Smart Parking Tools Suitability for Open Parking Lots: A Review
Vijay Paidi, Hasan Fleyeh, Johan Håkansson and Roger G. Nyberg
Dalarna University, School of Technology and Business Studies, Borlänge, Sweden
Keywords: Decision Support System, Sensors, Technologies, Applications.
Abstract: Parking a vehicle in traffic dense environments is a common issue in many parts of the world which often
leads to congestion and environmental pollution. Lack of guidance information to vacant parking spaces is
one of the reasons for inefficient parking behaviour. Smart parking sensors and technologies facilitate
guidance of drivers to free parking spaces thereby improving parking efficiency. Currently, no such sensors
or technologies are in use for the common open parking lot. This paper reviews the literature on the usage of
smart parking sensors, technologies, applications and evaluate their suitability to open parking lots. Suitability
was made in terms of expenditure and reliability. Magnetometers, ultrasonic sensors and machine vision were
few of the widely used sensors and technologies used in closed parking lots. However, this paper suggests a
combination of machine vision, fuzzy logic or multi-agent systems suitable for open parking lots due to less
expenditure and resistance to varied environmental conditions. No application provided real time parking
occupancy information of open parking lots, which is a necessity to guide them along the shortest route to
free space. To develop smart parking applications for open parking lots, further research is needed in the fields
of deep learning.
1 INTRODUCTION
In densely populated areas as in cities, the availability
of parking spaces is often less than the availability of
vehicles which leads to a shortage of parking space
(Chinrungrueng et al., 2007). A parking space is
defined as a location designated for parking which
can either be paved or unpaved. A parking lot is
referred as a group of parking spaces and parking
refers to a vehicle finding and occupying a vacant
parking space.
Studies showed that in worldwide traffic dense
environment, 30 to 50 percent of the drivers look for
free parking space (Polak and Vythoulkas, 1993,
Boltze and Puzicha, 1995, White, 2007, Gallivan,
2011). Based on previous studies drivers use between
3.5 and 14 minutes to find a parking space (Shoup,
2006a, Polycarpou et al., 2013). Examples of
consequences are; frustration of drivers, accidents,
lost business opportunities, congestion, and increased
air pollution. According to (Shoup, 2006b), studies
focus on areas where there is expected increase in
traffic and scarcity of parking spaces. Congestion can
be observed when traffic density is increased, which
is observed during peak traffic periods (Geroliminis
and Daganzo, 2008, Geroliminis and Sun, 2011). In
terms of air pollution, acceleration and cruising of a
vehicle lead to high amount of emissions compared to
idle or deceleration. A daily average of 100 inefficient
cruising’s in a parking lot would generate 24 tons of
CO
2
which is approximately equivalent to three times
of CO
2
per capita in Sweden (Frey et al., 2012, Jan
Minx, 2008).
In order to reduce parking related congestion and
air pollution, public transportations such as bus,
metro, etc., can be utilized. However, private vehicles
are still used for convenience. One way to address the
carbon emission issue is to replace all fuel
consumption vehicles with electric ones. Since, the
average age of a car in Europe or United States is
approximately 11 years (U.S)(Statista, 2017,
European Automobile Manufactures Association,
2017), it can be expected to take quite a long time
before electric vehicles replace all fuel consumption
vehicles. In spite of using electric vehicles, cruising
of a vehicle to find a vacant parking space would still
be time consuming leading to congestion. A solution
to that problem might be driverless cars. Driverless
vehicle is one of the upcoming technologies currently
being developed. Few companies such as Google
have already developed prototype vehicles which
support driverless mode in specific conditions
(Thierer and Hagemann, 2014). Vehicle to Vehicle
(V2V) technology could be implemented on all
600
Paidi, V., Fleyeh, H., Håkansson, J. and G. Nyberg, R.
Smart Parking Tools Suitability for Open Parking Lots: A Review.
DOI: 10.5220/0006812006000609
In Proceedings of the 4th International Conference on Vehicle Technology and Intelligent Transport Systems (VEHITS 2018), pages 600-609
ISBN: 978-989-758-293-6
Copyright
c
2019 by SCITEPRESS – Science and Technology Publications, Lda. All rights reserved
vehicles in order to improve the safety of all
driverless vehicles (Harding et al., 2014, Guler et al.,
2014). An automated driverless vehicle would depend
on a number of technologies such as GPS, sensors,
V2V, etc. and it would take few decades to replace
existing vehicles. Yet another, not that far-fetched,
solution might be a decision support system that uses
smart parking sensor and technologies to detect
parking occupancy information which facilitate the
drivers to make informed decision of where to go and
park their car (Yang et al., 2003). Parking occupancy
information can be gathered using sensors and
technologies such as; ultrasonic sensors,
magnetometers and multi-agent systems. Parking a
vehicle with the use of assisted applications or
technologies is known as smart parking. Smart
parking applications are available online which
facilitate in improving parking efficiency by
providing navigational directions to a reserved empty
parking space (Kotb et al., 2016).
