of traditional networks connecting millions
of smart objects. Thanks to the great attention of various stakeholders, the IoT
has spawned impressive applications as it spreads, such as smart healthcare,
smart warehouses, SCs, smart homes, and so on [52, 53, 54]. SCs have come
into focus in recent decades due to drastic urbanization around the globe. The
implementation of city operations with the help of ICTs has made cities effective
in several ways, i.e., smart transportation, smart healthcare, smart energy, and
Smart power
Smart wharehouse
Smart transportation
Smart healthcare
Smart Education
Smart home Smart community
Smart security
Figure 2: A generic composition of SC architecture
smart shopping. But the inclusion of ICT to carry out urban operations does not
yet imply the full interpretation of SCs [55]. The SCs were favored among the
other city models, i.e. digital city, information city, and telicity, as it represents
the abstraction of all the other models [51]. Since SCs is an application domain
of the IoT [56], therefore the operating mechanism of SCs is the same as that
of IoT.
The SC concept integrates ICTs and various physical devices connected to
the IoT network to optimize the efficiency of city operations and services pro-
vided to citizens. Broadly speaking, a SC is an urban area that uses ICTs and
· Infrastructure
· Social life
· Energy
· Climate change
· Pollution & waste
· Health
· Economy
· Technology
· Infrastructure
· Economy
· Governance
· Employment
· Emotional well-beings
· Financial well-beings
· Services
· Better facilities
· Smart environments
· Smart livings
· Smart mobility
· Smart governance
· Smart people
· Smart economy
of smart city
Sustainability Urbanization
Smartness Quality of life
Figure 3: Characteristics of SC [50, 51]
various types of IoT-based sensors to collect data and then use the insights
gained from that data to efficiently manage assets, resources, and services with
the goals of improving city operations performance and quality of service. This
includes data collected from citizens, devices, and assets that are processed
and analyzed to monitor and manage traffic and transportation systems, power
plants, utilities, water supply networks, waste management, crime detection,
information systems, schools, libraries, hospitals, and other municipal services.
A comprehensive definition of a SC is given by [50, 57] as:
Definition 3. A SC is an advanced modern city that utilizes ICTs and other
technologies to improve the quality of life, competitiveness, operational efficacy
of urban services while ensuring resource availability for present and future gen-
erations in terms of social, economic, and environmental aspects [50, 57].
3.2.1. Characteristics of Smart City
There are four major attributes of SC, including sustainability, smartness,
urbanization, and quality of life [51], as depicted in Fig. 3. Furthermore, it
is worth noting that some issues (such as infrastructure, economy, and gover-
nance) have robust links with more than one attributes (see Figure 3).
Sustainability: The sustainability of a SC includes (but not limited to)
urban infrastructure and governance, energy and climate change, pollution and
waste, and social issues, economy, and health.
Urbanization: The urbanization aspects of a SC include several aspects and
indicators, such as technology, infrastructure, governance, economy, etc.
Quality of life: Quality of life can be measured in terms of the emotional and
financial well-being of citizens.
Smartness: The smartness of a SC is conceptualized as the effort to improve
the social, economic, and environmental standards of the city and its inhabi-
tants. The various aspects of smartness of a city that are frequently cited include
smart environments, smart living, smart mobility, smart governance, smart peo-
ple, and smart economy [51, 58]. A smart environment represents an attractive
and clean natural state with the least pollution and sustainable management of
resources. Health conditions, the standard of living, safety, cultural and edu-
cational facilities are the key indicators contributing to smart living. The key
factors for smart mobility are safe, sustainable, and innovative transport sys-
tems, and the availability of ICT infrastructure availability. Smart governance
is concerned with a transparent government that involves its citizens in decision-
making and provides easy access to public and social services. Smart people are
characterized by creativity, skill level, open-mindedness, an affinity for learning,
and participation in public life. The elements that contribute to creating a city
with a smart economy are entrepreneurship, labor flexibility, productivity, and
international embeddedness.
3.2.2. Smart City Generations
This section discusses in detail five different stages of how cities have em-
braced technologies and developments, moving from the technology- driven (SC
1.0), to the city government-driven, citizen-driven, industry 4.0 (4G, 5G, electri-
cal vehicles, etc.), and to, finally, artificial intelligence and cognitive computing
(SC 5.0). Figure 4 also presents pictorial presentation of five phases of SC Evo-
Smart City 1.0: SC 1.0 is characterized by technology vendors driving the
adoption of their solutions in cities. These cities are often criticized for their
technology push and the influential role of large companies, such as IBM and
CISCO. The SC 2.0, unlike the technology providers in SC 1.0, is led by a gov-
ernment agency, mayors, and city councils.
Smart City 2.0: SC 2.0 is appropriate when technological tools are explicitly
developed to address problems such as pollution, sanitation, health, and trans-
portation in consultation with citizens. Unfortunately, citizen participation in
informal decision-making structures and assemblies is poor and appeals only to
a small minority [59]. While SC 1.0 and 2.0 are driven by technology and gov-
ernment decisions respectively, SC 3.0 is driven by citizen/user expectations.
Smart City 3.0: In SC 3.0, the public should be able to express their opinions,
with the government acting as a facilitator and definer of government-specific
user needs [60]. Thus, a SC represents the entire connected ecosystem that
brings together the technologies, solutions, actors, and audiences in the SC, in-
cluding IoT, 5G connectivity, transportation and smart automotive, energy and
utilities, health and public safety, artificial intelligence, and data analytics [61].
Smart City 4.0: By adopting industrial revolution 4.0, the benefits of SCs
are appreciated to overcome the cost of the city with the platformization of the
city [62]. SC 4.0 represents the best of the past, for instance, the technological
disruption of generation 1.0, the individualization of 2.0, and the engagement
of 3.0; however, it adds two critical success factors: a holistic approach and the
challenge of integrating solutions [63]. The holistic approach aims to integrate
not only new technologies with old, but also new technologies with technologies
that may not have been developed yet. Moreover, successful city leaders under-
stand the opportunities and limitations of new technologies and appreciate the
impact that SCs technologies can have on their communities, while also recog-
nizing that these positive impacts may not be felt equally by all members of the
Smart City 5.0: SC 5.0 is characterized by the cooperation between humans
and Artificial Intelligence system [64, 16], and can harmoniously balance all
aspects of life and conflicting interests of different city stakeholders. The city
5.0 provides the approach that can help to find a ”consensus” between different
services and, more importantly, with citizens. As it reflects real-life, it should
not only take into account past or current information, but constantly changing
interests, preferences, and constraints of all actors in real-time, which should
be continuously identified, analyzed, transformed into plans, implemented, and
Figure 4: Five generations of SC
4. Preliminary on Multiscale Modeling
Multiscale phenomena have become a part of our daily lives, for instance,
we have divided our time into years, months, and days based on the multiscale
dynamics of the solar system. Similarly, our hierarchical structure of society into
continents, countries, states, and cities is also based on the multiscale geographic
earth structure [20]. The term ”MM” was coined in the early 1980s and it
can be defined as: a modeling approach in which multiple models are used
simultaneously at different scales to describe a complex system. The term MM
was coined in the early 1970s and the number of articles has grown tremendously
over time, as shown in Table 1. The MM can be defined as:
Definition 4. MM is an approach that deals with several different scales in a
single framework [65].
In a complex system, objects consist of several interrelated parts. Com-
plexity is classified into two different types: Complex Physical System (CPS)
and Complex Adaptive System (CAS). CPS has fixed properties like atoms and
is expressed by differential equations. In CPS laws, the elements change with
time; the only element that changes is the positions of the objects. In CAS,
however, elements, also called agents, are not fixed and they learn and adapt
as they interact with other agents [66]. Complex systems focus on different
scales of resolution. These models are eventually combined to produce more
accurate results than if only one scale approach is taken. Both accuracy and
efficiency can be achieved when using MM in complex systems. In addition,
the topic of multiscaling covers three different areas, including multiscale anal-
ysis, multiscale models, and multiscale algorithms. A multiscale analysis is a
fundamental component of MM that allows us to understand the relationship
between different models at different scales. Multiscale models help us to build
a model that consists of different models at different scales. Finally, multiscale
algorithms are developed by employing multiscale ideas. Numerical simulations
of MM are typically divided into three different scales, including microscopic
(106), mesoscopic (104), and macroscopic (102) [23]. The decisions made at
the lower level (e.g., citizen level) are included in the micro scale in MM, the
decisions made between the lower level (citizen) and the upper level (govern-
ment) are included in the meso scale in MM, and the macro scale includes the
decisions made at the higher level, e.g., government level.
