sustainability

Article
Assessing Spatial Variations of Traffic Congestion Using Traffic
Index Data in a Developing City: Lessons from Johannesburg,
South Africa
Thembani Moyo 1, * , Siphiwe Mbatha 2 , Oluwayemi-Oniya Aderibigbe 2 , Trynos Gumbo 2
and Innocent Musonda 1

1 Centre for Applied and Research Innovation in the Built Environment (CARINBE), Department of
Construction Management and Quantity Surveying, University of Johannesburg, Corner Siemert & Beit
Streets, Doornfontein, Johannesburg 0184, South Africa; [email protected]
2 Sustainable and Smart Cities and Regions Research Group, Department of Urban and Regional Planning,
University of Johannesburg, Corner Siemert & Beit Streets, Doornfontein, Johannesburg 0184, South Africa;
[email protected] (S.M.); [email protected] (O.-O.A.); [email protected] (T.G.)
* Correspondence: [email protected]

Abstract: The COVID-19 pandemic has created unforeseen effects in public transport and the mobility
of people in cities globally. Johannesburg, being a developing city in one of the most affected countries
in Africa during the pandemic, has experienced severe changes in traffic management and travel
patterns as a result of the restrictions imposed on movement. Hence, this study examined the spatial
variations in traffic during the pandemic. The study utilized data obtained from the TomTom Traffic
Citation: Moyo, T.; Mbatha, S.;
Index for the city of Johannesburg from 2019 to 2021, with 2019 representing the period pre-COVID-19
Aderibigbe, O.-O.; Gumbo, T.;
with no lockdown restrictions, 2020 representing the period with restricted movement to limit spread
Musonda, I. Assessing Spatial
Variations of Traffic Congestion
of COVID-19, and 2021 representing a period of relaxed and minimized restrictions on movement.
Using Traffic Index Data in a Our findings revealed that there was a variation in congestion levels between 2019–2021 with year
Developing City: Lessons from 2020 having the least congestion from the beginning of the COVID-19 restrictions due to regulations
Johannesburg, South Africa. enforced in movement and reduced travel. Our findings further revealed that traffic congestion was
Sustainability 2022, 14, 8809. higher during weekdays than weekends during the three periods, with mini-bus taxis as the major
https://doi.org/10.3390/ contributors to congestion. Consequently, there is a need to discourage the use of single occupancy
su14148809 vehicles and invest in more sustainable means of transportation to ease the mobility of people and
Academic Editors: Mladen Jardas, reduce traffic on major roads.
Pietro Evangelista, Predrag Brlek,
David Brčić, Zlatko Sovreski and Keywords: congestion; traffic index; COVID-19 pandemic; city of Johannesburg
Ljudevit Krpan

Received: 4 May 2022

Accepted: 12 July 2022
1. Introduction
Published: 19 July 2022
Cities worldwide have found themselves on the brink of massive traffic expansions
Publisher’s Note: MDPI stays neutral
which has threatened their ability to manage traffic. The situation has worsened in devel-
with regard to jurisdictional claims in
oping regions due to already prevailing traffic congestion [1]. Traffic overcrowding refers
published maps and institutional affil-
to a situation where road networks converge and a large amount of traffic needs to traverse
iations.
through the area, thus inhibiting free traffic flow [2,3]. Consequently, Jain et al. [4] asserted
that poor traffic management systems are one of the causes of traffic congestion in most
developing and third world countries, thus preventing the free flow of traffic along major
Copyright: © 2022 by the authors.
roads and massive traffic jams for prolonged time periods.
Licensee MDPI, Basel, Switzerland. The optimized interaction between traffic and its infrastructure without interruption
This article is an open access article is characterized as traffic flow [5]. Thus, the study of traffic flow is the foundation for
distributed under the terms and understanding the behavior of traffic (pedestrians, public transport, and cargo) in a system.
conditions of the Creative Commons This behavior is primarily influenced by the users of such a system and by external influ-
Attribution (CC BY) license (https:// ences. Chief among these is the way in which infrastructure and users interact [6]. Traffic
creativecommons.org/licenses/by/ engineering, therefore, pertains to the analysis of the behavior of traffic and the design of
4.0/). facilities in a bid to facilitate the smooth, safe, and economical operation of traffic [5,7].

Sustainability 2022, 14, 8809. https://doi.org/10.3390/su14148809 https://www.mdpi.com/journal/sustainability

Sustainability 2022, 14, 8809 2 of 16

Aging physical transportation infrastructure such as the conditions of roads, high-

ways, or bridges are often blamed as one of the inherent causes of traffic congestion [8].
However, congestion is also partly responsible for accelerating the physical degradation of
transportation infrastructure and consequently for reducing the transportation network
performance [9,10]. According to a study by [11], there have been several recovery strate-
gies aimed at improving the damaged network performance, but most of these strategies
are not sufficient or even might not be applicable for congested road traffic conditions.
Furthermore, Kumarage [12] asserted that traffic congestion is a global issue which
challenges the development of a resilient and sustainable transportation system. Therefore,
the main thrust of this research paper was to contribute to the development of a sustainable
and resilient transportation management system which can minimize the negative impact
of congestion and ensure free traffic flow on major roads in Johannesburg. In this paper, we
examined the COVID-19 pandemic’s effect on traffic congestion in the city of Johannesburg,
comparing the changes in congestion volumes from 2019 to 2021.

Related Work
Due to the incremental challenges of urbanization, there has been a growth in trans-
portation studies to improve the quality of service. Research by [13] has revealed that
areas in and around central business districts suffer from peak hour traffic congestion.
This has now become a common burden that is hardly avoidable for many residents. To
mitigate such challenges, experts and scholars have focused on intelligent transportation
systems to analyze historical and real-time mobility trends so as to develop reactive sys-
tems to improve traffic flow conditions. This is evidenced by the growth in traffic research
using geographic information systems [14,15], traffic sensors [16], and traffic control mea-
sures [17]. Research by [18] defined traffic flow as the interactions between the travelers
(for both commuting and recreational trips) and their surrounding infrastructure which
supports travelling. These surroundings include road infrastructure such as buildings,
traffic control, stations, and bus stops, and the objective has been to develop an optimal
transportation system with efficient movement of traffic. Moreover, the level of traffic flow
can also be expressed as a function of free flow and exposure to obstacles that limit free
flow (these obstacles can be traffic control measures or congestion). According to [19–21],
such examples of studies to improve traffic flow have focused on developing prediction
and estimation models. From these studies, it is evident that a widely used deterministic
model of traffic flow systems utilizes discrete-time meanings, such as time of day [19,20],
month, and season [21] or discrete location [22]. Similar to these studies, we sought in
this paper to visualize traffic flow trends during the peak hours and also to assess any
seasonal variations. Regarding location, studies have revealed that various locations in
the city experience congestion differently [4,12,22]. This has led route planners to develop
adaptive response systems that redirect traffic from heavily congested to alternative routes.
Several scholars have identified variables affecting traffic flow conditions, although
these vary in relation to the specific area of analysis. For example, Zhang et al. [22] outlined
a relationship between congestion and land use. This led to urban planners seeking to
develop transit-oriented cities to ensure that locations along economic corridors in the city
have adequate transportation capacity [23,24]. Consequently, city authorities have now
begun using the nodal approach to planning, as it promotes accessibility between residential
areas and workplaces. In Johannesburg, South Africa, the city has invested in Transit
Oriented Development (TOD) as a means to improve economic activities along selected
corridors in the city [25]. Research by [26] highlighted how these TODs have also been
used as tools for addressing spatial inequalities by linking residents with areas of economic
opportunities. Land use has been identified as a variable that has a functional relationship
with the level of traffic flow. Therefore, planners need to ensure a balance of investment in
economic areas and the supporting public transportation infrastructure. However, given
the merits of the nodal approach to development, uncertainty still exists as to the possibility
Sustainability 2022, 14, 8809 3 of 16

of creating adequate nodes that have a critical mass to impact developmental trajectories
significantly [25,27].
COVID-19 has led to researchers tapping into new data collection approaches and
analyses. Table 1 outlines the emerging approaches to assess trends that have emerged in
transportation planning during to COVID-19 pandemic.