There has already been considerable amount of
research conducted to improve parking efficiency at
closed parking lots which are paid parking lots and
support reservation of parking spaces. There are
applications available online to provide decision
support through smart parking services in closed
parking lots. A parking space can be reserved by
paying it through mobile or web applications. Once
the reservation is made, the vehicle can enter the
closed parking lot and occupy a vacant or allocated
parking space. When reservation is made using
applications, the user would be given a reference code
which can be used for authorization. However
literature is very scarce with regard to another and
very common type of parking lot, the open parking
lots which does not support reservation, freely
available for a limited amount of time and are
commonly placed outdoors occupying large amount
of space. Therefore, there still exists a research gap in
order to improve parking efficiency at an open
parking lot. It is assumed that the driver behaviour
could be efficient if improved decision support to
drivers are offered. The areal size of the often very
large open parking lots could be reduced by
improving parking efficiency along with reduced
congestion and CO
2
emissions.
The aim of this paper is to review smart parking
sensors, technologies and applications suitability at
open parking lots. Challenges in an open parking lot
are; freely available where reservation is not possible
and they are subjected to environmental conditions
when placed outdoors. Therefore, suitability of
sensors and technologies was made by compatibility
and expenditure parameters.
The remaining part of this paper is organised in
the following way. Section 2, describes the methods
used to conduct the review of existing smart parking
sensors, technologies and applications. Section 3
describes and analyses smart parking sensors,
technologies and applications while Section 4
discusses about their suitability in open parking lots.
Finally, the paper is ended with conclusion in Section
5.
2 MATERIALS AND METHODS
In this paper, literature search was carried out to
identify articles using online databases. Two search
procedures have been performed. First search is to
find the smart parking sensors, technologies and a
second one with focus on smart parking applications.
An initial search was made in Google scholar and
based on the result in that search, a refined search
have been made in databases such as; ACM, IEEE,
Springer, Elsevier, Taylor & Francis, Google Books.
Keywords such as; smart parking sensors,
technologies, intelligent parking systems, wireless
parking systems, online parking, parking efficiency,
outdoor parking systems, etc., were used for literature
search. The initial search and review of articles
identified smart parking sensors and technologies, see
Section 3. Studies referring to parking issues or
development and testing of smart parking sensors,
technologies were selected in the initial search.
Further analysis of information collected from the
articles identified the suitability of smart parking
sensors, technologies for open parking lots. Since the
paper focuses on sensors, technologies and interfaces
which are used to collect and display real time
parking occupancy information, frameworks or
architecture of smart parking systems are not
reviewed in this paper. Parking meters which are used
for payments are placed either at a parking lot or
beside parking spaces are not referred in this paper.
From the literature search, previous literature reviews
on sensors and technologies are presented in Table 1,
since they cover the available sensors and
technologies described in the literature. However, it
should be mentioned that these reviews do not
emphasize on parking issues related to open parking
lots. Keywords such as; review of smart parking,
summary of intelligent parking, efficient parking
systems overview, etc. were used to identify review
articles using mentioned online databases. Surveys or
reviews referring to usage of smart parking sensors
and technologies were selected.
Smart Parking Tools Suitability for Open Parking Lots: A Review
601
Table 1: Literature reviews on smart parking sensors and technologies.
Smart Parking Tools
(Mimbel
a and Klein,
2000)
(Idris
et al.,
2009a)
(Revathi
and
Dhulipala,
2012)
(Mahm
ud et al.,
2013)
(Fraife
r and
Fernström,
2016)
(Hassou
ne et al.,
2016)
(Enríquez
et al., 2017)
Infrared sensors
Ultrasonic sensors
Inductive loop detectors
Parking guidance systems
Radio frequency tags
Magnetometer
Microwave radar
GPS
Machine vision
Vehicular Ad hoc networks
(VANET)
Multi-agent systems
Neural network
Fuzzy logic
Online search engine such as Google and mobile
application stores are used to find existing smart
parking applications. Keywords such as parking
applications, smart parking applications, mobile
parking were used to identify applications. In mobile
application store, keywords such as; parking, e-
parking, smart parking, etc. were used. There are
hundreds of parking applications available in mobile
application store. However, in this paper parking
applications which provide real time parking
occupancy information and navigational directions to
a reserved parking space are selected. All the referred
applications support reservation which is possible
only in closed parking lots.
3 SMART PARKING TOOLS
Smart parking tools consists of sensors, technologies
and applications which are used to identify parking
occupancy information and facilitate to improve
parking efficiency. As shown in Table 1, (Revathi and
Dhulipala, 2012, Mimbela and Klein, 2000, Hassoune
et al., 2016) are reviews about various types of smart
parking sensors, technologies along with their uses
while (Fraifer and Fernström, 2016), identifies
research gap in designing smart parking system for
stakeholders. Furthermore, (Mahmud et al., 2013),
reviews about intelligent parking technologies and
their economic analysis and (Enríquez et al., 2017,
Idris et al., 2009a) reviews the advantages and
drawbacks of sensors and technologies. As shown in
Table 1, machine vision using visual camera is
referred by all the review articles. Sensors like
ultrasonic and magnetometers are widely reviewed
and tested which are used in various smart parking
applications. Vehicular Ad hoc networks and fuzzy
logic were not reviewed prior to 2013. Detailed
description of each sensors and technologies can be
found in this section along with emphasis on
expenditure.