The authors in [67] stated that the use of ICT has enabled an era of advanced
services to improve living standards. The research community has identified in-
formation services that include starting new businesses in a country, making
reservations for medical examinations, a system for remote patient monitoring,
and waste disposal by the city administration. The authors of [68] analyzed
the case of Barcelona, the second-largest city in Spain, in terms of becoming a
SC. The aim of the research was to find out the necessary infrastructures/assets
requirements for the transformation of Barcelona into a SC and to identify
the challenges during the transformation. The methodology adopted was a case
study approach using interviews, observations, site analysis, and the use of other
various sources. The Barcelona model of the SC consists of four main compo-
nents, including 1) Smart People, 2) Smart Economy, 3) Smart Living, and 4)
Smart Governance. Smart Governance includes better and free access to gov-
ernment information through the Open Data project. Smart Economy consists
of providing an interactive platform for individuals, companies, and enterprises
to interact, collaborate and boost their businesses. Smart living involves smart
public transportation. Smart People involves digital literacy. The three corner-
stones of the Barcelona SC model are ubiquitous infrastructures, human capital,
and information. Infrastructures include buildings, roads, and fiber connections.
Information includes data that comes from sensors, Open Data, or social me-
dia. Human capital are agents that help make the city smarter. Components
of Barcelona SC Strategy include smart labs for living, smart infrastructures,
neighborhoods, new and fast services for citizens, data that is open and acces-
sible to all, and the management of the SC.
4.1. Categories of Multiscale Modeling
This section unfolds two main categories of MM, i.e., sequential and concur-
4.1.1. Sequential Multiscale Modeling
In the sequential MM, the microscale models are used first to precompute
and generate the data inputs that can then be further used for the macro model.
In this way, the information is communicated from lower to upper scales and
operations/activities at the upper levels wait until operations at a lower level
are completed. Based on the above fact, sequential multiscale models are also
called ”parameter passing” because some parameters are passed between the
macro and micro scale models [19]. Fig. 5 presents a typical example of sequen-
tial MM, where it can be seen that vehicle speed, weather forecasting, traffic
control, and navigation control operations are performed at the micro level;
subsequently, based on data given by micro scale, traffic flow prediction, travel
time prediction, and routs are optimized at meso scale. Eventually, based on
input from meso scale, whole traffic is managed and new policies and rules are
4.1.2. Concurrent Multiscale Modeling
In the concurrent MM, both micro and macro models operate simultaneously
and the data required by the macro model is generated on-the-fly from the micro
models. The concurrent model is further divided into two subcategories called
”partitioned-domain” and ”hierarchical” methods. The partitioned-domain con-
current approach deals with the physical problem that is partitioned into two
or more contiguous regions, where a different model scale is used in each region.
On the other hand, hierarchical methods, use both scales micro and macro ev-
erywhere [69]. Figure 6 shows a typical example of the concurrent multiscale
model that was developed for social systems [70], where three scales are con-
sidered, i.e., micro, meso, and macro. It can be seen from the figure, there are
Vehicle speed
Traffic flow prediction
Travel time prediction
management New rules
& policies
Meso Micro
Traffic signal
Route optimization
Figure 5: A typical example of sequential MM for smart transportation systems in SC envi-
three modeling scales along with decision-makers at a particular scale. Based on
Figure 6, there are three decision-makers at the micro level, i.e., industry, com-
mercial, and citizens, and the main decisions made at this level are how much
power to consume, which utility company to buy electricity from, whether elec-
tricity is safe or not, etc. The main decision-makers at the meso level are the
electricity utilities, which are responsible to provide electricity. In addition,
their decisions influence both macro-level and micro-level decision-makers. Fi-
nally, the government is on the top (macro level) to make global decisions and
their decisions also influence others.
Compared to existing methods dealing with social systems, we can say that
the approach of MM offers the possibility to break down the case into levels
of abstraction, which can be very practical when we try to model concepts like
society 5.0. In addition, it also provides more insight into each system that is at
certain levels, and these insights can allow us to review the model from differ-
ent perspectives. The study presented in [70] conducts several experiments to
compare the MM approach with state-of-the-art methods, and the results show
the effectiveness of the MM-based approach.
Economic Policy making
Power companies
Energy type
Industries Commercials Citizens
Micro Meso Macro
Figure 6: A typical example of concurrent MM developed for social systems in SCs (taken
from [70])
5. Emerging Applications of Multiscale Modeling in Megacities and
Smart Cities
In this section, we present some emerging applications of MM in SC environ-
ment, including urban expansion modeling, atmospheric dispersion modeling,
social systems modeling, diseases and viruses modeling, energy forecasting, and
traffic control modeling.
5.1. Urban Expansion Modeling in Megacities
For the growth of urban extensions in megacities, there are various factors
such as infrastructure, housing, industry, hospitals, population, etc., which need
to be monitored from the administrative point of view. A multiscaling approach
is used in megacities to model and monitor the urban expansion of Wuhan city
in Central China [71]. Three approaches, including a geographic information
system, remote sensing, and spatial analysis, were combined to monitor ur-
ban expansion from 1995 to 2010. For the purpose of exploring the driving
mechanisms underlying urban expansion in megacities, twenty variables were
categorized into different groups, i.e., proximity, density, and characteristics.
The simulation results showed the supremacy of the spatial regression models,
and with the increase in scales, a better fit is obtained. Another study also
uses MM in megacities to examine the relative importance of policy factors
and socioeconomic at different administrative levels on urban expansion and
the associated conversion of cultivated land in China [72]. They conduct the
analysis for urban hot-spot counties across the country and use multilevel mod-
eling approaches to inspect how policy factors and socioeconomic at different
administrative levels affect the conversion of cultivated land over three-time in-
tervals (1989-to-1995, 1995-to-2000, and 2000-to-2005). A study presented in
[73] also uses a multi-level approach (i.e., pixel, grid, and city block) in megac-
ities and proposes a general framework for a precise analysis of urban change.
Two typical megacities of China, i.e., Wuhan and Beijing, are selected for the
experiments. Experimental results affirm the productivity and accuracy of the
developed MM -based multi-level approach for monitoring subtle urban changes,
achieving kappa coefficients of 0.8 at the pixel-level and 93-95% accuracy at
the grid-level.
5.2. Atmospheric Dispersion Modeling
Measuring air quality by running atmospheric dispersion models is time-
consuming. Integrated assessment modeling is a new development for measuring
air pollutants, greenhouse gases for a smart environment. [74] used a multiscale
UK Integrated Assessment Model (IAM) for the UK to measure emissions of
sulphur, nitrogen (SO2, NH3, NOx, etc.) in air quality for better health and
ecosystem protection. The proposed multiscale IAM is applied in various sce-
narios, including agriculture [75], road transport [76], and energy projections
[77], and compared with other UK models. The experimental results demon-
strate the higher performance of their proposed multiscale-based model against
its counterparts. Another study [78] also uses MM for the atmospheric environ-
ment and it is proved from the results that the proposed approach better solves
the diurnal variations in temperature, humidity, and wind speed over complex
urban areas.
5.3. Social Systems Modeling
MM is able to break-down the social systems in SCs and megacities into
multiple layers or levels. This subdivision provides better insight into each sys-
tem at a particular level. These insights give decision makers the ability to
review and control the model from different perspectives. They can also solve
any problem at a particular level/layer without interrupting work at other lev-
The Super Smart Society is a novel concept that has attracted a lot of attention
in Japan. It consists of almost 12 interconnected systems and all systems con-
tain multiple subsystems simultaneously [70]. Furthermore, there are citizens or
society at the end of each system. So, it is quite a difficult task to model these
kinds of systems. The authors of [70] have developed a multiscale approach for
modeling social systems where three steps are proposed, i.e., 1) scale separation,
2) identification of decision-makers, and scale bridging. The first step further
contains two phases, the selection of scale separation characteristic and the set-
ting of scale granularity. The scale separation characteristic is selected as the
amount of energy. Moreover, the developed model consists of three scales, i.e.,
macro, meso, and micro.
In the second step (identification of decision-makers), there are three decision-
makers in a micro scale, such as citizens, commercials, and industries. In this
stage, various decisions are made such as the amount of energy to be consumed,
which utility company will supply energy, whether the electricity is safe or not,
etc. At the meso scale, decisions are made regarding energy companies and
decisions at this scale affect both scales, i.e., upper and lower (macro and mi-
cro). The macro scale is responsible for the decisions made by the government.
Finally, the last step deals with the scale bridging, which shows the type of
MM (information flow), i.e., sequential and concurrent models (see figures 5
and 6). This study adopts both models for information flow and decisions. Sev-
eral experiments were also conducted to show the effectiveness of the proposed
model, which is confirmed by the results. It is worth noting that the proposed
method becomes slow due to the large amount of data generation when real-
time participants in a social system are considered. This is because every node
(participant) generates its own data and sometimes every single node gener-
ates data every minute. So, there is a need for Big Data analysis approaches
to deal with big data during MM in social system modeling in SCs or megacities.
5.4. Diseases and Viruses Modeling
The multiscaling approach is widely applied in various fields of biology, i.e.,
public health management [79, 80], cell to organ modeling [81], MM in medicine
[82]. One of the powerful tools is microrheology, which is used to find out the
mechanical properties of living cells. Microrheology deals with the study of cells
at different length scales and over different time spans. Biological functions can
span a variety of elements, including cell migration, cellular adhesion, and di-
vision. Viruses and bacterial cells use the actin polymerization machinery for
their growth and propulsion. Cancer is caused by cell migration and eventually
leads to metastasis, which results in the death of living organisms. Therefore,
cell motility and migration are important areas of biology that cause tumor
growth in which MM is applied [83].