Table 1. Emerging approaches in transportation planning.

Method Basic Idea and Content Authors

Using supervised learning and unsupervised learning to
Machine Learning and Deep Learning assess the impacts of the COVID-19 pandemic on human [28–30]
mobility and on air quality
Assessing the implications of the COVID-19 pandemic for
different regions and identification of measures or actions
Case Study [31–34]
that help policymakers and traffic planners to better develop
sustainable strategies to improve transport system efficiency
Evaluation of Internet of Things concepts and the 5v’s of big
data for leveraging the sustainability and resilience of urban
Internet of Things and Big Data [35–37]
mobility systems and providing directions to future
research agendas

Given the above-mentioned challenges to traffic flow, city authorities during the
COVID-19 pandemic shifted to measures that reduced the need for travel as a response to
reduce transmission [38]. This had an impact on how people conduct their work, with more
people working from home as compared to the office. This dramatically reduced congestion
on major roads in cities [39]. Lessons learned from this shift revealed some possibility of
people traveling at different times to reduce travel demand during the morning and peak
hours.

2. Materials and Methods

The metropolitan city of Johannesburg is at the heart of Gauteng, South Africa’s
economic hub. Given the recent COVID-19 pandemic, urban mobility has been affected
by several restrictions to movement. The paper assessed the impacts of the lockdown
regulations on movement, and Table 2 summarizes the COVID-19 movement restriction
timeline in South Africa from 2019 to 2021.
The paper used the TomTom Traffic Index  for the city of Johannesburg from 2019 to
2021, with 2019 representing the period pre-COVID-19 with no lockdown restrictions, 2020
representing the period with restricted movement to limit the spread of COVID-19, and
2021 presenting a period with relaxed movement restrictions. The TomTom Traffic Index e
reflects the extra average travel time which is calculated from the baseline for the city by
analyzing free-flow travel times of all vehicles on the entire road network [40]

e = ( Ta − T f f )/T f f (1)

where Ta represents the actual travel time, which is calculated from the baseline for the
city by analyzing free-flow travel times T f f of all vehicles on the entire road network. To
visualize the congestion level in the city, we assumed Ta > 0 as time is discrete. For each
road network, we denote the density of vehicles per km2 at time T which results in the state
of the road network. It should also be noted the density of vehicles changes dynamically
over time. If a model of travel time Ta is available, then one can deduce a trend graph
for low, moderate, and high congestion levels. The main assumption of the time series
is that an observation at a given time depends only on its previous values and a random
noise. Hence, curve fitting with radial basis functions was used for the model for the time
series [41,42].
Sustainability 2022, 14, 8809 4 of 16

Table 2. Timeline of COVID-19 movement restriction in South Africa.

Date Event
December 2019 WHO announces outbreak of COVID-19 in Wuhan City, China.
South Africa records first COVID-19 case, and President declares National lockdown (21-day stay at
March 2020
home lockdown).
First recorded COVID-19 death. President outlines a phased relaxation of the COVID-19 lockdown
April 2020
restrictions and details a 5-level alert system.
May 2020 South Africa to Level 4
June 2020 South Africa moves to Level 3, Domestic air travel permitted for business purposes.
South Africa becomes the 5th worst country affected by the COVID-19 globally with more than
July 2020
360,000 infections.
August 2020 South Africa moves to Level 2.
September 2020 South Africa moves to Level 1.
December 2020 2nd wave of COVID-19 and South Africa moves to an adjusted level 3 lockdown.
January 2021 South Africa’s vaccine plan revealed.
March 2021 South Africa moves to Level 1.
June 2021 South Africa moves to Level 2.
July 2021 South Africa moves to adjusted Level 4 and 3.
September 2021 South Africa moves to adjusted Level 3 and 2.
October 2021 South Africa moves to adjusted Level 1.

Level of service (LOS) was determined as the saturation flow rate, expressed between
the range 0 to 1 calculated using the below equation.

s = so f w f p f bb (2)

where, so = adjusted saturation flow rate (number of vehicles per hour on road segment),
and f w = adjustment factor for road width (all road widths were taken as 1 as all roads had
a standardized width). f p = adjustment factor for existence of parking activity adjacent
to road. f bb = adjustment factor for blocking effect on road such as accident, stationery
vehicle, etc. (this was rated based on number of reported blocks on the road and then
expressed as a percentage).
To validate traffic data, standard measurement devices for traffic, namely loop detec-
tors, magnetometers, radar traffic detectors, or video detection systems, were used. These
were positioned at fixed and predefined positions in the network to monitor a section of
road, and they are able to detect and assign a timestamp to the event such as a vehicle
crossing the road section. Information is then aggregated in time slots.

3. Results
3.1. Level of Service
A spatial analysis of congestion trends for the city of Johannesburg in 2020 revealed
congestion trends during peak hours. Figure 1 shows the average congestion during
the morning peak hour from 06:00 to 08:00 and the corresponding average speed limit
in the city. During the morning peak hour, most of the freeway roads such as the M1
north, N1 northwestern bypass, and N3, northeastern bypass were highly congested. In
Johannesburg, freeways connect commuters from residential areas to places of economic
opportunities. An example is the M1 which connects commuters residing in the low-income
residences located south of Johannesburg (such as Soweto, Orange farm, and Southgate)
with the inner city (Braamfontein CBD and Rosebank CBD) and the northern parts of
Johannesburg (Sandton and Midrand) which are characterized as economic nodes with
 city. During the morning peak hour, most of the freeway roads such as the M1 north, N1
northwestern bypass, and N3, northeastern bypass were highly congested. In Johannes-
burg, freeways connect commuters from residential areas to places of economic opportu-
nities. An example is the M1 which connects commuters residing in the low-income resi-
Sustainability 2022, 14, 8809 dences located south of Johannesburg (such as Soweto, Orange farm, and Southgate) 5with of 16
the inner city (Braamfontein CBD and Rosebank CBD) and the northern parts of Johan-
nesburg (Sandton and Midrand) which are characterized as economic nodes with mixed
land
mixed use. Hence,
land the M1the
use. Hence, south
M1was also
south washeavily congested
also heavily duringduring
congested the evening peak hour,
the evening peak
as
hour, as commuters were traveling to their homes. Additionally, there was high congestion
commuters were traveling to their homes. Additionally, there was high congestion
along
along major
major roads
roads in
in and
and around
around the Fourways, Illovo,
the Fourways, Illovo, Midrand,
Midrand, and
and Sandton
Sandton suburbs
suburbs
which have a mix of residential and business land uses.
which have a mix of residential and business land uses.