3.1 Smart Parking Sensors
There are various sensors which facilitate in detecting
parking occupancy information and these are
mentioned in the following sections. Sensors are one
of the common tools which were widely tested in
several previous literatures. Descriptions of these
sensors are mentioned in the following sections.
3.1.1 Passive or Active Infrared Sensor
Passive sensors detect changes in energy and when a
vehicle occupies a parking space, these sensors
identify the change in energy and detects occupancy
(Shaheen, 2005, Mouskos et al., 2007). Passive
sensor observes a change in energy when a vehicle is
placed or a person standing above the sensor. Based
RESIST 2018 - Special Session on Resilient Smart city Transportation
602
on the amount of energy change, it can be used to
isolate outliers. Active sensors would emit infrared
energy and detect any object or vehicle by the amount
of energy reflected. However, both passive and active
infrared sensors are sensitive to environment and they
would not be accurate when there is snow or rain.
Passive infrared sensors should be placed under the
ground or on the ceiling while active sensors are
normally mounted above a parking spot. Both the
sensors require high investment for procurement and
maintenance. These sensors would be suitable for
closed parking lots which are inside buildings and are
not suitable for outdoor open parking lots.
3.1.2 Ultrasonic Sensor
These sensors would emit sound waves between 25 to
50 kHz and detect objects based on reflected energy.
They are usually mounted on ceiling and are sensitive
to environmental changes such as rain and snow
(Kianpisheh et al., 2012a). Therefore, they are
suitable for indoor parking lots rather than open
parking lots. Based on the distance at which waves
are reflected it can distinguish between a vehicle and
a person. In order to get parking occupancy status
these sensors should be placed on top of every
parking space. These sensors would be available for
low cost but installation and maintenance of multiple
sensors and connecting them to a grid would be
expensive in the long run. Wireless ultrasonic sensors
are also used to gather parking occupancy
information. They are connected using wireless
sensor networks such as ZigBee protocol or other
similar networks (Idris et al., 2009b). However,
wireless sensors involves periodic maintenance costs.
In another study, ultrasonic sensors are used on a
drive-by vehicle and parking occupancy information
is collected at regular periods (Mathur et al., 2010).
Real time parking occupancy information cannot be
attained using drive-by vehicle.
3.1.3 Inductive Loop Detectors
These detectors are installed using underground
wiring system and they use principles of
electromagnetism to detect the presence of a vehicle
(Shaheen, 2005). They are commonly used at the
entrance and exit to get the count of vehicles which
can be used to know availability of parking spaces.
These detectors are expensive to install and maintain
(Mouskos et al., 2007) and they can be used in indoor
and outdoor parking lots to get the count of available
parking spaces. Accurate count of vehicles would be
provided using these detectors and these are in use at
multiple commercial parking lots.
3.1.4 Parking Guidance Systems
Parking guidance systems is another smart parking
system which provides information about number of
parking spaces available on display screens and these
are usually placed near the parking lots as the driver
can see and decide the parking space to
occupy.(Waterson et al., 2001) (Idris et al., 2009b)
Inductive loop detectors or visual camera can be used
at the entrance and exit of a parking lot to know the
count of the vehicles in a parking lot which would be
displayed on the screens. However, they do not guide
the driver to a particular parking space which is found
empty. Therefore, there is every possibility that the
driver would cruise for several minutes before finding
an empty space to occupy. The driver can make a
decision about the parking lot only after viewing the
display screens (Kianpisheh et al., 2012b). Since
sensors or visual cameras would be deployed only to
get the count of vehicles the expenditure for
installation and maintenance would be minimal
making them suitable for open parking lots.
3.1.5 Radio-Frequency Identification
(RFID)
Radio frequency tags are used to identify vehicle.
Each vehicle will be given a radio frequency tag for
identification. A transceiver and antenna would be
installed at the entrance of a parking lot to identify the
tag and allow the vehicle to occupy a parking lot
(Rahman et al., 2009). These are suited for closed and
indoor parking lots which are controlled. It is not
suitable for an open parking lot as they are freely
available. RFID is used to authorize movement of
vehicles at a parking lot. However, it does not
providing individual parking occupancy status nor
facilitates the driver in finding a vacant parking space.
3.1.6 Magnetometer
These sensors detect the presence of vehicle by
detecting the change in electromagnetic field. They
need to be in close proximity to the vehicle, therefore,
they are placed beneath the surface. They are not
sensitive to the environment (Shaheen, 2005). These
are suitable for both open and closed parking lots.
