Assessing Spatial Variations of Traffic Congestion Using Traffic
Assessing Spatial Variations of Traffic Congestion Using Traffic
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
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.
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.
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
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
(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.
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.
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.
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.
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]