(a) (b)
Figure
Figure 1.
1. (a)
(a)Average
Averagecongestion over
congestion over12 12
months during
months the the
during morning peakpeak
morning hourhour
(06:00 a.m. to
(06:00 08:00
a.m. to
a.m.) in 2020, (b) average speed limit.
08:00 a.m.) in 2020, (b) average speed limit.

A
A spatial
spatialanalysis
analysisofofcongestion
congestion trends
trendsduring
duringthethe
afternoon peakpeak
afternoon between 12:0012:00
between and
14:00 revealed that major roads along business land uses were highly congested
and 14:00 revealed that major roads along business land uses were highly congested (see (see Fig-
ure 2). 2).
Figure Notable
Notableexamples
examples include
includethethe
Braamfontein
BraamfonteinCBD,CBD,Rosebank
RosebankCBD,
CBD,and
and Sandton
Sandton
CBD areas.
areas. Furthermore,
Furthermore,access
accessroads
roadsfrom
fromthe
the major
major freeways
freeways (M1,
(M1, N1,N1,
N3,N3,
M2,M2,
N17,N17,
and
and the M5)
the M5) werewere also highly
also highly congested.
congested. The density
The density of various
of various road classes
road classes and theand the re-
respective
spective speed
speed limits limits demonstrated
demonstrated some intriguing
some intriguing results.with
results. Locations Locations
minor with
road minor
densityroad
that
density thatlimits
have speed haveranging
speed limits
from 18ranging
km/h to from 18 km/h
40 km/h weretoassociated
40 km/h with
werelower
associated with
congestion,
while those with a higher road density and higher speed limits from 60 km/h to 120 km/h
tended to be associated with more congestion.
Sustainability 2022, 14, x FOR PEER REVIEW 6 of 15

Sustainability 2022, 14, 8809 6 of60

lower congestion, while those with a higher road density and higher speed limits from 16

km/h to 120 km/h tended to be associated with more congestion.

(a) (b)
Figure
Figure 2.
2.(a)
(a)Average
Averagecongestion
congestionover 12 12
over months during
months the the
during afternoon peakpeak
afternoon hourhour
(12:00(12:00
p.m. to 14:00
p.m. to
p.m.) in 2020, (b) average speed limit.
14:00 p.m.) in 2020, (b) average speed limit.

During
During the the evening
evening peak
peak hour,hour, the
the major
major freeways
freeways suchsuch asas the
the M1
M1 south;
south; the
the N1N1
south,
south, thethe western
western bypass;
bypass; and and thethe N3
N3 south,
south, eastern
eastern bypass
bypass were
were heavy
heavy congested
congested as as
commuters
commuters traveled
traveled from
from their
their workplaces
workplaces to to residential
residential areas
areas in
in the
the south
south ofof the
the city
city (see
(see
Figure
Figure 3).3). The
The merging
merging roads
roads (ramps)
(ramps) to to these
these freeways
freeways may may experience
experience congestion
congestion due due
to the increase
to the increase in demand to access the freeways. Road
Road segments along residential areas
segments along residential areas
with
with speed
speedlimits
limitsbetween
between1818km/h km/h to 40
to km/h
40 km/h werewere
less congested. Moreover,
less congested. majormajor
Moreover, pub-
lic transportation
public transportation routes served
routes served by by
mini-bus
mini-bus taxis
taxiswere
werealso
alsohighly
highlycongested
congestedduring
during thethe
evening.
evening. The mini-bus
mini-bus taxi
taxi isisregarded
regardedasaspart partofofthe
theinformal
informal public
public transportation
transportation sys-
system
tem
and and as such,
as such, therethere
has has
beenbeen limited
limited control
control overover
theirtheir operation.
operation. Mini-bus
Mini-bus taxitaxi drivers
drivers are
are
paidpaid on commission;
on commission; therefore,
therefore, theythey
havehave resorted
resorted to making
to making their their
own own decisions
decisions on a
on a daily
daily
basis basis
as to as
how to they
how they operate
operate so asso toas to make
make the most
the most money.
money. The The result
result of this
of this free-
freedom
dom regarding
regarding operations
operations has the
has been been the increased
increased number number of mini-bus
of mini-bus taxis on taxis
theon the roads,
roads, which
has ledhas
which to congestion.
led to congestion.
 x FOR PEER REVIEW
Sustainability 2022, 14, 8809 7 7of
of 15
16

(a) (b)
Figure
Figure 3.
3. (a)
(a)Average
Averagecongestion over
congestion over1212
months during
months thethe
during evening peakpeak
evening hourhour
(166:00 p.m. p.m.
(166:00 to 18:00
to
p.m.) in 2020, (b) average speed limit.
18:00 p.m.) in 2020, (b) average speed limit.

3.2. Analysis of Congestion

Figure 4 depicts the variations in congestion levels for two epochs: the the period
period without
without
movement
movement restrictions
restrictions or limited restrictions
restrictions during
during 20192019 and 2021 and the period when
movement was
movement wasrestricted
restrictedinin2020.
2020. During
During the the months
months of January
of January and March
and March for the forthree
the
three periods,
periods, therefew
there were were few variations
variations in congestion
in congestion levels trends,
levels trends, as theseasperiods
these periods had
had similar
similarwith
traffic traffic withtolittle
little to no movement
no movement restrictions
restrictions beingbeing imposed.
imposed. GivenGiven this similarity,
this similarity, we
we could assume that the average congestion in a month does not
could assume that the average congestion in a month does not change for different years. change for different
years.
We then We then
used theused
datathe
fromdatathefrom
year the
2019year 2019
as the as the
normal normal congestion
congestion level for the level for the
respective
respective
months andmonths
the data andin the
2020data in 2020
to assess to assess
whether thewhether
movement the movement
restrictionsrestrictions had a
had a significant
significant
impact impact on levels
on congestion congestion
in thelevels in the
city. We usedcity.
theWedata used
fromthe datatofrom
2021 assess2021
the to assess
implica-
the implications
tions of relaxing of relaxing movement
movement restrictions.restrictions. Figure the
Figure 4 outlines 4 outlines
observed thetrends.
observed trends.
These re-
These revealed that movement restrictions led to reduced congestion
vealed that movement restrictions led to reduced congestion in the city. However, with in the city. However,
with continued
continued relaxation
relaxation of movement
of movement restrictions,
restrictions, congestion
congestion levels
levels returned
returned to the
to the normnorm
of
of 2019. Additionally, there was a correlation between an increase
2019. Additionally, there was a correlation between an increase in congestion with further in congestion with
further relaxation
relaxation of movement of movement restrictions.
restrictions. The dataThealsodata also revealed
revealed that the introduction
that the introduction of move-
of movement restrictions alone was not enough to reduce congestion;
ment restrictions alone was not enough to reduce congestion; there is a need to improve there is a need to
public transportation facilities to make them more attractive to the general public. public.
improve public transportation facilities to make them more attractive to the general
Sustainability
Sustainability 2022,
2022, 14,
14, x8809
FOR PEER REVIEW 8 8ofof15
16

Figure 4. Average congestion levels per month.