There are wireless sensors with a battery life time of
few years which can be used to detect real time
parking occupancy information. The sensors should
be placed under every parking space to know the
occupancy of parking spaces. However, it is
expensive to install and maintain these sensors on a
large scale.
Smart Parking Tools Suitability for Open Parking Lots: A Review
603
3.1.7 Microwave Radar
A microwave radar transmits microwave beam and
based on the reflected signal it estimates the velocity
of the moving target. However, it does not detect
stationary objects. In order to eliminate this
restriction, dual microwave Doppler radar can be
used to detect both moving and stationary vehicles
(Bao et al., 2017). These can be mounted or placed
beneath the surface for vehicle detection. These
radars are not sensitive to environment and can be
used in open and closed parking lots. They should be
placed in every parking space to detect parking
occupancy status making them expensive to install
and maintain these microwave radars on a large scale.
3.2 Smart Parking Technologies
Sensor technologies are tools which facilitate the
driver in occupying a vacant parking space and
descriptions of these technologies can be found in the
following sections below.
3.2.1 Global Positioning System (GPS)
GPS based navigational directions are provided to the
driver for occupying a vacant parking space. GPS will
facilitate in finding the shortest/optimal route from
the current location. However, GPS alone cannot
gather occupancy information of parking spaces. In
one study occupancy of parking spaces is estimated
using historical occupancy information and
navigational directions are provided using GPS to the
estimated parking space (Pullola et al., 2007). The
accuracy of the GPS with a single frequency receiver
is less than or equal to 7.8 meters. If a dual frequency
receiver is used, the accuracy is less than 0.71 meters.
A normal parking space would be between 2.3 to 2.7
meters and most of the smartphones are provided with
single frequency receiver which have higher error
compared to dual frequency receivers. Dual
frequency receivers are usually used for military
products for greater accuracy. A GPS is also
perceptible to errors when the signal is blocked due
to tall towers, walls within a building or under the
ground. Therefore, navigational directions using GPS
will be prone to errors in a closed indoor parking lot.
Usage of GPS is suited for outdoor open parking lots
where there is less chance for signal blocking.
Accuracy of the GPS signal is also dependent on
availability of satellite.
3.2.2 Machine Vision
A visual camera can be used for license plate
recognition or identifying parking lot occupancy
using machine vision. The camera should be placed
near the entrance of a closed parking lot for license
plate recognition (LPR). Based on the number of
vehicles entered and exited it can help to get the count
of vacant parking spaces. However, occupancy status
of parking spaces cannot be attained using this
system. Video processing of parking lot using a
camera is not ideal as it requires continuous transfer
of large bandwidths. Therefore, a video should be
broken to images at regular intervals and frame rates
to facilitate continuous monitoring of the parking lot
(Enríquez et al., 2017). For parking spaces occupancy
detection, a camera can be installed overhead to a
parking lot and relevant image detection algorithms
can be used to segment vehicles and detect occupancy
of parking space. A camera is suited for open parking
lots as it can cover large number of parking spaces
(Ichihashi et al., 2009). However, it is susceptible to
limitations such as; occlusion and shadow effects,
distortion and lightning change. These limitations can
be removed with the use of 3-D scene information
(Huang et al., 2013). Since limited number of cameras
can cover large number of parking spaces the
expenditure is considered minimal.
3.2.3 Vanet
This system uses wireless communication devices to
provide services such as; smart parking and antitheft.
Road side unites (RSUs) would be widely placed
across parking lots and vehicles should be installed
with on-board units (OBU). A Trusted authority will
be responsible for registrations of OBU and RSUs.
(Lu et al., 2010) Therefore, once a vehicle approaches
the parking lot installed with RSUs navigational
information to the vacant parking space will be
provided to OBU. These devices are not sensitive to
environment and are suitable for closed and open
parking lots. However, installation and maintenance
of RSUs in the parking lot would be expensive. In
order to achieve accurate parking occupancy data and
navigational information all the vehicles must install
OBU. Parking occupancy data is prone to errors if
there are vehicles without OBU are parked.
3.2.4 Multi-Agent Systems
These kind of systems makes use of multiple
mediums such as sensors, mobile, algorithms, visual
camera, etc. These systems are also capable of
incorporating aspects such as user preference,
RESIST 2018 - Special Session on Resilient Smart city Transportation
604
importance, etc., in finding a vacant parking space for
the driver. Multi-agent systems are considered as
foundation for automation of smart parking systems.
A user can select a parking space using a mobile or
web application and based on the user importance and
preference, a parking space will be selected. The user
will also receive navigational information to reach the
parking space. Java tools such as JaCaMo and
environment such as CArtAgo can be used in the
architecture (Bilal et al., 2012). Machine vision
systems or VANETs can be used instead of using
sensors. Usage of multiple systems are supported in
this architecture. These systems are suitable for both
open and closed parking lots. The expenditure would
be dependent on the usage of technology to identify
occupancy status of parking spaces.