Cancer Cell Growth Modeling: Cancer is a disease in which a cell deviates
from its normal pathway into adjacent tissue and forms a tumor. Cancer is a
complex biological phenomenon that requires a multiscale approach to monitor
all activities in which cancer cells communicate with their microenvironment to
grow and survive. The authors in [84] studied cancer cells using a multiscale
approach including (micro, meso, and macroscopic scales) in continuation of
using various biological modeling techniques including Boolean networks, dif-
ferential equations, stochastic methods, agent-based systems, etc. Since cancer
follows and uses a common progression schedule, it is possible to use an appro-
priate modeling approach for better clinical outcomes and better understanding
of cancer treatment approaches.
MM approaches that incorporate biological and physical modeling are com-
monly used to study the tumor microenvironment. The work [85] focused on a
multiscale approach to melanoma metastasis in which tumor cells interact with
polymorphonuclear neutrophils. Finally, the latest technological biochemistry
and structural models related to the tumor environment at the micro level and
cell growth and population at the macro level have been summarized.
Influenza A Virus (IAV) infection has increased rapidly in recent years, and most
studies have addressed on one scale modeling either tissue-level or population-
scale infection. A study presented in [86] reviewed IAV infection and emphasized
the use of a MM approach that includes spatial intrahost models and links viral
load to transmission.
5.5. Energy Forecasting
To cope with the problems of carbon emission and costly power generation
in SCs, the world is moving towards renewable energy sources, solar panel, wind
turbine, etc [87]. In particular, electricity generation from solar panels is consid-
ered more attractive because there is no cost to operate solar panels other than
the initial installation cost [88, 89]. However, since solar energy is significantly
affected by weather conditions, accurate and efficient solar energy prediction
methods are crucial for managing energy supply with power demand in a SC.
There are many research works that provide various deep learning and ML mod-
els for predicting solar or wind energy for proper energy supply management
[90, 91, 92, 93]. However, the structure of most predictive models is not based
on MM. Since it is observed that irregular factors have a negative impact on the
prediction results of a very short-term energy prediction, the overall prediction
performance is degraded. To solve this problem, multiscale forecasting models
are needed. For example, the authors of [89] have developed a deep learning
model based on Long-Short-Term-Memory (LSTM) with multiple scales, which
is able to perform very short-term energy forecasting for efficient power supply
management. The developed multiscale LSTM method concatenates two differ-
ent scaled LSTM modules to overcome the degradation caused by the irregular
5.6. Traffic Control Modeling
Modern transportation technology in SCs is witnessing the emergence of a
new era of transportation systems commonly known as cooperative systems.
Recent advances in Autonomous Vehicles (AVs), Connected Vehicles (CVs),
and autonomous ships are expected to completely change the way people use
and perceive modern transportation [94]. Therefore, in order to improve the
efficiency, throughput, and safety of traffic flow, researchers are becoming in-
creasingly interested in exploring vehicle-to-vehicle and vehicle-to-infrastructure
communication technologies. In the last decade, the research community has
been exploiting the AVs and CVs for efficient control, proper management, im-
proved safety, and efficient throughput in the SC traffic flows [95, 96]. The main
motive for using MM in SC traffic control is to properly understand and analyze
CVs and AVs in a traffic flow. For instance, in [96], a multiscale control archi-
tecture is developed for vehicle-to-everything traffic control and analysis in a
SC environment. Whereby in the newly developed framework, CVs are treated
as discrete objects that can be controlled microscopically, and in this way, they
can macroscopically influence and control the overall traffic flows.
5.7. Multiscaling Approach in Behavioral Sciences
In behavioral science, animal and human behavior is studied through a sys-
tematic approach. This behavior is studied through controlled and disciplined
scientific experiments. The following are some areas of behavioral science in
which a multiscaling approach is used.
Crowd’s Modeling: Modeling and analyzing crowd dynamics at the micro
and macro levels is one of the motivating research areas that are studied sepa-
rately. A multiscale approach based on microscopic and macroscopic levels has
never been studied in crowd dynamics, leading to new consequences of prop-
erties such as self-organization and emergent behavior. In [97], the authors
modeled and coupled micro and macro scales of crowd dynamics together in a
rigorous mathematical framework to interact and obtain more accurate results.
The study and analysis of crowds is an important concern in today’s world. It
is a difficult task to use multiscale texture analysis for inferring unsupervised
crowds without having contextual knowledge. A novel algorithm is presented
to find out crowds and backgrounds in still images without any prior knowledge
[98]. Dense crowds are successfully inferred using the following three steps: i)
extracting the crowd features from the image in pixels and storing them in a
vector ii) using the binary classification approach to distinguish the crowd from
its background iii) optimizing the time and data volume for computation.
A smart and intelligent environment requires visual crowd detection for various
applications such as security, surveillance, etc. The authors in [99] studied 83
papers and showed the progress and trends in crowd analysis and modeling for
the seven years from 2007 to 2014. The conclusion drawn shows inadequacies
and challenges in datasets, assumptions, modeling, and methods for crowds,
including density dependence, and the focus was directed to the study of crowd
behavior at both macro and micro-levels understandings of crowds.
Surveillance through Micro Aerial Vehicles: Micro aerial vehicles, also
called micro air vehicles, are used to monitor an object for surveillance. An
algorithm is developed that can effectively observe multiple scenes [100]. The
methodology used the centralized quad-tree strategy of micro aerial vehicles for
monitoring two or more moving targets. Mobility options were also considered,
which achieved a common goal of maximizing the observation of interactivity of
multiple moving targets.
6. Current Challenges and Future Directions in Multiscale Modeling
Up to this point, this paper has focused on understanding and reviewing the
current literature on MM in SCs, including recent advances in developments
and the latest trends. Although this area has received a lot of attention from
the research community in the last decade, there are still many topics/gaps in
this field that need to be explored as future research. In this section, this study
presents a list of current challenges along with future directions for using MM
to improve the performance of SCs. The summary of current challenges and
future directions is presented in Table 2.
6.1. People and Communities
One of the main challenges of MM is to model different systems at different
scales having different complexities. It is quite a difficult task to deal with
different systems simultaneously when each system has a particular complexity.
However, modeling and analyzing crowd dynamics at the micro and macro scales
is one of the motivating research areas that are being studied separately. A
multiscale approach based on microscopic and macroscopic levels has never been
investigated in crowd dynamics, leading to new consequences of properties such
as self-organization and emergent behavior. [97] modeled and coupled micro
and macro scales of crowd dynamics to obtain accurate results.
6.2. Smart Mobility
Situational awareness is an important aspect of intelligent mobility that
spans multiple scales, including time and space. In [101], the authors addressed
the challenge of multiscale spatio-temporal tracking using active cameras, real-
time video surveillance, and analysis, pattern analysis along with multiple ob-
ject models to create non-intrusive, and comprehensive situational awareness.
A multiscale approach was adopted for traffic flow. Traffic could be monitored
and controlled using discrete-event simulation tools. The authors in [102] used
the DEUS software tool to simulate SCs. DEUS provides discrete event simu-
lation to monitor key city business processes. The authors used online Google
Map Application Programming Interfaces to monitor and visualize peer-to-peer
traffic information systems. Monitoring a view and scene at different spatial
scales from micro aerial vehicles is complex. All the above examples have used
the MM approach to focus on a specific problem domain of SCs. In short, it
can be concluded that applying the MM approach towards smart mobility could
help the planners to better understand the problems at micro and macro levels.
6.3. Monitoring of Urban Infrastructure Growth
Urbanization is a complex phenomenon that requires spatio-temporal char-
acteristics and changes in Land Use/Land Cover (LULC) are due to available
choices such as individual preferences, policies, sustainability of sites. Un-
planned growth with respect to urban areas is a very serious problem and the
authors have studied the characteristics of peri-urban and urban growth in the
paper [103]. By applying predictive models, the authors determined the impact
at different stages, in the city of Lahore in Pakistan. The results were compiled
after examining LULC maps from 1999 to 2011 at different scales and it was
accumulated that there is a major land transition in the metropolitan region
from peri-urban and urban zones.
The monitoring of infrastructure growth, such as ports, buildings, roads,
airports, and markets, encompasses a complex environment that has fragmen-
tation and functional gaps between different institutions/government agencies.
All these infrastructures are managed by multiple scale-dependent institutions.
The challenge of interplay between institutions for better governance has been
presented in [104] to improve monitoring and cost efficiency while reducing frag-
6.4. Smart Energy Management
Energy management can be done at a micro level, taking into account the
cycles of individual appliances. The peak consumption cycles of the appliances
are shifted to off-peak hours to reduce the electricity bill of the consumers [105,
106]. At an aggregate level, different appliances are considered simultaneously
and their peak load is shifted to off-peak hours [107, 108]. At a higher level,
energy is shifted from surplus areas in a SC to deficit areas according to certain
criteria and tariffs agreed upon among users. Thus, there is an urgent need
to use MM in smart energy management systems. By using MM approaches
in SCs and megacities, various nodes of the energy system can be managed at
multiple levels. For instance, energy generation can be managed at the first
level, energy consumption can be analysed at another level, and load-demand
balance modelling, theft detection, and other aspects of smart energy can also
be managed at the third level.