Figure 4. Average congestion levels per month.
The city of Johannesburg is one of the most populated cities in the Republic of South
The city of Johannesburg is one of the most populated cities in the Republic of South
Africa (RSA) due to the different kinds of opportunities afforded, as the city attracts people
Africa (RSA) due
from different to theofdifferent
corners the Republickinds andof the
opportunities
world. During afforded, as the COVID-19
2019, when city attracts peo-
started
ple from different
globally, the RSA was corners of the Republic
not affected and the world.
by the pandemic in any way, Duringand2019, when COVID-19
everything functioned
started globally, the RSA was not affected by the pandemic
normally. Specifically, traffic congestion was very high, which was a normal in any way, and everything
problem
functioned
throughoutnormally.
the year. Specifically, traffic congestion was very high, which was a normal
problem throughout
In 2020, the RSAthe wasyear.
affected by the COVID-19 pandemic. For the first three months
In 2020, the RSA was
(January, February, and March), affected thebycountry
the COVID-19
was stillpandemic.
functioning For the first as
normally three months
in previous
(January,
years, and February,
the traffic and March), the
congestion was country
at its was
normalstillrate
functioning
which was normally
high. as In in previous
late March,
years,
the Republic was under COVID-19 lockdown where movement was prohibited, andthe
and the traffic congestion was at its normal rate which was high. In late March, all
Republic
persons werewas under
required COVID-19
to remain lockdown
home indoors. where Anymovement
movement was from
prohibited, and alltoper-
one location the
sons
next were required
required to remain
a permit home indoors.
to be produced, and Any
as canmovement
be seen in from one4,location
Figure the level to of
thetraffic
next
required a permit to be produced, and as can be seen in Figure
congestion was at 2% which was the lowest rate at any point. From May to December, 4, the level of traffic con-
gestion was on
restrictions at 2% which was
movement weretheslowly
lowest being
rate atlifted
any point.
due toFrom May toreasons
numerous December, suchre- as
strictions
opening the on movement
economy and were slowly some
allowing beingactivities.
lifted dueHence,
to numerous
the graphreasons
trends such
wereas slowly
open-
ing the economy
increasing. and allowing
However, in 2020, thesome activities.
level Hence,
of traffic the graph
congestion was trends
very low were slowlywere
as there in-
creasing. However, in 2020,
strict regulations put in place. the level of traffic congestion was very low as there were strict
regulations
In 2021, put in place. were lifted as the economy was slowly opening, and there was
restrictions
moreInmovement
2021, restrictions
on roads. were liftedcongestion
Traffic as the economywas not wasthe slowly
sameopening, and there
as in previous was
months
more
when movement
compared with on roads. Traffic
the same months congestion wasyears
of different not the same
prior as COVID-19
to the in previouspandemic.
months
when compared
In 2021, with the same
traffic congestion wasmonths
lower as of different years prior
the government to the COVID-19
encouraged pandemic.
hybrid working to
In 2021, traffic
delineate congestion
high public was lower
engagement as the government
for individuals encouraged
due to skepticism hybrid
of the working
pandemic to
being
deadly. However,
delineate high public it is noted that with
engagement the absence
for individuals dueandto the presence
skepticism ofof
thethe pandemic,
pandemic the
being
city of Johannesburg still suffered high traffic congestion levels
deadly. However, it is noted that with the absence and the presence of the pandemic, the per month.
city of Johannesburg still suffered high traffic congestion levels per month.
3.3. Comparison of Average Hourly Congestion Level
Traffic congestion
3.3. Comparison of Averagediffered
Hourly hourly, as there
Congestion Level were different rush hours between the
mornings, afternoons, and evenings. This
Traffic congestion differed hourly, as therewas mostly
were caused
differentbyrush
individuals commuting
hours between the
from home to work or work to home and/or any place of interest. This
mornings, afternoons, and evenings. This was mostly caused by individuals commuting movement from
originhome
from to destination
to work or could
work not
tobe avoided
home by any
and/or travelers;
place hence, the traffic
of interest. This congestion
movement when
from
travelling took place at the same time. Figure 5 shows the average hourly congestion
origin to destination could not be avoided by travelers; hence, the traffic congestion when levels
from 2019took
travelling to 2021.
place at the same time. Figure 5 shows the average hourly congestion levels
from 2019 to 2021.
Sustainability
Sustainability 2022,
2022, 14,
14, x8809
FOR PEER REVIEW 9 9ofof15
16

Figure 5.
Figure Average hourly
5. Average hourly congestion
congestion (2019–2021).
(2019–2021).

The city
The city of
of Johannesburg
Johannesburg traffictraffic congestion
congestion levels levels areare not
not the
the same
same for
for the
the weekdays
weekdays
and the weekend. In 2019 and early 2020, 2019 representing
and the weekend. In 2019 and early 2020, 2019 representing the period pre-COVID-19 with the period pre-COVID-19 with
no lockdown
no lockdown restrictions,
restrictions, the the daily
daily patterns
patterns during
during weekdays
weekdays appearedappeared the the same
same as as the
the
weekend patterns. The busiest time on the city of Johannesburg
weekend patterns. The busiest time on the city of Johannesburg roads was the morning roads was the morning
around 7:00
around 7:00 a.m.
a.m.throughout
throughoutthethe weekweek as most
as mostindividuals
individualsare rushing for business,
are rushing school,
for business,
school, work, etc. As can be seen from Figure 5, the morning rush hour had overtraffic
work, etc. As can be seen from Figure 5, the morning rush hour had over 70% 70%
congestion.
traffic This was
congestion. Thisfollowed by the afternoon
was followed rush hour
by the afternoon later
rush in the
hour dayinaround
later the day17:00aroundp.m.
with the traffic congestion around 60% to 68% from Mondays
17:00 p.m. with the traffic congestion around 60% to 68% from Mondays to Thursdays, to Thursdays, and on Fridays,
the rush hour was around 16:00 p.m.
and on Fridays, the rush hour was around 16:00 p.m.
On Saturdays, the volume of traffic congestion was very low which is evident from
On Saturdays, the volume of traffic congestion was very low which is evident from
Figure 5. The index showed that the traffic congestion was not over 20%, only high early in
Figure 5. The index showed that the traffic congestion was not over 20%, only high early
the afternoon around 12:00 p.m., and kept on dropping every hour. Accordingly, Sundays
in the afternoon around 12:00 p.m., and kept on dropping every hour. Accordingly, Sun-
had lower traffic congestion compared to all the weekdays, as the high traffic congestion
days had lower traffic congestion compared to all the weekdays, as the high traffic con-
was below 14% around 11:00 a.m. Therefore, there were low traffic congestion volumes
gestion was below 14% around 11:00 a.m. Therefore, there were low traffic congestion
on weekends as people are not travelling to work, business, or school and these are the
volumes on weekends as people are not travelling to work, business, or school and these
factors contributing mostly to travel around the city. A trend was evident of low or high
are the factors contributing mostly to travel around the city. A trend was evident of low
congestion every hour from 5:00 a.m. in the morning, and levels of traffic congestion were
or high congestion every hour from 5:00 a.m. in the morning, and levels of traffic conges-
noted throughout the afternoon.
tion were
During noted
thethroughout the afternoon.
months of January and February in all the years of 2019, 2020, and 2021,
During the months of
there were few variations in congestion January and levels,
February in all periods
as these the years hadof similar
2019, 2020,
trafficand 2021,
without
there were few variations in congestion levels, as these periods
any movement restrictions being imposed. Given this similarity, we could assume the had similar traffic without
any movement
average congestionrestrictions
in a month being imposed.
does not changeGivenfor this similarity,
different years.weWecould
thenassume
used thethedata
av-
erage congestion in a month does not change for different
from the year 2019 and 2021 as the normal congestion level for the respective months andyears. We then used the data
from
used thethe year
data 2019
in 2020and to2021 as whether
assess the normal thecongestion
movement level for thehad
restrictions respective months
a significant and
impact
used the data in 2020 to assess whether the movement restrictions
on congestion levels in the city. Figures 6 and 7 show the observed trends. These revealed had a significant impact
on congestion
that the congestion levelspatterns
in the city.
were Figures
similar6 for
andall7 show the observed
the months, trends.
as essential These revealed
activities still need
that the congestion patterns were similar for all the months,
cars and contribute to congestion. During the commencement of the COVID-19 pandemic, as essential activities still need
cars
thereand wascontribute
an increase to congestion. Duringto
in the preference the commencement
use private vehicles, of the
as COVID-19 pandemic,
public transportation
there was anasincrease
was viewed a potentialin thehotpreference to use private
spot for contracting vehicles,
the virus. The as public transportation
summary data observed
was
fromviewed
2020 and as a2021
potential hot spot
were below the forobserved
contracting datathe virus.
from The
2019; summary
however, thedata observed
introduction
from 2020 and 2021 were below the observed data from 2019;
of restrictions alone was not enough to reduce congestion, and there is a need to improve however, the introduction
of restrictions
public alone was
transportation not enough
facilities in order to toreduce
makecongestion,
them more and there is
attractive toathe
need to improve
general public.
public transportation facilities in order to make them more attractive to the general public.
14, x FOR PEER REVIEW 10 of 15
14, x FOR PEER REVIEW 10 of 15
Sustainability 2022, 14, 8809 10 of 16