3.2.5 Neural Networks
Neural network is a data processing system which is
inspired by brain nervous system. Neural networks
have evolved over the years and various types of
neural networks were developed such as; fuzzy,
neural network, fluid neural network, feed forward
and convolution neural network. Neural networks can
be combined with machine vision to achieve
automation. Neural networks were used in efficient
recognition of license plates in real time videos
(Rahman et al., 2003, Villegas et al., 2009). In one of
the study, images from morning and night were taken
separately to train the neural network and a two
layered feed-forward network with hidden sigmoid is
used to produce accurate results in detection of
available parking spaces (Jermsurawong et al., 2012).
Deep learning is a branch of machine learning which
uses neural networks in object detection and
classification. There is another evolving technology
such as convolution neural networks which would
take images as input and is more efficient in analysing
images. In a recent study, convolution neural
networks was used along with machine vision to
capture parking occupancy information efficiently
(Amato et al., 2017). This technology would function
as an efficient tool in data processing while it is not
involved in real time data capturing. Therefore, it is
suitable for open and closed parking lots with
minimal expenditure.
3.2.6 Fuzzy Logic
Fuzzy logic is an approach which incorporates
multivalued logic in evaluating. Fuzzy logic can be
used to develop forecasting models based on sample
data. Similar to neural networks, fuzzy logic can also
be used in multi-agent systems. According to a study,
a sample of parking spaces availability information
for 5 days using machine vision was taken to predict
the availability of parking spaces in future dates using
fuzzy logic (Chen et al., 2013). Fuzzy logic supports
autonomy in providing information on the availability
of parking spaces. The accuracy of forecast models
would not be high without validating with real time
data. Therefore, combination of fuzzy logic models
with machine vision or sensor technologies would
increase the accuracy of the overall system. These
systems are suitable for both open and closed parking
lots. The expenditure would be minimal if image
processing is used along with fuzzy logic to estimate
available parking spaces for the future as well as
provide real time availability of parking spaces. Since
this technology is not involved in the real time data
capturing process, it can be used for closed and open
parking lots with minimal expenditure.
Not all of the sensors and technologies are suitable
for gathering real time occupancy information of
open parking lots. Even though sensors are widely
used to acquire parking occupancy information, they
would be expensive to install and maintain on large
number of parking spaces. Few sensors such as
infrared and ultrasonic are sensitive to environment
and are not suitable for outdoor open parking lots.
Technologies such as; machine vision, multi-agents
systems and fuzzy logic are suitable for open parking
lots to acquire parking occupancy information and
GPS can be used to provide navigational directions.
3.3 Smart Parking Applications
There are many free smart parking applications
available in Google Play Store for Android and the
Apple application store for iOS. Previously,
reservation of parking space was done by calling to
the service provider and now with the current usage
of internet and smartphones, these services are
provided online using mobile and web applications.
These applications serve as decision support systems
for the driver in occupying a vacant parking space.
For instance, if the application shows a particular
parking lot of choice to be full, the driver can search
for nearby parking lots with available parking spaces
or choose another destination. In this way smart
parking applications serve as decision support
systems in occupying available parking spaces. Table
2, shows smart parking applications available online
provide parking occupancy and guidance
information. All the applications are provided in
limited cities of mentioned countries.
Smart Parking Tools Suitability for Open Parking Lots: A Review
605
Table 2. Smart parking applications and their use of
technologies and sensors.
Country
Sensors/Technology
used
Austria,
Germany
Sensors
New Zealand
Sensors, RFID
Japan, US,
UK,
Germany,
Brazil
Sensors
US
M4 Smart sensors,
LPR
US
Sensors
Canada
Sensors
France
Sensors
US
Sensors
UK
Magnetometer
Sweden,
Denmark, etc.
Transactional data
and crowdsourcing
US
Sensors, Machine
vision
US
Magnetometer
US
Sensors
US,
Germany,
Sweden, etc.
Predictive analytics,
sensors
US
Sensors
US
Crowdsourcing
All the applications mentioned in Table 2 provide
smart parking services in closed parking lots which
support reservation and all of them are available in
mobile application stores. However, Open Spot is the
only application which is currently unavailable and
discontinued by Google. Navigational directions to
the free available parking space are provided which
improve the efficiency of parking behaviour. A
percentage or number of available parking spaces are
shown for selected parking lots. None of the
applications give the driver a choice in selecting a
particular parking space. Most of the applications use
underground wireless sensors such as magnetometers
to get real time parking occupancy information. Setup
and maintenance cost of wireless sensors will be high
and it is mandatory to replace all sensors again when
the battery life is depleted. Since, a sensor needs to be
placed in all the parking spaces, operational and
maintenance costs of such sensors would be high.
Crowd sourcing application are suited for
open/closed parking lots and the expenditure is
considered very minimal as no hardware installations
or maintenance is required. However, the users need
to update occupancy details of parking spaces every
time they park the vehicle and the accuracy of the
information will be dependent on the user updating
the occupancy details and is prone to human errors.