6.5. Integration of Machine Learning with Multiscale Modeling
Due to the varying complexity of systems at different scales, the parameters
of the developed models are insufficient to provide a basic set for generating
the dynamics of systems at higher scales. According to [25], MM and ML
models can be set up to work in parallel to provide independent confirmation of
parameters’ sensitivity. For instance, smart home or smart health care systems
in a SC environment generate relatively simple dynamics; however, they depend
on several complex underlying parameters that can be handled by MM. There is
an opportunity to integrate ML with MM to identify both the underlying high
dimensionality and low dimensionality dynamics [109].
6.6. Uncertainty Quantification
Quantifying uncertainty helps in making several important decisions in the
SC environment. Usually, most of the houses in a SC are equipped with green
energy such as solar and wind turbines. Prediction of energy is also necessary
for efficient energy planning, and prediction results are used at the micro-scale
in a multiscale energy management system. Furthermore, prediction results
without uncertainty quantification are unreliable and untrustworthy [110]. To
understand the multiscale systems (related to energy management or any other
prediction domain, i.e., weather forecast, traffic forecast, accident forecast, and
more in the SC environment), it is necessary to first understand the quantifi-
cation of uncertainty. For example, machine/deep learning models start with
collecting appropriate datasets, selecting an appropriate forecasting model based
on the performance goals, training the model by using a labeled dataset, and
optimizing various learning parameters that help in achieving satisfactory per-
formance. There are several uncertainties involved in the ML steps that need
to be quantified. These include, for example, the selection/collection of training
data, the accuracy and completeness of the training data, the understanding
of the deep learning models along with their performance limitations and con-
straints, and uncertainties due to operational data [110, 111]. The primary goal
of uncertainty quantification is to reveal reliable confidence values for forecasting
results generated by ML methods and what the developed methods have not
learned properly. Because several decisions on the next step (meso or macro
level) are made based on these results.
In the field of energy management and forecasting area, uncertainty quan-
tification has attracted noticeable attention of the research community in recent
years. Current studies show its applications and advantages, for example, the
application of energy management in smart grid [112] and uncertainty quantifi-
cation in wind power forecasting [113, 114]. However, the uncertainty quantifi-
cation of multiscale systems in the SC environment remains open for future work
to improve the performance, reliability, and accuracy of multiscale systems.
6.7. Growing and Pruning Multiscale Models
Growing and pruning are novel techniques that can be adapted to improve
the performance and reduce the computational complexity of multiscale models.
Since multiscale systems are more complex compared to single scale models, it is
important to use such approaches that can help in reducing the complexity. In
this technique (Growing and Pruning), an architecture of the multiscale system
is first designed with the fewest necessary parameters. Then, new parameters
are incorporated into the architecture by applying the growing approach. In
contrast, when the pruning approach is applied, a number of parameters are
removed from the architecture. Both the growing and pruning approach based
architectures repeat three key operations until acceptable performance (low
complexity) is achieved [115]: i) training the model, ii) changing the weights
(parameters) based on the growing or pruning criteria, and iii) re-training the
model. In recent years, the field of growing and pruning has received consider-
able attention from the research community and several studies have discussed
its effectiveness in various research domains, including self-care activities [115],
health service improvement [116], and speech emotion recognition [117]. There-
fore, the implementation of growing and pruning approaches for multiscale mod-
els in a SC environment is still an open direction for researchers and industry.
7. Conclusion
In this study, we conduct a comprehensive survey on the application of MM
in megacities and SCs, which is becoming increasingly important due to its
emerging applications in the SC environment. The reason for this study is the
need and lack of a comprehensive survey on the integration of MM in megaci-
ties and SCs. Due to the growing population and urbanization trend, we first
Table 2: Summary of Current Challenges and Future Directions
Main Area Sub Area Challenges MM Features Future Directions References
People and
communities Crowed dynamics Crowed mo delling
at various scales
Covered micro and macro
scales of crowed dynamics
Various scales at micro and
macro levels can be opted
for more accurate results of
crowd dynamics monitoring
Smart mobility
Situational awareness
of multiple moving
targets via micro aerial
Algorithm to work
and monitor differ-
ent moving targets
MM algorithm for used the
centralized quad-tree strat-
egy of micro aerial vehicles
for monitoring two or more
moving targets
Real-time monitoring of sev-
eral targets using MM algo-
rithms usage
Situational awareness
of traffic flow
Real-time video
surveillance of
traffic flow
Spatio-temporal tracking us-
ing active cameras
Discreate and real-time
monitoring at multiple
scales, including time and
Situational awareness
of key city business
Traffic flow moni-
Monitoring a view and scene
at different spatial scales
from micro aerial vehicles
Real time traffic monitoring
through micro aerial vehicles
Monitoring of
Monitoring of LULC Due to spatio-
temporal char-
acteristics mon-
itoring of urban
growth is complex
Examining LULC maps at
different scales
LULC monitoring via
spation-temporal char-
acteristics via predictive
Monitoring of buildings Fragmentation and
functional gaps
analysis between
different institu-
agencies of differ-
ent buildings
Using MM for better gover-
Monitoring and management
of scale-dependent institu-
tions for better infrastruc-
ture growth is another area
to explored
Smart energy
Energy Management at
micro level
Monitoring of dif-
ferent cycles used
by an appliance
Using MM micro models for
better monitoring of appli-
ance cycles
Shifting peak appliance cy-
cles to off-peak electricity
pricing signals
[105, 106]
Energy textblueM-
anagement at macro
Monitoring and
shifting of aggre-
gate appliances
Using MM macro models for
better monitoring of aggre-
Shifting of aggregate appli-
ances peak load to off-peak
[107, 108]
Integration of
machine learning
with MM
Overview of MM in
behavioral, biomedical,
and biological fields
Integration of ML
with MM
MM and ML models can be
set up to work in parallel to
provide independent confir-
mation of parameters’ sensi-
To integrate ML with MM
to identify both the underly-
ing high dimensionality and
low dimensionality dynamics
is another area for future
[25, 109]
Uncertainty quantifica-
tion in wind power fore-
Involvement of
several uncertain-
ties, forecasting,
and prediction of
Use of MM features for un-
certainty quantification at
various levels, including col-
lection, training, deep learn-
ing modelling, and opera-
tional of data phases
Uncertainty quantification
of multiscale systems in the
SC environment remains
open for future work to
improve the performance,
reliability, and accuracy of
multiscale systems
[113, 114]
Growing and
techniques in
Growing and pruning
techniques at various
Reduction of the
complexity of
multiscale models
Design MM architecture
with few parameters and
growing its size at later
stages (growing) while in
pruning approach a number
of parameters are removed
from the architecture
The implementation of grow-
ing and pruning approaches
for multiscale models in a SC
environment is still an open
direction for researchers and
[115, 116,
provide an overview of cities along with megacities and their transformation
into SCs. We then discuss the various challenges associated with the transfor-
mation of megacities into SCs, including mobility, health, housing, and secu-
rity. Moreover, this study introduces new emerging application areas of MM to
help planners and government authorities for a better understanding of complex
problems associated with the transformation of SCs. In addition to assessing
current trends related to MM in megacities and SCs, we also carefully review
challenges related to MM in megacities and SCs. These include smart com-
munity, smart mobility, monitoring urban infrastructure growth, smart energy
management, uncertainty quantification, and growing and pruning multiscale
models. Reflecting the needs raised by these challenges, we further outline fu-
ture research needs and directions for the research community and industry.
Finally, based on the literature review, this study concludes that the existing
MM solutions in megacities and SCs are still at an early stage of development;
however, they can improve life in megacities and SCs. In the future, real-time
MM-based solutions, to improve the quality of life in megacities and SCs, which
can provide quick insights will be needed.
[1] S. Brenner, Brenner n., schmid c, The urban age in ques-
tion//International Journal of Urban and Regional Research 38 (2014)
[2] B. Cohen, Urbanization in developing countries: Current trends, future
projections, and key challenges for sustainability, Technology in society
28 (1-2) (2006) 63–80.
[3] A. E. Osorio, B. ¨
Ozkazan¸c-Pan, Defining the ‘urban’in urban en-
trepreneurship: Implications for economic development policy, in:
Academy of Management Proceedings, Vol. 2014, Academy of Manage-
ment Briarcliff Manor, NY 10510, 2014, p. 12486.
[4] L. Zhang, R. LeGates, M. Zhao, Understanding China’s urbanization: The
great demographic, spatial, economic, and social transformation, Edward
Elgar Publishing, 2016.
[5] Nature, The century of the city, Nature 467 (2010) 900–901.