Figure 6. Congestion Index for period with no restriction (normal) and period with movement re-
Figure 6.(limited)
striction Congestion Index
Figure
from for to
period
6. Congestion
January June. withfor
Index noperiod
restriction (normal)
with no and
restriction period
(normal) with
and movement
period re-
with movement
restriction (limited)
striction (limited) from January to June. from January to June.

Figure 7. Congestion Figure

Index 7. Congestion
for periodIndex
with for
no period with no
restriction restriction
(normal) (normal)
and periodand period
with with movement
movement re-
Figure 7. Congestion Index
restriction for period
(limited)
striction (limited) from July to December. fromwith
July no
to restriction
December. (normal) and period with movement re-
striction (limited) from July to December.
Year 2020 was different in all aspects and experiences in the Republic of South Africa.
Year 2020 was Thedifferent in allofaspects
general trend andhigh
hours with experiences in the Republic
and low congestion of South
was similar for allAfrica.
the years,
Year 2020 was
The general trendwith different
of hours in all aspects
with lower
2020 having and
high congestion experiences
and low congestion in the Republic
was similar
due to the introduction of South Africa.
for all therestrictions.
of movement years,
The general
with 2020 havingtrendObservations
of hours
lower with
revealed
congestion highthatand
due thethe
to low congestion
restrictions ofwas
had a great
introduction similar
impact
movement for
with the all the years,
freeways
restrictions. across
Ob-the
with 2020 having city
lowerbeing empty and with very little traffic congestion in rush hour periods, allowing good
servations revealed that congestion
the restrictions due to hadthea introduction
great impactof movement
with restrictions.
the freeways
flow of movement on the roads. Notably, the major freeways such as the M1 south, the
across Ob-
the
servations
city revealed
being empty andthat thevery
with restrictions had acongestion
little traffic great impact with hour
in rush the freeways
periods, across
allowingthe
N1 south, western bypass, N3 south, and the eastern bypass which normally have heavy
city being
good flow of empty and congestion
movement
traffic with
on thevery little
roads.
had good traffic
flowingcongestion
Notably, the major
traffic in rush
freeways
throughout hour
such
the year. periods,
as the M1
Consequently, allowing
south,
where traffic
good flow of movement on the roads. Notably, the major
the N1 south, western bypass, N3 south, and the eastern bypass which normally have in
congestion took place, the movement on the freeways
roads was such
quicker as
than the M1
normal, south,
resulting
the N1traffic
heavy south,congestion
western
less timebypass,
wasted on
had good N3 south,
traffic, and and
flowing the across
traveling
traffic easternthe bypass
throughout thewhich
city was good normally
in the
year. year 2020.have
Consequently, Table 3
heavy traffic highlights
trafficcongestion
congestiontook the trend
hadplace, from
good the 2019
flowing to 2021.
trafficonthroughout the quicker
year. Consequently,
where movement the roads was than normal,
where traffic congestion took place, the movement on
resulting in less time wasted on traffic, and traveling across the city was good the roads was quicker than normal,
in the year
resulting in less time wasted on traffic,
2020. Table 3 highlights the trend from 2019 to 2021. and traveling across the city was good in the year
2020. Table 3 highlights the trend from 2019 to 2021.
Table 3. Traffic congestion.
Table 3. Traffic congestion.
Morning Rush Evening Rush Congestion Average
Year 2019 Morning
67% Rush Evening
67%Rush Congestion
ExtremelyAverage
high
Year 2020
Year 2019 67%
45% 67%
41% Extremely
Low high
Sustainability 2022, 14, 8809 11 of 16

Table 3. Traffic congestion.

Morning Rush Evening Rush Congestion Average

Year 2019 67% 67% Extremely high
Year 2020 45% 41% Low
Year 2021 37% 36% Low
Decrease from 2019 to 2020 22% 26% Extremely Low
Decrease from 2020 to 2021 8% 5% Moderate low
Decrease from 2019 to 2021 30% 31% Extremely Low

The worst day of the week with the worst traffic congestion was Mondays from 7 a.m.
to 8 a.m. Travelling before 7 a.m. on Monday could save someone up to 4 h per year for a
30 min commute. During 2019, 20+ min per 30 min trip was spent on the morning rush
hour and 20+ min per 30 min trip was spent during the evening rush. In 2020, 13+ min
per 30 min trip was spent on the morning rush hour and 12+ min per 30 min trip was
spent during the evening rush. In 2021, 11+ min per 30 min trip was spent on the morning
rush hour and 11+ min per 30 min trip was spent during the evening rush. A total of
153 h was lost in traffic congestion in 2019, which is equivalent to 6 days 9 h; 98 h was
lost in traffic congestion in 2020, which totals 4 days 2 h. A total of 84 h was lost in traffic
congestion in 2021 which is equivalent to 3 days 12 h. The decrease between 2021 and 2019
of hours spent in traffic congestion was 2 days 21 h, while the decrease between 2021 and
2020 of hours spent in traffic congestion was 14 h. The decrease between 2020 and 2019
of hours spent in traffic congestion was 2 days 7 h. Consequently, from Table 3, it can be
noted that the morning rush had the highest traffic congestion. Furthermore, the overall
traffic congestion for 2021 seemed to be lower than the traffic congestion in 2020. The
reason for this was because the first 3 months of the year 2020 had high traffic congestion
and the COVID-19 restrictions were not yet enforced, hence the slightly higher overall
congestion in 2020. However, the analysis conducted by the authors revealed that the year
2020 had the lowest traffic congestion overall due to the mentioned reason. Using the
spatiotemporal region of Johannesburg, the following was used to express variations in
low traffic congestion: COVID-19 restrictions and COVID-19 severe restrictions with low
traffic congestion. The COVID-19 pandemic’s impact on traffic congestion in the city of
Johannesburg is demonstrated further in Table 4 below.