Applications such as; Parkopedia and EasyPark
Group use historical data for predicting parking
occupancy information. Algorithms like fuzzy logic,
time series are used to predict parking occupancy
information based on historical or sample data. These
applications are operational in more number of cities
and countries compared to other applications. All
these applications provide an overview of parking
occupancy which facilitates in decision making.
4 DISCUSSION
The review of existing smart parking applications
show that most of these applications use sensors for
parking occupancy detection in closed parking lots
which would require considerable amount of
expenditure for installation and maintenance. Few
applications use predictive analytics or crowd
sourcing which can also be used on open parking lots.
The applications which used predictive analytics
were operational in more number of cities and
countries than the applications which used sensors for
parking occupancy detection. The difference might be
due to less expenditure in using predictive analytics
than deploying sensors in all parking spaces.
However, real time occupancy information of parking
spaces cannot be acquired using predictive analytics
and accuracy of parking occupancy information by
crowdsourcing is not reliable. None of the
applications provide real time parking occupancy
information of open parking lots. Lack of economic
returns can be one of the reasons for not providing
real time occupancy information of open parking lots.
Since there would not be any immediate economic
gains with the use of smart parking tools at open
parking lots it would be well suited to use a sensor or
technology with minimal expenditure to make it
financially viable.
According to previous literatures, several smart
parking sensors and technologies were tested and
reviewed. However, sensors or technologies were
largely used for closed parking lots which support
reservation. None of the technologies were used to
improve efficiency of parking in open parking lots.
Ultrasonic and infrared sensors are sensitive to
RESIST 2018 - Special Session on Resilient Smart city Transportation
606
environmental conditions which might lead to
inaccurate parking occupancy information. Sensors
such as magnetometers and microwave radar are not
sensitive to environmental conditions but are
expensive to install and maintain at open parking lots.
Therefore, sensors are not ideal for open parking lots
as expenditure should be minimal. Technologies such
as VANET are also expensive and can be ruled out
for open parking lots. However, VANET can be
suitable in the future when all the vehicles are
connected using on-board and road side units. As
expenditure plays an important role in the choice of
technology, it is found that; machine vision, multi-
agent systems, along with convolution neural network
or fuzzy logic are suitable to provide parking
occupancy information. Neural networks are used
mostly for license plate recognition and there are
limited studies which use convolution neural
networks for parking occupancy detection. In deep
learning, large datasets such as Alexnet are already
available and additional data would facilitate in
further improving image classification. Convolution
neural networks and deep learning are contemporary
technologies which can be used in parking occupancy
detection of open parking lots.
The parking occupancy detection using machine
vision will vary between closed and open parking lots
as there can be varying light conditions in an open
outdoor parking lot. As mentioned in Section 3.2,
challenges faced by low lighting and shadow
conditions can be addressed. Therefore, machine
vision using visual camera is one of the feasible smart
parking technology for acquiring real time parking
occupancy information in open parking lots. Machine
vision can be associated with convolution neural
networks and deep learning for efficient image
classification. Along with real time occupancy
information, a short term prediction of occupancy can
be acquired using algorithms like fuzzy logic and
time series. Since, reservation is not possible in an
open parking lot, a short term prediction would
facilitate the driver in the decision making process of
choosing a parking lot. The prediction of parking
space availability can also be made based on the
location of the driver and its distance to the desired
parking lot or destination. An optimal route can be
calculated from the location to destination based on
the distance of the route using which estimated arrival
time can be acquired. Therefore, parking space
occupancy prediction based on estimated arrival time
can be shown to the driver. In this way smart parking
applications can serve as efficient decision support
systems. Due to various types of parking lots, such as
outdoor, multi-storey, basement or indoor, a
combination of smart parking technologies can be
used to improve the efficiency of parking behaviour
with minimal expenditure.
5 CONCLUSIONS
This paper reviews smart parking tools suitability for
open parking lots. All the existing smart parking
technologies and applications are not suitable for
open parking lots due to varying environmental
conditions and high expenditure. As there are no
immediate economic gains from providing smart
parking services in an open parking lot, expenditure
plays an important role in the choice of smart parking
technologies. Parking guidance system can be used to
get the count of available parking spaces while
machine vision can be used to acquire real time
parking occupancy information on open parking lots
due to its minimal expenditure. However, there is no
single ideal technology suitable for parking
occupancy detection. Based on the type of parking lot
and size, different combination of smart parking
technologies can be used for efficient and financially
viable parking occupancy detection. In order to
further improve parking efficiency, navigational
directions should be provided to a vacant parking
space. Therefore, in order to address this challenge
further research in the use of deep learning and multi-
agent systems would help to provide real time parking
occupancy information along with navigational
directions to the available parking space in an open
parking lot.
REFERENCES
Amato, G., Carrara, F., Falchi, F., Gennaro, C., Meghini, C.