[6] B. Halder, J. Bandyopadhyay, P. Banik, Monitoring the effect of urban
development on urban heat island based on remote sensing and geo-spatial
approach in kolkata and adjacent areas, india, Sustainable Cities and So-
ciety (2021) 103186.
[7] J. N. Pelton, Smart cities, megacities and meta-cities, in: Preparing for
the Next Cyber Revolution, Springer, 2019, pp. 67–77.
[8] E. Arcaute, E. Hatna, P. Ferguson, H. Youn, A. Johansson, M. Batty,
Constructing cities, deconstructing scaling laws, Journal of the royal so-
ciety interface 12 (102) (2015) 20140745.
[9] R. J. Fuchs, E. Brennan, F.-C. Lo, J. I. Uitto, J. Chamie, Mega-city
Growth and the Future, United Nations University Press, 1994.
[10] J. Lee, J. Arts, F. Vanclay, Stakeholder views about land use and transport
integration in a rapidly-growing megacity: Social outcomes and integrated
planning issues in seoul, Sustainable Cities and Society (2021) 102759.
[11] M. Mapar, M. J. Jafari, N. Mansouri, R. Arjmandi, R. Azizinezhad, T. B.
Ramos, A composite index for sustainability assessment of health, safety
and environmental performance in municipalities of megacities, Sustain-
able Cities and Society 60 (2020) 102164.
[12] T. Nam, T. A. Pardo, Conceptualizing smart city with dimensions of
technology, people, and institutions, in: Proceedings of the 12th annual
international digital government research conference: digital government
innovation in challenging times, 2011, pp. 282–291.
[13] T. V. Kumar, B. Dahiya, Smart economy in smart cities, in: Smart econ-
omy in smart cities, Springer, 2017, pp. 3–76.
[14] S. Aslam, H. Herodotou, S. M. Mohsin, N. Javaid, N. Ashraf, S. Aslam,
A survey on deep learning methods for power load and renewable energy
forecasting in smart microgrids, Renewable and Sustainable Energy Re-
views 144 (2021) 110992.
[15] G. Rzevski, S. Kozhevnikov, M. Svitek, Smart city as an urban ecosystem,
in: 2020 Smart City Symposium Prague (SCSP), IEEE, 2020, pp. 1–7.
[16] M. Sv´ıtek, P. Skobelev, S. Kozhevnikov, Smart city 5.0 as an urban ecosys-
tem of smart services, in: International Workshop on Service Orientation
in Holonic and Multi-Agent Manufacturing, Springer, 2019, pp. 426–438.
[17] S. Kozhevnikov, M. Svitek, P. Skobelev, Multi-agent approach for smart
resilient city, in: Service Oriented, Holonic and Multi-Agent Manufactur-
ing Systems for Industry of the Future: Proceedings of SOHOMA 2020,
Springer International Publishing, 2021, pp. 215–227.
[18] W. Garira, The research and development process for multiscale models
of infectious disease systems, PLoS computational biology 16 (4) (2020)
[19] E. Weinan, J. Lu, Multiscale modeling, Scholarpedia 6 (10) (2011) 11527.
[20] E. Weinan, Principles of multiscale modeling, Cambridge University Press,
[21] D. G. Vlachos, Multiscale modeling for emergent behavior, complexity,
and combinatorial explosion, AIChE journal 58 (5) (2012) 1314–1325.
[22] M. F. Horstemeyer, Multiscale modeling: a review, Practical aspects of
computational chemistry (2009) 87–135.
[23] M. Jebahi, F. Dau, J.-L. Charles, I. Iordanoff, Multiscale modeling of com-
plex dynamic problems: an overview and recent developments, Archives
of Computational Methods in Engineering 23 (1) (2016) 101–138.
[24] Z. Welsh, M. J. Simpson, M. I. H. Khan, M. Karim, Multiscale modeling
for food drying: State of the art, Comprehensive reviews in food science
and food safety 17 (5) (2018) 1293–1308.
[25] M. Alber, A. B. Tepole, W. R. Cannon, S. De, S. Dura-Bernal,
K. Garikipati, G. Karniadakis, W. W. Lytton, P. Perdikaris, L. Pet-
zold, et al., Integrating machine learning and multiscale model-
ing—perspectives, challenges, and opportunities in the biological, biomed-
ical, and behavioral sciences, NPJ digital medicine 2 (1) (2019) 1–11.
[26] Q. T. Ho, J. Carmeliet, A. K. Datta, T. Defraeye, M. A. Delele, E. Her-
remans, L. Opara, H. Ramon, E. Tijskens, R. van der Sman, et al., Mul-
tiscale modeling in food engineering, Journal of food Engineering 114 (3)
(2013) 279–291.
[27] Y. Li, B. C. Abberton, M. Kr¨oger, W. K. Liu, Challenges in multiscale
modeling of polymer dynamics, Polymers 5 (2) (2013) 751–832.
[28] D. Li, J. Ma, T. Cheng, J. van Genderen, Z. Shao, Challenges and oppor-
tunities for the development of megacities, International Journal of Digital
[29] P. Puchol-Salort, J. O’Keeffe, M. van Reeuwijk, A. Mijic, An urban plan-
ning sustainability framework: Systems approach to blue green urban
design, Sustainable Cities and Society 66 (2021) 102677.
[30] K. Su, J. Li, H. Fu, Smart city and the applications, in: 2011 international
conference on electronics, communications and control (ICECC), IEEE,
2011, pp. 1028–1031.
[31] UN Population, (January 2020).
[32] D. D. Parrish, H. B. Singh, L. Molina, S. Madronich, Air quality progress
in north american megacities: A review, Atmospheric Environment 45 (39)
(2011) 7015–7025.
[33] C. K. Chan, X. Yao, Air pollution in mega cities in china, Atmospheric
environment 42 (1) (2008) 1–42.
[34] B. R. Gurjar, J. Lelieveld, New directions: megacities and global change,
Atmospheric Environment 39 (2) (2005) 391–393.
[35] M. J. Molina, L. T. Molina, Megacities and atmospheric pollution, Journal
of the Air & Waste Management Association 54 (6) (2004) 644–680.
[36] M. Eremia, L. Toma, M. Sanduleac, The smart city concept in the 21st
century, Procedia Engineering 181 (2017) 12–19.
[37] P. Zhao, H. Hu, Geographical patterns of traffic congestion in growing
megacities: Big data analytics from beijing, Cities 92 (2019) 164–174.
[38] Y. Wen, S. Zhang, J. Zhang, S. Bao, X. Wu, D. Yang, Y. Wu, Map-
ping dynamic road emissions for a megacity by using open-access traffic
congestion index data, Applied Energy 260 (2020) 114357.
[39] F. J. Kelly, J. C. Fussell, Global nature of airborne particle toxicity and
health effects: a focus on megacities, wildfires, dust storms and residential
biomass burning, Toxicology research 9 (4) (2020) 331–345.
[40] L. Tian, Y. Yan, Y. Wu, L. Shao, Breaking the land monopoly: Can col-
lective land reform alleviate the housing shortage in china’s mega-cities?,
Cities 106 (2020) 102878.
[41] A. L. Britto, A. Maiello, S. Quintslr, Water supply system in the rio de
janeiro metropolitan region: open issues, contradictions, and challenges
for water access in an emerging megacity, Journal of Hydrology 573 (2019)
[42] A. Khan, N. Javaid, Jaya learning-based optimization for optimal sizing of
stand-alone photovoltaic, wind turbine, and battery systems, Engineering
6 (7) (2020) 812–826.
[43] R. Biswas, K. Arya, S. Deshpande, More toilet infrastructures do not nul-
lify open defecation: a perspective from squatter settlements in megacity
mumbai, Applied Water Science 10 (4) (2020) 1–9.
[44] A. Khan, N. Javaid, Tacma: total annual cost minimization algorithm for
optimal sizing of hybrid energy systems, Journal of Ambient Intelligence
and Humanized Computing 11 (11) (2020) 5785–5805.
[45] J. Peng, S. Zhao, J. Dong, Y. Liu, J. Meersmans, H. Li, J. Wu, Applying
ant colony algorithm to identify ecological security patterns in megacities,
Environmental Modelling & Software 117 (2019) 214–222.
[46] S. Hoor-Ul-Ain, Public sexual harassment mayhem on public transport
in megacities-karachi and london: A comparative review, Aggression and
violent behavior (2020) 101420.
[47] Y.-J. Chang, D. Zhu, Water security of the megacities in the yangtze
river basin: Comparative assessment and policy implications, Journal of
Cleaner Production 290 (2021) 125812.
[48] V. Kharisma, N. Abe, Food insecurity and associated socioeconomic fac-
tors: Application of rasch and binary logistic models with household
survey data in three megacities in indonesia, Social Indicators Research
148 (2) (2020) 655–679.
[49] M. Esteban, L. Jamero, C. Chadwick, Actual adaptation to sea level rise:
Learning from subsidence in small islands and coastal megacities, Coastal
Engineering Proceedings (36v) (2020) 8–8.