Table 4. COVID-19 effect on traffic congestion 2020.

Day 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31
Jan
Feb
Mar
Apr
May
Jun
July
Aug
Sep
Oct
Nov
Dec
COVID-19 COVID-19 severe restriction with low traffic
Day with low traffic congestion
restriction congestion

Table 4 is a breakdown of the traffic congestion when the COVID-19 pandemic was
strongly affecting the city of Johannesburg. It shows the traffic congestion analysis from the
Sustainability 2022, 14, 8809 12 of 16

time when the pandemic regulatory system was enforced up to the last day of 2020. From
the 27th of March 2020 to the first of May 2020, the travelling restrictions were very severe,
and the traffic congestion was very low at 2% during the entire month of April. During
the period of May, June, July, and August, it can be noted that the restrictions were not as
severe, though travelling was still restricted and the percentage of days with low traffic
congestion overall was precisely 12%, as indicated by the color green. Further, the grey
area indicates that as there were still COVID-19 restrictions, and these days had a slightly
higher traffic congestion compared to other days when the restrictions were still active;
however, that period experienced precisely 19% of traffic congestion. This was due to the
opening of some of the city’s activities to boost the economy. Traffic congestion volume
decreased immensely across the city of Johannesburg, and COVID-19 brought about a huge
decrease in traffic congestion in many cities globally including the city of Johannesburg.
Accordingly, the city had 130 days with very much less congestion in 2020.

3.4. Toward Sustainable Urban Mobility

Sustainable cities require smart policy practices, smart governance, and smart mobility.
Placing urban mobility as focus for improving quality of life requires three key phases
to be achieved in developing cities. The first phase is identification of points of interest
in the city which are hot spots for congestion. This is essential as it navigates us to
catchment areas or areas of interest. The analysis of points of interest reveals areas with
mixed land uses that have the highest levels of congestion. Harris et al. [8] outlined that
investment in improving road infrastructure at such hot spots has the potential to also
improve overall productivity. The second phase is improving monitoring systems. The
literature reveals that investment in location-based services (LBSs) is essential for prediction
planning and improving responsiveness to traffic challenges. A lack of coordination
and structured information management has significant impacts on the effectiveness of
monitoring systems. As such, LBSs avail data-driven assurance and confidence for traffic
management. As outlined in the earlier sections, TomTom data captures movement trends
with geo-referencing data [43]. With the growth in big data and Internet of Things, LBSs
can provide cities with real-time data that can be used for monitoring and mitigating
traffic challenges. The last phase is for cities to tap into machine learning and deep
learning. This can bring opportunities of autonomous traffic management either through
supervised or unsupervised learning of data obtained through LBSs [44,45]. Globally digital
innovation has proven to be at the heart of transformation and increasing opportunities.
It is a driving force for innovative, sustainable growth and facilitating change. However,
for many developing cities, there is very little investment in machine learning and deep
learning [46,47]. As such, the shift toward digital transformation should be initiated through
raising awareness of the comprehensive capability of machine learning and deep learning.

4. Discussions, Implications, Reflections, and Generalizability of Findings

This work has several implications with regard to new knowledge generation for
academic purposes and policy formulation and implementation in the local, provincial,
and national spheres. Efficient public transportation systems provide economic and social
benefits by improving market accessibility, promoting labor mobility, and connecting
communities. However, key factors in the long-term reduction of congestion are dependent
on the willingness to adopt new technologies [48,49]. It must be stressed that urban mobility
is essential for the maintenance and well-being of the society and economy. Furthermore,
the benefits of reduced congestion levels spread far beyond the transportation sectors since
it is essential for economic development and improving the quality of life of people. From
the findings, it is clear that limiting movement may yield desired results of reducing overall
congestion. However, there is a need to develop plans for the post-COVID-19 period, as
the lockdown regulations cannot be sustained indefinitely [50,51]. It is thus important
that more attention be paid to soft and functional factors that can attract commuters to
utilize public transportation regularly. This would lead to fewer vehicles being on the roads
Sustainability 2022, 14, 8809 13 of 16

while also allowing commuters to traverse between their homes, workplaces, and places
of interest.
This statistical analysis of traffic congestion provided an insight into traffic congestion
between the years 2019 and 2021. It was noted that prior to the COVID-19 pandemic, the
level of traffic congestion was very high during rush hours, specifically during the morning
and evening rushes on weekdays. The overall traffic congestion of 2019 was 30%, while in
2020 it was 21%; thus, the decrease from 2019 to 2020 was 9%. The overall traffic congestion
of 2020 was 21%, with 2021 at 19%, and thus, the decrease from 2020 to 2021 was 2%;
furthermore, the decrease from 2019 to 2021 was 11%. The time spent in traffic congestion
per day during the morning, afternoon, and the evening rushes had a negative impact on
the daily productivity for the economy. During 2020, when the city of Johannesburg was
challenged with COVID-19, the volume of the congestion drastically decreased to a very
low rate throughout the year. This was due to the strict regulations enforced to control
travelling for the management of the COVID-19 pandemic. This clearly indicates that traffic
congestion can be controlled by the enforcement of strict regulatory systems for travelling
and the use of better technological systems. Globally, there are improvements in reducing
traffic congestion [52]. This is especially true in developed countries, where adaptive traffic
signals, real-time traffic monitoring, smart corridors, and pedestrian tracking systems with
advanced V2I technologies are deployed, which has drastically reduced congestion [53].
The developing cities such as Johannesburg can look into initiating such technological
systems so as to bring about a positive impact on the challenged roads. As such, this
growing interest in the development of resilient cities is bringing the issue of sustainable
mobility to the forefront of public discourse. This is because the provision of sound and
adequate urban transportation is of paramount importance to achieve such sustainability
and economic growth. On a macro level, ecological transportation management facilitates
efforts of ecosystem restoration through supporting the area’s physical and social growth.
The role of ecosystem restoration is thus critical in managing urban mobility systems.

5. Conclusions
This paper presented the potential of sustainable and resilient transportation manage-
ment by focusing on two phases, namely, the first phase of identification of points of interest
in the city which are hot spots for congestion and the second phase of outlining. This was
undertaken through an assessment of congestion levels across three critical periods for the
city of Johannesburg. With more and more assessments of urban transportation, transport
management authorities can use these trends to guide future policy incentives to enhance
the operational performance of mobility systems. In 2020, the reduced congestion levels
during the national lockdown revealed that if mobility in the city is restricted, this leads to
better air quality. However, introducing a lockdown is not an advisable option to reduce
vehicular emissions as this negatively impacts economic activities. The results demon-
strated that mining and analyzing traffic data have a significant role for future mobility
planning in both the developed and developing worlds, and more generally, for improving
the quality of commuting trips in the city.
The results showed a continuous decrease from 2020 to 2021 in traffic congestion in the
city due to the COVID-19 pandemic. Traffic congestion is a global challenge and is not easy
to be prevented or controlled. Transportation authorities with various stakeholders in the
city of Johannesburg work toward implementing strategies to reduce congestion so as to
improve the flow of movement on the roads. However, the effort is still challenging as high
traffic congestion volumes are still evident because Johannesburg is the city with the third
highest traffic congestion on the African continent and second in the Republic of South
Africa after the city of Cape Town. This calls for urgent solutions; otherwise, when the
country transitions to a developed state, there may be no movement on the roads during
rush hours. Lastly, the paper calls for investment in autonomous traffic management in
developing cities. This may lead to less hours lost in congestion. Machine learning and
Sustainability 2022, 14, 8809 14 of 16

deep learning have the potential for ensuring autonomous decision making, which can
lead to better regulatory systems on the roads.