& Vairo, C. 2017. Deep learning for decentralized
parking lot occupancy detection. Expert Systems with
Applications, 72, 327-334.
Bao, X., Zhan, Y., Xu, C., Hu, K., Zheng, C. & Wang, Y.
2017. A novel dual microwave Doppler radar based
vehicle detection sensor for parking lot occupancy
detection. IEICE Electronics Express, 1-12.
Bilal, M., Persson, C., Ramparany, F., Picard, G. &
Boissier, O. Multi-agent based governance model for
Machine-to-Machine networks in a smart parking
management system. IEEE International Conference
on Communications 2012 Ottawa.
Boltze, M. & Puzicha, J. 1995. Effectiveness of the parking
guidance system in Frankfurt am Main. Parking Trend
International, 27-30.
Chen, Z., Xia, J. C. & Irawan, B. 2013. Development of
Fuzzy Logic Forecast Models for Location-Based
Smart Parking Tools Suitability for Open Parking Lots: A Review
607
Parking Finding Services. Mathematical Problems in
Engineering, 2013, 1-6.
Chinrungrueng, J., Sunantachaikul, U. & Triamlumlerd, S.
2007. Smart Parking: an Application of opticalWireless
Sensor Network. IEEE.
Enríquez, F., Soria, L. M., Álvarez-García, J. A., Velasco,
F. & Déniz, O. Existing Approaches to Smart Parking:
An Overview. International Conference on Smart
Cities, 2017. Springer, 63-74.
European Automobile Manufactures Association. 2017.
Average Vehicle Age [Online]. ACEA. Available:
http://www.acea.be/statistics/tag/category/average-
vehicle-age [Accessed].
Fraifer, M. & Fernström, M. Investigation of smart parking
systems and their technologies. Thirty Seventh
International Conference on Information Systems. IoT
Smart City Challenges Applications (ISCA 2016),
Dublin, Ireland, 2016. 1-14.
Frey, C. H., Unal, A., Rouphail, N. M. & Colyar, J. D. 2012.
On-Road Measurement of Vehicle Tailpipe Emissions
Using a Portable Instrument. Journal of the Air &
Waster Management Association, 992-1002.
Gallivan, S. 2011. IBM global parking survey: Drivers
share worldwide parking woes technical report.
Technical report, IBM.
Geroliminis, N. & Daganzo, C. F. 2008. Existence of urban-
scale macroscopic fundamental diagrams: Some
experimental findings. Transportation Research Part
B: Methodological, 42, 759-770.
Geroliminis, N. & Sun, J. 2011. Properties of a well-defined
macroscopic fundamental diagram for urban traffic.
Transportation Research Part B: Methodological, 45,
605-617.
Guler, S. I., Menendez, M. & Meier, L. 2014. Using
connected vehicle technology to improve the efficiency
of intersections. Transportation Research Part C:
Emerging Technologies, 46, 121-131.
Harding, J., Powell, G., Yoon, R., Fikentscher, J., Doyle,
C., Sade, D., Lukuc, M., Simons, J. & Wang, J. 2014.
Vehicle-to-vehicle communications: Readiness of V2V
technology for application.
Hassoune, K., Dachry, W., Moutaouakkil, F. & Medromi,
H. Smart parking systems: A survey. Intelligent
Systems: Theories and Applications (SITA), 2016 11th
International Conference on, 2016. IEEE, 1-6.
Huang, C.-C., Tai, Y.-S. & Wang, S.-J. 2013. Vacant
Parking Space Detection Based on Plane-based
Bayesian Hierarchical Framework. IEEE Transactions
on Circuits and Systems for Video Technology, 1-13.
Ichihashi, H., Notsu, A., Honda, K., Katada, T. & Fujiyoshi,
M. Vacant parking space detector for outdoor parking
lot by using surveillance camera and FCM classifier.
Fuzzy Systems, 2009. FUZZ-IEEE 2009. IEEE
International Conference on, 2009. IEEE, 127-134.
Idris, M., Leng, Y., Tamil, E., Noor, N. & Razak, Z. 2009a.
Саг park system: a review of smart parking system and
its technology. Information Technology Journal, 8,
101-113.
Idris, M., Tamil, E., Noor, N., Razak, Z. & Fong, K. 2009b.
Parking guidance system utilizing wireless sensor
network and ultrasonic sensor. Information Technology
Journal, 8, 138-146.
Jan Minx, K. S., Glen Peters, John Barrett 2008. An
Analysis of Sweden's Carbon Footprint.
JermsuraWong, J., Ahsan, M. U. & Haidar, A. Car Parking
Vacancy Detection and Its Application in 24-Hour
Statistical Analysis. 10th International Conference on
Frontiers of Information Technology, 2012.
Kianpisheh, A., Mustaffa, N., Limtrairut, P. &
Keikhosrokiani, P. 2012a. Smart Parking System (SPS)
Architecture using ultrasonic detector. International
Journal of Software Engineering and its Applications,
6, 51-58.