[50] B. N. Silva, M. Khan, K. Han, Towards sustainable smart cities: A review
of trends, architectures, components, and open challenges in smart cities,
Sustainable Cities and Society 38 (2018) 697–713.
[51] S. P. Mohanty, U. Choppali, E. Kougianos, Everything you wanted to
know about smart cities: The internet of things is the backbone, IEEE
Consumer Electronics Magazine 5 (3) (2016) 60–70.
[52] S. R. Islam, D. Kwak, M. H. Kabir, M. Hossain, K.-S. Kwak, The internet
of things for health care: a comprehensive survey, IEEE access 3 (2015)
[53] S. Jabbar, M. Khan, B. N. Silva, K. Han, A rest-based industrial web of
things’ framework for smart warehousing, The Journal of Supercomputing
74 (9) (2018) 4419–4433.
[54] J. Jin, J. Gubbi, S. Marusic, M. Palaniswami, An information framework
for creating a smart city through internet of things, IEEE Internet of
Things journal 1 (2) (2014) 112–121.
[55] R. G. Hollands, Will the real smart city please stand up? intelligent,
progressive or entrepreneurial?, City 12 (3) (2008) 303–320.
[56] B. N. Silva, M. Khan, C. Jung, J. Seo, Y. Yoon, J. Kim, K. Han, Plan-
ning of smart cities: Performance improvement using big data analytics
approach, in: Fourth international conference on advances in computing,
electronics and communication institute of research engineers and doctors,
2016, pp. 51–55.
[57] S. Kondepudi, V. Ramanarayanan, A. Jain, G. Singh, N. Nitin Agarwal,
R. Kumar, R. Singh, P. Bergmark, T. Hashitani, P. Gemma, et al., Smart
sustainable cities analysis of definitions, The ITU-T focus group for smart
sustainable cities.
[58] T. V. Kumar, E-governance for smart cities, in: E-governance for smart
cities, Springer, 2015, pp. 1–43.
[59] Smart City 1.0 to 3.0,
(January 2021).
[60] Smart City 3.0,
(January 2020).
[61] 5G and IoT in Smart Cities,
(January 2021).
[62] Y. Yun, M. Lee, Smart city 4.0 from the perspective of open innovation,
Journal of Open Innovation: Technology, Market, and Complexity 5 (4)
(2019) 92.
[63] Smart Cities 4.0—The undiscovered generation - Leading Cities,
(January 2021).
[64] M. Rosemann, J. Becker, F. Chasin, City 5.0, Business & Information
Systems Engineering (2020) 1–7.
[65] G. S. Rosales, N. T. Darias, Introduction to multiscale modeling, in:
O. Wolkenhauer (Ed.), Systems Medicine, Academic Press, Oxford, 2021,
pp. 68 – 76. doi:
[66] J. H. Holland, Complexity: A very short introduction, OUP Oxford, 2014.
[67] G. Piro, I. Cianci, L. A. Grieco, G. Boggia, P. Camarda, Information
centric services in smart cities, Journal of Systems and Software 88 (2014)
[68] T. Bakıcı, E. Almirall, J. Wareham, A smart city initiative: the case of
barcelona, Journal of the knowledge economy 4 (2) (2013) 135–148.
[69] E. B. Tadmor, R. E. Miller, Modeling materials: continuum, atomistic
and multiscale techniques, Cambridge University Press, 2011.
[70] N. Nikhanbayev, N. Fujii, T. Kaihara, D. Kokuryo, A study on multiscale
modeling and simulation approach for social systems, in: Proceedings of
the International Symposium on Flexible Automation 2018 International
Symposium on Flexible Automation, The Institute of Systems, Control
and Information Engineers, 2018, pp. 473–476.
[71] C. Zeng, M. Zhang, J. Cui, S. He, Monitoring and modeling urban ex-
pansion—a spatially explicit and multi-scale perspective, Cities 43 (2015)
[72] L. Jiang, X. Deng, K. C. Seto, Multi-level modeling of urban expansion
and cultivated land conversion for urban hotspot counties in china, Land-
scape and Urban Planning 108 (2-4) (2012) 131–139.
[73] X. Huang, D. Wen, J. Li, R. Qin, Multi-level monitoring of subtle ur-
ban changes for the megacities of china using high-resolution multi-view
satellite imagery, Remote sensing of environment 196 (2017) 56–75.
[74] T. Oxley, A. J. Dore, H. ApSimon, J. Hall, M. Kryza, Modelling future
impacts of air pollution using the multi-scale uk integrated assessment
model (ukiam), Environment international 61 (2013) 17–35.
[75] N. Hasnain, A multi-criteria approach to environmental policy: a case
study in agricultural ammonia emissions, Ph.D. thesis, Imperial College
London (2010).
[76] T. Oxley, A. Elshkaki, L. Kwiatkowski, A. Castillo, T. Scarbrough, H. Ap-
Simon, Pollution abatement from road transport: cross-sectoral implica-
tions, climate co-benefits and behavioural change, Environmental Science
& Policy 19 (2012) 16–32.
[77] H. ApSimon, O. Tim, Technical report on UK supplementary assessment
under the Air Quality Directive (2008/50/EC), the Air Quality Framework
Directive (96/62/EC) and Fourth Daughter Directive (2004/107/EC) for
2019, (January 2020).
[78] C. Yan, S. Miao, Y. Liu, G. Cui, Multiscale modeling of the atmospheric
environment over a forest canopy, Science China Earth Sciences 63 (6)
(2020) 875–890.
[79] R. Sadsad, G. McDonnell, Multiscale modelling for public health manage-
ment: a practical guide, Discrete-event simulation and system dynamics
for management decision making (2014) 280–294.
[80] Y. Bar-Yam, Improving the effectiveness of health care and public health:
a multiscale complex systems analysis, American Journal of Public Health
96 (3) (2006) 459–466.
[81] J. B. Bassingthwaighte, H. J. Chizeck, L. E. Atlas, Strategies and tactics
in multiscale modeling of cell-to-organ systems, Proceedings of the IEEE
94 (4) (2006) 819–831.
[82] M. Meier-Schellersheim, I. D. Fraser, F. Klauschen, Multiscale model-
ing for biologists, Wiley Interdisciplinary Reviews: Systems Biology and
Medicine 1 (1) (2009) 4–14.
[83] A. Chauvi`ere, L. Preziosi, C. Verdier, Cell mechanics: from single scale-
based models to multiscale modeling, CRC Press, 2010.
[84] A. Masoudi-Nejad, G. Bidkhori, S. H. Ashtiani, A. Najafi, J. H. Bo-
zorgmehr, E. Wang, Cancer systems biology and modeling: microscopic
scale and multiscale approaches, in: Seminars in cancer biology, Vol. 30,
Elsevier, 2015, pp. 60–69.
[85] R. F. Kunz, B. J. Gaskin, Q. Li, S. Davanloo-Tajbakhsh, C. Dong, Multi-
scale biological and physical modelling of the tumour micro-environment,
Drug Discovery Today: Disease Models 16 (2015) 7–15.
[86] L. N. Murillo, M. S. Murillo, A. S. Perelson, Towards multiscale modeling
of influenza infection, Journal of theoretical biology 332 (2013) 267–290.
[87] A. Khan, N. Javaid, Optimal sizing of a stand-alone photovoltaic, wind
turbine and fuel cell systems, Computers & Electrical Engineering 85
(2020) 106682.
[88] W. Liu, Y. Gong, G. Geng, Q. Jiang, Refined ramp event characterisation
for wind power ramp control using energy storage system, IET Renewable
Power Generation 13 (10) (2019) 1731–1740.
[89] D. Kim, D. Kwon, L. Park, J. Kim, S. Cho, Multiscale lstm-based deep
learning for very-short-term photovoltaic power generation forecasting in
smart city energy management, IEEE Systems Journal.
[90] K. Chaudhari, A. Ukil, K. N. Kumar, U. Manandhar, S. K. Kollimalla,
Hybrid optimization for economic deployment of ess in pv-integrated ev
charging stations, IEEE Transactions on Industrial Informatics 14 (1)
(2017) 106–116.
[91] U. K. Das, K. S. Tey, M. Seyedmahmoudian, M. Y. Idna Idris, S. Mekhilef,
B. Horan, A. Stojcevski, Svr-based model to forecast pv power generation
under different weather conditions, Energies 10 (7) (2017) 876.
[92] X. G. Agoua, R. Girard, G. Kariniotakis, Probabilistic models for spatio-
temporal photovoltaic power forecasting, IEEE Transactions on Sustain-
able Energy 10 (2) (2018) 780–789.
[93] A. Ahmarinejad, A multi-objective optimization framework for dynamic
planning of energy hub considering integrated demand response program,
Sustainable Cities and Society 74 (2021) 103136.
[94] S. Aslam, M. P. Michaelides, H. Herodotou, Internet of ships: A survey
on architectures, emerging applications, and challenges, IEEE Internet of
Things journal 7 (10) (2020) 9714–9727.