Author Contributions: Conceptualization, T.M. and S.M.; methodology, T.M.; software, O.-O.A.;
validation, T.G., O.-O.A. and I.M.; formal analysis, T.M. and S.M.; investigation, O.-O.A.; resources,
T.G. and I.M.; data curation, O.-O.A.; writing—original draft preparation, T.M., S.M. and O.-O.A.;
writing—review and editing, T.G. and I.M.; visualization, T.M. and S.M.; supervision, T.G. and I.M.;
project administration, O.-O.A.; funding acquisition, T.G. and I.M. All authors have read and agreed
to the published version of the manuscript.
Funding: This research was funded by Centre of Applied Research and Innovation in the Built
Environment (CARINBE).
Institutional Review Board Statement: Not applicable.
Informed Consent Statement: Not applicable.
Data Availability Statement: All the supporting data are available in the manuscript.
Acknowledgments: The authors acknowledge the co-operation and data availed by the Johannesburg
Metropolitan City Municipality.
Conflicts of Interest: The authors declare no conflict of interest.

References
1. Kahachi, H.A.H. Analyzing Public Transportation Plans to Addressing Traffic Congestions in Developing Countries. Iraqi J.
Archit. Plan. 2019, 18, 44–57. [CrossRef]
2. Venkatesham, S.; Kumar, S. Analysis of the Measurable Results of Novel Structure Construction for Road Traffic. In Proceedings
of the 2021 7th International Conference on Advanced Computing and Communication Systems (ICACCS), Coimbatore, India,
19–20 March 2021; pp. 1490–1494. [CrossRef]
3. Sun, L.; Song, R. Improving Efficiency in Congested Traffic Networks: Pareto-Improving Reservations through Agent-Based
Timetabling. Sustainability 2022, 14, 2211. [CrossRef]
4. Jain, V.; Sharma, A.; Subramanian, L. Road traffic congestion in the developing world. In Proceedings of the 2nd ACM Symposium
on Computing for Development, Atlanta, GA, USA, 11–12 March 2012; pp. 1–10.
5. Chen, L.; Zheng, L.; Yang, J.; Xia, D.; Liu, W. Short-term traffic flow prediction: From the perspective of traffic flow decomposition.
Neurocomputing 2020, 413, 444–456. [CrossRef]
6. Khan, N.U.; Shah, M.A.; Maple, C.; Ahmed, E.; Asghar, N. Traffic Flow Prediction: An Intelligent Scheme for Forecasting Traffic
Flow Using Air Pollution Data in Smart Cities with Bagging Ensemble. Sustainability 2022, 14, 4164. [CrossRef]
7. Deng, S.; Jia, S.; Chen, J. Exploring spatial–temporal relations via deep convolutional neural networks for traffic flow prediction
with incomplete data. Appl. Soft Comput. 2019, 78, 712–721. [CrossRef]
8. Harris, N.; Shealy, T.; Klotz, L. Choice architecture as a way to encourage a whole systems design perspective for more sustainable
infrastructure. Sustainability 2016, 9, 54. [CrossRef]
9. Kiunsi, R.B. A Review of Traffic Congestion in Dar es Salaam City from the Physical Planning Perspective. J. Sustain. Dev. 2013, 6,
94–103. [CrossRef]
10. Olayode, I.O.; Tartibu, L.K.; Okwu, M.O.; Uchechi, U.F. Intelligent transportation systems, un-signalized road intersections and
traffic congestion in Johannesburg: A systematic review. Procedia CIRP 2020, 91, 844–850. [CrossRef]
11. Afrin, T.; Yodo, N. A survey of road traffic congestion measures towards a sustainable and resilient transportation system.
Sustainability 2020, 12, 4660. [CrossRef]
12. Kumarage, A.S. Urban traffic congestion: The problem and solutions. Asian Econ. Rev. 2004, 2, 10–19.
13. Gubins, S.; Verhoef, E.T. Dynamic bottleneck congestion and residential land use in the monocentric city. J. Urban Econ. 2014, 80,
51–61. [CrossRef]
14. Ogunbodede, E.F. Assessment of traffic congestions in Akure (Nigeria) using GIS approach: Lessons and challenges for urban
sustenance. In Proceedings of the Conference on Whole Life Urban Sustainability, Glasgow, Scotland, 27–29 June 2007; pp. 1–25.
15. Wang, C. The Relationship between Traffic Congestion and Road Accidents: An Econometric Approach Using GIS. Ph.D. Thesis,
Loughborough University, Loughborough, UK, 2010.
16. Petrovska, N.; Stevanovic, A. Traffic congestion analysis visualisation tool. In Proceedings of the IEEE 18th International
Conference on Intelligent Transportation Systems, Gran Canaria, Spain, 15–18 September 2015; pp. 1489–1494.
17. Jain, N.K.; Saini, R.K.; Mittal, P. A review on traffic monitoring system techniques. In Soft Computing: Theories and Applications;
Springer: Cham, Switzerland, 2019; pp. 569–577.
18. Gerlough, D.L.; Huber, M.J. Traffic Flow Theory; National Research Council: Washington, DC, USA, 1976.
Sustainability 2022, 14, 8809 15 of 16