Kianpisheh, A., Mustaffa, N., Limtrairut, P. &
Keikhosrokiani, P. 2012b. Smart parking system (SPS)
architecture using ultrasonic detector. International
Journal of Software Engineering and Its Applications,
6, 55-58.
Kotb, A. O., Shen, Y.-C., Zhu, X. & Huang, Y. 2016.
iParker—A New Smart Car-Parking System Based on
Dynamic Resource Allocation and Pricing. IEEE
Transactions on Intelligent Transportations Systems,
17, 2637-2647.
Lu, R., Lin, X., Zhu, H. & Shen, X. 2010. An intelligent
secure and privacy-preserving parking scheme through
vehicular communications. IEEE Transactions on
Vehicular Technology, 59, 2772-2785.
Mahmud, S., Khan, G., Rahman, M. & Zafar, H. 2013. A
survey of intelligent car parking system. Journal of
applied research and technology, 11, 714-726.
Mathur, S., JIN, T., Kasturirangan, N., Chandrasekaran, J.,
Xue, W., Gruteser, M. & Trappe, W. Parknet: drive-by
sensing of road-side parking statistics. Proceedings of
the 8th international conference on Mobile systems,
applications, and services, 2010. ACM, 123-136.
Mimbela, L. E. Y. & Klein, L. A. 2000. Summary of vehicle
detection and surveillance technologies used in
intelligent transportation systems.
Mouskos, K., Boile, M. & Parker, N. A. 2007. Technical
solutions to overcrowded park and ride facilities. New
Jersey Department of Transportation.
Polak, J. & Vythoulkas, P. 1993. An assessment of the
state-of-the-art in the modelling of parking behaviour.
TSU REF.
POlycarpou, E., Lambrinos, L. & Protopapadakis, E. Smart
parking solutions for urban areas. World of Wireless,
Mobile and Multimedia Networks (WoWMoM), 2013
IEEE 14th International Symposium and Workshops on
a, 2013. IEEE, 1-6.
Pullola, S., Atrey, P. K. & EL Saddik, A. Towards an
intelligent GPS-based vehicle navigation system for
finding street parking lots. Signal Processing and
Communications, 2007. ICSPC 2007. IEEE
International Conference on, 2007. IEEE, 1251-1254.
Rahman, C. A., Badawy, W. & Radmanesh, A. A real time
vehicle's license plate recognition system. Advanced
Video and Signal Based Surveillance, 2003.
Proceedings. IEEE Conference on, 2003. IEEE, 163-
166.
RESIST 2018 - Special Session on Resilient Smart city Transportation
608
RAHMAN, M. S., PARK, Y. & KIM, K.-D. Relative
location estimation of vehicles in parking management
system. Advanced Communication Technology, 2009.
ICACT 2009. 11th International Conference on, 2009.
IEEE, 729-732.
REVATHI, G. & DHULIPALA, V. S. Smart parking
systems and sensors: A survey. Computing,
Communication and Applications (ICCCA), 2012
International Conference on, 2012. IEEE, 1-5.
SHAHEEN, S. 2005. Smart parking management field test:
A bay area rapid transit (bart) district parking
demonstration. Institute of Transportation Studies.
SHOUP, D. C. 2006a. Cruising for parking. Transport
Policy, 13, 479-486.
SHOUP, D. C. 2006b. Cruising for Parking. Transport
Policy.
STATISTA. 2017. Average age of light vehicles in the U.S.
from 2003 to 2016 (in years) [Online]. The Statistics
Portal. Available:
https://www.statista.com/statistics/261881/average-
age-of-light-vehicles-in-the-united-states/ [Accessed
17 Nov 2017].
THIERER, A. & HAGEMANN, R. 2014. Removing
Roadblocks to Intelligent Vehicles and Driverless Cars.
Available:
https://www.mercatus.org/system/files/Thierer-
Intelligent-Vehicles.pdf [Accessed Sep].
VILLEGAS, O. O. V., BALDERRAMA, D. G.,
DOMÍNGUEZ, H. & SÁNCHEZ, V. G. C. 2009.
License plate recognition using a novel fuzzy
multilayer neural network. International journal of
computers, 3, 31-40.
WATERSON, B., HOUNSELL, N. & CHATTERJEE, K.
2001. Quantifying the potential savings in travel time
resulting from parking guidance systems-a simulation
case study. Journal of the Operational Research
Society, 52, 1067-1077.
WHITE, P. 2007. No Vacancy: Park Slopes Parking
Problem And How to Fix It. Internet: http://www.
transalt. org/newsroom/releases/126 [7 Januari 2016].
YANG, Z., LIU, H. & WANG, X. The research on the key
technologies for improving efficiency of parking
guidance system. Intelligent Transportation Systems,
2003. Proceedings. 2003 IEEE, 2003. IEEE, 1177-
1182.
Smart Parking Tools Suitability for Open Parking Lots: A Review
609