[95] B.-J. Chang, J.-M. Chiou, Cloud computing-based analyses to predict ve-
hicle driving shockwave for active safe driving in intelligent transportation
system, IEEE Transactions on Intelligent Transportation Systems 21 (2)
(2019) 852–866.
[96] P. Kachroo, S. Agarwal, B. Piccoli, K. ¨
Ozbay, Multiscale modeling and
control architecture for v2x enabled traffic streams, IEEE Transactions on
Vehicular Technology 66 (6) (2017) 4616–4626.
[97] E. Cristiani, B. Piccoli, A. Tosin, Multiscale modeling of granular flows
with application to crowd dynamics, Multiscale Modeling & Simulation
9 (1) (2011) 155–182.
[98] A. Fagette, N. Courty, D. Racoceanu, J.-Y. Dufour, Unsupervised dense
crowd detection by multiscale texture analysis, Pattern Recognition Let-
ters 44 (2014) 126–133.
[99] M. S. Zitouni, H. Bhaskar, J. Dias, M. E. Al-Mualla, Advances and trends
in visual crowd analysis: A systematic survey and evaluation of crowd
modelling techniques, Neurocomputing 186 (2016) 139–159.
[100] A. Khan, B. Rinner, A. Cavallaro, Multiscale observation of multiple mov-
ing targets using micro aerial vehicles, in: Intelligent Robots and Systems
(IROS), 2015 IEEE/RSJ International Conference on, IEEE, 2015, pp.
[101] A. Hampapur, L. Brown, J. Connell, A. Ekin, N. Haas, M. Lu, H. Merkl,
S. Pankanti, Smart video surveillance: exploring the concept of multiscale
spatiotemporal tracking, IEEE signal processing magazine 22 (2) (2005)
[102] M. Picone, M. Amoretti, F. Zanichelli, Simulating smart cities with deus,
in: Proceedings of the 5th International ICST Conference on Simulation
Tools and Techniques, 2012, pp. 172–177.
[103] S. S. Bhatti, N. K. Tripathi, V. Nitivattananon, I. A. Rana, C. Mozumder,
A multi-scale modeling approach for simulating urbanization in a
metropolitan region, Habitat International 50 (2015) 354–365.
[104] L. R. Gon¸calves, P. Fidelman, A. Turra, O. Young, Correction: The dy-
namics of multiscale institutional complexes: the case of the s˜ao paulo
macrometropolitan region, Environmental Management 67 (1) (2021) 119–
[105] A. Khan, N. Javaid, A. Ahmad, M. Akbar, Z. A. Khan, M. Ilahi, A
priority-induced demand side management system to mitigate rebound
peaks using multiple knapsack, Journal of Ambient Intelligence and Hu-
manized Computing 10 (4) (2019) 1655–1678.
[106] A. Khan, N. Javaid, A. Ahmed, S. Kazmi, H. M. Hussain, Z. A. Khan,
Efficient utilization of hem controller using heuristic optimization tech-
niques, in: International Conference on Emerging Internetworking, Data
& Web Technologies, Springer, 2017, pp. 573–584.
[107] S. Aslam, A. Khalid, N. Javaid, Towards efficient energy management
in smart grids considering microgrids with day-ahead energy forecasting,
Electric Power Systems Research 182 (2020) 106232.
[108] K. Aurangzeb, S. Aslam, S. M. Mohsin, M. Alhussein, A fair pricing
mechanism in smart grids for low energy consumption users, IEEE Access
9 (2021) 22035–22044.
[109] G. C. Peng, M. Alber, A. B. Tepole, W. R. Cannon, S. De, S. Dura-Bernal,
K. Garikipati, G. Karniadakis, W. W. Lytton, P. Perdikaris, et al., Mul-
tiscale modeling meets machine learning: What can we learn?, Archives
of Computational Methods in Engineering (2020) 1–21.
[110] M. Abdar, F. Pourpanah, S. Hussain, D. Rezazadegan, L. Liu,
M. Ghavamzadeh, P. Fieguth, A. Khosravi, U. R. Acharya,
V. Makarenkov, et al., A review of uncertainty quantification in
deep learning: Techniques, applications and challenges, arXiv preprint
[111] E. Begoli, T. Bhattacharya, D. Kusnezov, The need for uncertainty quan-
tification in machine-assisted medical decision making, Nature Machine
Intelligence 1 (1) (2019) 20–23.
[112] A. K. Raz, P. C. Wood, L. Mockus, D. A. DeLaurentis, System of systems
uncertainty quantification using machine learning techniques with smart
grid application, Systems Engineering 23 (6) (2020) 770–782.
[113] H. Quan, A. Khosravi, D. Yang, D. Srinivasan, A survey of computational
intelligence techniques for wind power uncertainty quantification in smart
grids, IEEE Transactions on Neural Networks and Learning Systems.
[114] B. Wang, J. Lu, Z. Yan, H. Luo, T. Li, Y. Zheng, G. Zhang, Deep uncer-
tainty quantification: A machine learning approach for weather forecast-
ing, in: Proceedings of the 25th ACM SIGKDD International Conference
on Knowledge Discovery & Data Mining, 2019, pp. 2087–2095.
[115] K. Akyol, Comparing of deep neural networks and extreme learning ma-
chines based on growing and pruning approach, Expert Systems with Ap-
plications 140 (2020) 112875.
[116] S. Rajaraman, J. Siegelman, P. O. Alderson, L. S. Folio, L. R. Folio, S. K.
Antani, Iteratively pruned deep learning ensembles for covid-19 detection
in chest x-rays, arXiv preprint arXiv:2004.08379.
[117] M. S´anchez-Guti´errez, P. P. Gonz´alez-P´erez, Discriminative neural net-
work pruning in a multiclass environment: A case study in spoken emotion
recognition, Speech Communication.
... It is important to strike a balance between the goals of the subsystems and bring the whole system closer to the desired goals. Based on previous literature, in general, there are two main methods for modeling cities-analytical and heuristic ( Khana et al. 2021). In an analytical model involving the state of a system, examining all Input variables and measurement uncertainty for complex systems such as smart cities are very challenging (Rzevski 2016). ...
Full-text available
In smart city architecture, information and communication technologies are used to improve living standards and its management by citizens and government. Most researchers have divided this structure into six main components: smart people, the smart government, smart environment, smart transportation, smart economy, and smart life. Due to the connection between smart cities and the challenges resulting from their implementation and especially its integration, there exists no perfect solution for the concept of an integrated smart city so far according to our studies. Some more general concepts such as security, ICT infrastructure, and knowledge are not seen integrative in these structures. Therefore, it seems that new sub-components and general extra-components should be added to the existing models to form an integrated structure in such a way that the executive projects are located in their proper place in this structure and create and guarantee the integration of the smart city. Therefore, the requirements engineering of the smart city can also be explained more precisely. This study presents a model of an integrated graph in such a way that besides maintaining and improving the model of the smart city and existing models, it will fully cover the integration and requirements engineering and methodologies of the smart city in the future. The present paper offers an upgraded model of a six-component smart city structure as a flexible integrated dynamic graph so that beside maintaining the features of existing smart city models, it ensures its integrity, dynamism, flexibility and performance and prevents the failure of smart operations. Due to its flexibility, adaptability and localization, the proposed model presented in this paper can create an integrated solution and facilitating the life cycle of executive systems and enable governments and communities to predict and prevent sudden events such as natural disasters, pandemics like Covid-19 and the like as well as managing and leading their target community in the best way.
Full-text available
Climate change and the rapid urbanization process can trigger the current urban heat island effect. An urban heat island is categorized by temperature variation between urban and rural environments. Urban areas are mostly hotter than rural areas due to urbanization, population pressure, industrialization, vegetation insufficiency, and transportation system. Thermal remote sensing was used to carry out the exploration of urban heat islands over Kolkata municipality and surrounding urban and rural-urban fringe areas, India. This study was identify the land use/land cover change dynamics and land surface temperature analysis as well as the correlation analysis of LST, LULC, NDVI, and NDBI. Built-up areas were located 22.01% (1990), 31.44% (2000), 44.21% (2010) and 55.768% (2020). The Annual temperature was increased 2.01°C per year. The relationship between LST and NDVI was showing a negative correlation because of the vegetated area affected due to urban expansion. The relation between LST and NDBI resulted in a positive correlation because of anthropogenic activities. The CA-ANN was used for LULC prediction of the year 2030 and 2050 for future urban expansion and environmental degradation. Those areas need proper planning for better healthy livelihood and also protect our environment otherwise climate change will affect the entire location.
Full-text available
Land use and transport integration (LUTI) has been promoted around the world as essential for sustainable urbanization. Using Seoul as a case study, this paper critically examines if and how LUTI policy can bring positive social outcomes, including accessibility and quality of life, in rapidly-growing megacities. We explored stakeholder views about the experience of these outcomes in different localities, and about critical elements and processes needed to enhance social outcomes. Our research identified that facilitating compact urban form around transport nodes does not guarantee increased accessibility or livelihood opportunities. The outcomes are closely related to the quality of development around nodes, accessibility to public facilities, local mobility, and functionality of pedestrian paths. Ensuring sustainability of social outcomes from LUTI approach requires: balancing restrictions and incentives to control quality of development around nodes; creating cost-effective strategies to maintain local environmental quality; applying flexible rules to address varying priorities and opportunities in different localities; and multi-level planning that balances the responsibilities of metropolitan and local stakeholders to facilitate desired outcomes. We conclude that, in rapidly-growing cities, a flexible and holistic approach to integration should be applied to reflect diverse needs and local circumstances and to ensure fair benefits across the city.