19. Christidis, P.; Rivas, J.N.I. Measuring Road Congestion; Institute for Prospective Technological Studies (IPTS), European Commission
Joint Research Centre: Brussels, Belgium, 2012; Available online: http://ipts.jrc.ec.europa.eu/publications/pub.cfm (accessed on
12 March 2022).
20. Yue, J.S.; Mandayam, C.V.; Merugu, D.; Abadi, H.K.; Prabhakar, B. Reducing Road Congestion through Incentives: A Case Study.
In Proceedings of the Transportation Research Board 94th Annual Meeting, Washington, DC, USA, 11–15 January 2015.
21. Yin, Y.; Ieda, H. Assessing performance reliability of road networks under nonrecurrent congestion. Transp. Res. Rec. 2001, 1771,
148–155. [CrossRef]
22. Zhang, T.; Sun, L.; Yao, L.; Rong, J. Impact analysis of land use on traffic congestion using real-time traffic and POI. J. Adv. Transp.
2017, 2017, 7164790. [CrossRef]
23. Renne, J.L. From transit-adjacent to transit-oriented development. Local Environ. 2009, 14, 1–15. [CrossRef]
24. Harrison, P.; Rubin, M.; Appelbaum, A.; Dittgen, R. Corridors of freedom: Analyzing Johannesburg’s ambitious inclusionary
transit-oriented development. J. Plan. Educ. Res. 2019, 39, 456–468. [CrossRef]
25. Ndebele, R.; Aigbavboa, C.; Ogra, A. Urban transport infrastructure development in African Cities: Challenges and opportunities.
In Proceedings of the International Conference on Industrial Engineering and Operations Management, Johannesburg, South
Africa, 29 October–1 November 2018; p. 833.
26. Pieterse, E.; Owens, K. Johannesburg: Confronting Spatial Inequality; World Resources Institute: Washington, DC, USA, 2018.
27. Afrin, T.; Yodo, N. A probabilistic estimation of traffic congestion using Bayesian network. Meas. J. Int. Meas. Confed. 2021,
174, 109051. [CrossRef]
28. Rahman, M.M.; Paul, K.C.; Hossain, M.A.; Ali, G.G.; Nawaz, M.; Rahman, M.S.; Thill, J. Machine Learning on the COVID-19
Pandemic, Human Mobility and Air Quality: A Review. IEEE Access 2021, 9, 72420–72450. [CrossRef]
29. Wijnands, J.S.; Zhao, H.; Nice, K.A.; Thompson, J.; Scully, K.; Guo, J.; Stevenson, M. Identifying safe intersection design through
unsupervised feature extraction from satellite imagery. Comput. Aided Civ. Infrastruct. Eng. 2021, 36, 346–361. [CrossRef]
30. Mesgarpour, M.; Abad, J.M.N.; Alizadeh, R.; Wongwises, S.; Doranehgard, M.H.; Jowkar, S.; Karimi, N. Predicting the effects
of environmental parameters on the spatio-temporal distribution of the droplets carrying coronavirus in public transport—A
machine learning approach. Chem. Eng. J. 2022, 430, 132761. [CrossRef]
31. Benning, O.; Calles, J.; Kantarci, B.; Khan, S. Transit Networks, Social Contacts, and Open Data Meet Public Transportation Plans
for Post-COVID-19: A Canadian Case Study. IEEE Eng. Manag. Rev. 2021, 49, 30–41. [CrossRef]
32. Nie, Q.; Qian, X.; Guo, S.; Jones, S.; Doustmohammadi, M.; Anderson, M.D. Impact of COVID-19 on paratransit operators and
riders: A case study of central Alabama. Transp. Res. Part A Policy Pract. 2022, 161, 48–67. [CrossRef]
33. Attard, M. Active travel and sustainable transport. Commun. Transp. Res. 2022, 2, 100059. [CrossRef]
34. Yan, R.; Wang, S.; Zhen, L.; Laporte, G. Emerging approaches applied to maritime transport research: Past and future. Commun.
Transp. Res. 2021, 1, 100011. [CrossRef]
35. Kakderi, C.; Oikonomaki, E.; Papadaki, I. Smart and Resilient Urban Futures for Sustainability in the Post COVID-19 Era: A
Review of Policy Responses on Urban Mobility. Sustainability 2021, 13, 6486. [CrossRef]
36. Arai, I.; El-Tawab, S.; Salman, A.; Elnoshokaty, A. The Effect of COVID-19 on the Transit System in Two Regions: Japan and USA.
In Proceedings of the 2021 IEEE Global Conference on Artificial Intelligence and Internet of Things (GCAIoT), Dubai, United
Arab Emirates, 12–16 December 2021; pp. 51–56.
37. Moyo, T.; Musakwa, W. Using crowdsourced data (Twitter & Facebook) to delineate the origin and destination of commuters of
the Gautrain public transit system in South Africa. ISPRS Ann. Photogramm. 2016, 3, 143–150.
38. Nouvellet, P.; Bhatia, S.; Cori, A.; Ainslie, K.E.; Baguelin, M.; Bhatt, S.; Boonyasiri, A.; Brazeau, N.F.; Cattarino, L.; Cooper, L.V.
Reduction in mobility and COVID-19 transmission. Nat. Commun. 2021, 12, 1090. [CrossRef]
39. Nižetić, S. Impact of coronavirus (COVID-19) pandemic on air transport mobility, energy, and environment: A case study. Int. J.
Energy Res. 2020, 44, 10953–10961. [CrossRef]
40. Tomtom. Traffic-Index; Tomtom: Amsterdam, The Netherlands, 2022.
41. Kibangou, A.Y.; Moyo, T.; Musakwa, W. Dynamic Linear Model for Urban Essential Traffic Congestion and Emissions. IFAC
PapersOnLine 2021, 54, 151–156. [CrossRef]
42. Wan, C.; Yang, Z.; Zhang, D.; Yan, X.; Fan, S. Resilience in transportation systems: A systematic review and future directions.
Transp. Rev. 2018, 38, 479–498. [CrossRef]
43. TomTom/EC Team on Data for Road Safety. 2020. Available online: just-auto.com (accessed on 20 November 2021).
44. Wang, C.; Li, X.; Chen, P.; Xie, Y.; Liu, W. Spatial pattern and developing mechanism of railway geo-systems based on track gauge:
A case study of Eurasia. J. Geogr. Sci. 2020, 30, 1283–1306. [CrossRef]
45. Harrou, F.; Zeroual, A.; Sun, Y. Traffic congestion monitoring using an improved kNN strategy. Meas. J. Int. Meas. Confed. 2020,
156, 107534. [CrossRef]
46. Neu, D.A.; Lahann, J.; Fettke, P. A systematic literature review on state-of-the-art deep learning methods for process prediction.
Artif. Intell. Rev. 2021, 55, 801–827. [CrossRef]
47. Jiang, F.; Ma, X.; Zhang, Y.; Wang, L.; Cao, W.; Li, J.; Tong, J. A new form of deep learning in smart logistics with IoT environment.
J. Supercomput. 2022, 78, 11873–11894. [CrossRef]
48. Akhtar, M.; Moridpour, S. A review of traffic congestion prediction using artificial intelligence. J. Adv. Transp. 2021, 2021, 8878011.
[CrossRef]
Sustainability 2022, 14, 8809 16 of 16

49. Jiang, H.; Li, Q.; Jiang, Y.; Shen, G.; Sinnott, R.; Tian, C.; Xu, M. When machine learning meets congestion control: A survey and
comparison. Comput. Netw. 2021, 192, 108033. [CrossRef]
50. Currie, G.; Jain, T.; Aston, L. Evidence of a post-COVID change in travel behaviour–Self-reported expectations of commuting in
Melbourne. Transp. Res. Part A Policy Pract. 2021, 153, 218–234. [CrossRef]
51. Xu, P.; Li, W.; Hu, X.; Wu, H.; Li, J. Spatiotemporal analysis of urban road congestion during and post COVID-19 pandemic in
Shanghai, China. Transp. Res. Interdiscip. Perspect. 2022, 13, 100555. [CrossRef]
52. Farda, M.; Balijepalli, C. Exploring the effectiveness of demand management policy in reducing traffic congestion and environ-
mental pollution: Car-free day and odd-even plate measures for Bandung city in Indonesia. Case Stud. Transp. Policy 2018, 6,
577–590. [CrossRef]
53. Aleksander, R.; Paweł, C. Recent advances in traffic optimisation: Systematic literature review of modern models, methods and
algorithms. IET Intell. Transp. Syst. 2020, 14, 1740–1758. [CrossRef]