Full-text available
Based on energy demand, consumers can be broadly categorized into low energy consumers (LECs) and high energy consumers (HECs). HECs use heavy load appliances, e.g., electric heaters and air conditioners, and LECs do not use heavy load appliances. Thus, HECs demand more energy compared to LECs. The usage of high energy consumption appliances by HECs leads to peak formation in various time intervals. Different pricing schemes, i.e., time of use (ToU), real time pricing (RTP), inclined block rate (IBR), and critical peak pricing (CPP), have been proposed previously. In ToU, an energy tariff is divided into three blocks, i.e., on-peak (high rates), off-peak (low rates), and mid-peak (between on-peak and off-peak rates) hours, and these rates are applied to all electricity users without distinction. The high energy demand by HECs causes the high peak formation; thus, higher rates should be applied to only HECs rather than all consumers, which is not the case in existing billing mechanisms. LECs are also charged higher rates in on-peak intervals and this billing mechanisms are unjustified. Thus, in this paper, a fair pricing scheme (FPS) based on power demand forecasting is developed to reduce extra bills of LECs. First, we developed a machine learning-based electricity load forecasting method, i.e., an extreme learning machine (ELM), in order to differentiate LECs and HECs. With the proposed FPS, electricity cost calculations for LECs and HECs are based on the actual energy consumption; thus, LECs do not subsidize HECs. Simulations were conducted for performance evaluation of our proposed FPS mechanism, and the results demonstrate LECs can reduce electricity cost up to 11.0075%, and HECs are charged relatively higher than previous pricing schemes as a penalty for their contribution to the on-peak formation.
Full-text available
There is widespread concern in academia and the media in general that in the future many coastal communities will be forced to relocate in the face of rising water levels. However, there is little evidence of any relocation actually taking place, even when there are a number of examples of coastal areas being regularly flooded due to relative sea level rise caused by groundwater extraction or earthquake induced subsidence. To better understand the consequences of future sea level rise this keynote speech will analyse a number of instances of land subsidence that have taken place in the 20th and early 21st centuries (such as low-lying coastal areas of Tokyo, ports in Jakarta, and the experience of the small coral islands on the Danajon bank in the Philippines). In all cases the inhabitants of such densely populated coastal areas remain in place, despite the challenge of living with higher water levels. Through such case studies the actual adaptation pathways of both residential areas and ports can be better understood. Thus, while it is clear that sea level rise will pose an additional financial strain on urbanised coastal areas, there is presently no evidence that any major coastal settlements will surrender a significant portion of their land area to the sea, given the range of adaptation options available. Rather, the opposite would appear to be true, and that new lines of defence will be built further into the water, effectively meaning that humans will continue to encroach on the sea.Recorded Presentation from the vICCE (YouTube Link):
Full-text available
The climate emergency and population growth are challenging water security and sustainable urban design in cities worldwide. Sustainable urban development is crucial to minimise pressures on the natural environment and on existing urban infrastructure systems, including water, energy, and land. These pressures are particularly evident in London, which is considered highly vulnerable to water shortages and floods and where there has been a historical shortage of housing. However, the impacts of urban growth on environmental management and protection are complex and difficult to evaluate. In addition, there is a disconnection between the policy and decision-making processes as to what comprises a sustainable urban development project. We present a systems-based Urban Planning Sustainability Framework (UPSUF) that integrates sustainability evaluation, design solutions and planning system process. One of the features of this master planning framework is the spatial representation of the urban development in a Geographical Information System to create an operational link between design solutions and evaluation metrics. UPSUF moves from an initial baseline scenario to a sustainable urban development design, incorporating the requirements of governance and regulatory bodies, as well as those of the end-users. Ultimately, UPSUF has the potential to facilitate partnership between the public and the private sectors.
Full-text available
Since air pollutants are difficult and expensive to control, a strong scientific underpinning to policies is needed to guide mitigation aimed at reducing the current burden on public health. Much of the evidence concerning hazard identification and risk quantification related to air pollution comes from epidemiological studies. This must be reinforced with mechanistic confirmation to infer causality. In this review we focus on data generated from four contrasting sources of particulate air pollution that result in high population exposures and thus where there remains an unmet need to protect health: urban air pollution in developing megacities, household biomass combustion, wildfires and desert dust storms. Taking each in turn, appropriate measures to protect populations will involve advocating smart cities and addressing economic and behavioural barriers to sustained adoption of clean stoves and fuels. Like all natural hazards, wildfires and dust storms are a feature of the landscape that cannot be removed. However, many efforts from emission containment (land/fire management practices), exposure avoidance and identifying susceptible populations can be taken to prepare for air pollution episodes and ensure people are out of harm's way when conditions are life-threatening. Communities residing in areas affected by unhealthy concentrations of any airborne particles will benefit from optimum communication via public awareness campaigns, designed to empower people to modify behaviour in a way that improves their health as well as the quality of the air they breathe.
The integrated operation of multi-carrier energy systems has attracted special attention due to its increased efficiency, improved flexibility and reduced pollution. It should be noted that accurate and optimal design of these systems will have a high impact on their optimal operation. Hence, this paper presents a multi-objective optimization framework for long-term planning of energy hub, in which equipment degradation and integrated demand response (IDR) programs are considered. The planning problem has been solved for 20 years and all short-term operation constraints have been taken into account. The problem is modeled using the fuzzy max-min method as a three-objective optimization problem, in which the objective functions include total cost, emission and hub losses. The simulation results demonstrate that the IDR program has reduced the total cost by about 9% by reducing equipment capacity and operating costs. In addition, the results illustrate that the reduction of 29% and 31% of losses and emission, led to a 49% increase in total costs. Overall, the results demonstrate that the proposed model, considering the equipment degradation and the annual load growth, leads to the optimal design of the hub structure and guarantees the supply of load until the end of the planning period.
Having intensive economic development and rapid urbanization, the Yangtze River basin, namely the heart of China’s prosperity, has faced challenges in the accompanying deterioration of water security. How to closely inspect the features and development of water security for the major cities in the basin and to compare the water security conditions between the major cities at the basin scope is a keystone to better support water management practice in the cities as well as regions. Hence, this study refined the previous framework by applying 19 indicators to describe the conditions of resource, infrastructure, waterway, efficiency, risk, and capacity and then integrated the data-driven weighting approach, the Criteria Importance Through Inter-criteria Correlation (CRITIC) method, to objectively evaluate the development and characteristics of water security of the megacities in the Yangtze River basin, i.e., Shanghai, Nanjing, Wuhan, Chongqing, and Chengdu, during 2011-2017. Based on the aggregated scores, Chongqing had the best overall water security condition (0.696) in 2017, followed by Chengdu (0.613), Shanghai (0.581), Nanjing (0.496), and Wuhan (0.471). During 2011-2017, Chongqing and Shanghai had a greater improvement in the water security condition, while Wuhan had the least. From a basin perspective, the upstream megacities had the advantage of their water availability and depletion conditions, river quality, pollutant discharge, government's support of water affairs, and the societal investment in water conservancy. On the other hand, the middle- and downstream megacities had shown the better performance of the water affordability, the density of the sewage network, and water intensity. The sensitivity analysis detected the average of the standard variations of the score changes as 2.03% in the context of different indicator sets and thus assured the outcome robustness. This study enhances the assessment frameworks, facilitates the applications of temporal and spatial comparative evaluation of water security conditions on city and river basin level, and identifies the policy gaps for enhancing water management in the magacities and the basins.
System‐of‐Systems capability is inherently tied to the participation and performance of the constituent systems and the network performance which connects the systems together. It is imperative for the SoS stakeholders to quantify the SoS capability and performance to any uncertain variations in the system participation and network outages so that the system participation is incentivized and network design optimized. However, given the independent operations, management, and objectives of constituent systems, along with an increasing number of systems that collectively become a part of SoS, it becomes difficult to obtain a closed analytical function for SoS performance characterization. In this paper, we investigate and compare two machine learning techniques, Artificial Neural Network and Parametric Bayesian Estimation, to obtain a predictive model of the SoS given the uncertainty in the constituent system participation and the network conditions. We demonstrate our approach on a smart grid SoS application example and describe how the two machine learning techniques enable SoS robustness and resilience analysis by quantifying the uncertainty in the model and SoS operations. The results of smart grid example establish the value of SoS uncertainty quantification (UQ) and show how smart grid operators can utilize UQ models to maintain the desired robustness as operating conditions evolve and how the designers can incorporate low‐cost networks into the SoS while maintaining high performance and resilience.