=Paper=
{{Paper
|id=Vol-3028/D2-05_ESAAMM_2021_paper_6
|storemode=property
|title=Algorithmic Planning, Simulation and Validation of Smart, Shared Parking Services using Edge Hardware
|pdfUrl=https://ceur-ws.org/Vol-3028/D2-05_ESAAMM_2021_paper_6.pdf
|volume=Vol-3028
|authors=Muralikrishna Thulasi Raman,Andreas Graf,Benedikt Nieheus,Marco Aiello
}}
==Algorithmic Planning, Simulation and Validation of Smart, Shared Parking Services using Edge Hardware==
https://ceur-ws.org/Vol-3028/D2-05_ESAAMM_2021_paper_6.pdf
Algorithmic Planning, Simulation and Validation of
Smart, Shared Parking Services using Edge
Hardware
Muralikrishna Thulasi Raman Andreas Graf Benedikt Nieheus Marco Aiello
Itemis AG Itemis AG Itemis AG IAAS, University of Stuttgart
Stuttgart, Germany Stuttgart, Germany Lünen, Germany Stuttgart, Germany
mthulasi@itemis.de graf@itemis.de niehues@itemis.de marco.aiello@iaas.uni-stuttgart.de
Abstract—A major problem with vehicles used in densely optimization and mobility models. Our proposed design is
populated areas is parking. Part of the problem is the scarcity illustrated in Section III and its implementation in Section IV.
of the resource in busy areas and the non optimal utilization of The evaluation of the proposal by means of simulations is
private spots. We propose the ’Smart Shared Private Parking’
model that aims at including in the parking pool also private presented in Section V. A discussion and final remarks are
parking spots. In an urban area with both commercial and offered in Section VI.
residential buildings, frequently household owners do not utilize
their own parking lots during office hours, long duration of II. S TATE OF THE A RT
shopping and vacations. At the same time, such owners want to be Innovative parking and mobility models have been investi-
able to park in their lots when needed. Therefore, to engage the
residents and allow them to provide their parking spaces when gated before, e.g., Some of the works that are used to derive
unused one needs a real-time, distributed infrastructure that the points of common knowledge are [4, 17, 9, 18, 3, 12,
supports dynamic allocation of spaces. We propose a conceptual 7, 6]. One of the motivations for designing parking solutions
deployment of edge hardware and communication mechanisms is to solve the problem of commute time in searching for a
to enable shared private parking sharing. We illustrate an parking spot [18]. Not only time, but also vehicular (CO2 )
infrastructure setup required to analyse various aspects of the
model and show its effectiveness in parking allocation in terms emissions increase with the difficulty of finding a spot. The
of vehicular emission, utilisation rate, and revenue as metrics effect of increased vehicular (CO2 ) emissions contributes to
considering all the stakeholders. A simulation based on a district poor air quality of the considered urban area. Vehicular (CO2 )
of Dublin provides quantitative instances for these metrics. emission is an important parameter to quantify the efficiency
Index Terms—V2X, VANET, Vehicular Communication, Smart of parking and mobility solutions [12]. Apart from vehicular
Shared Private Parking, Smart Gate, Edge Computing
(CO2 ) emissions, traffic congestion is another important factor
I. I NTRODUCTION to be considered to quantify the quality of a parking model
[17]. The goal of a typical innovative parking model is to
Urbanization and Industrialization are the major factors improve the air quality by reduced emissions and reduced
contributing to the increased dependence on vehicles [3]. traffic congestion [9]. These points form the current common
The increased dependence on vehicles leads to problems like knowledge and are predominantly used in the related work.
traffic congestion and pollution. Due to the recent COVID
pandemic, the use of public transport has decreased while A. Parking models
the use of private transport has increased [2]. The increase Modelling the parking demand and supply management has
in usage of private transport leads to higher turnover and been the goal of several works. The importance of different
demand for parking. Hence, there is a need for new model parking modelling strategies is highlighted in [5]. Also, the
that can increases the supply of parking lots. The present work authors of [5] propose to bring up a balance of the new
proposes a parking model that aims to increase the supply of additions through geospatial analysis [5]. A parking allocation
parking lots. The contributions include: model using a centroid based approach for finding an optimal
• A novel model for the sharing of private parking lots that parking spot is used.
considers user preferences while offering the possibility 1) Shared parking model I - Focus on overtime parking
to optimize utilization rate. information: The boons of shared parking and managing the
• A design and implementation of the model. overtime stay through different correlation analyses are illus-
• A simulation based on the city of Dublin to evaluate the trated in [17]. In particular, the correlation between various
model effectiveness. parameters such as vehicle arrival numbers, parking duration
The rest of the paper is organized as follows. Section II and overtime occupancy are considered and the authors define
overviews related work that addresses the problem of parking a model where residential areas can be used to contribute to
Copyright © 2021 for this paper by its authors. Use permitted under Creative Commons License Attribution 4.0 International (CC BY 4.0).
the parking demand in an urban area. The model maximizes C. Existing commercial providers
the revenue by considering the optimal combination of the In addition to research work, a number of companies are al-
given parameters. ready providing services in the field of smart parking. Zenpark
2) Shared parking model II- Focus on walking distance is a shared parking solution available in France and Belgium
and other preferences: Driver and owner preferences are the [26]. The idea is to allow individuals and other commercial
focus of [9].The basic parameters for the shared parking model landowners to rent out their unused parking spots. Mobypark
are the location of the parking lot with respect to the most also offers a similar solution for the capitals of France,
visited destination points in the area, the distance between Belgium and the Netherlands [22]. Bosch community-based
the parking lot and the destination being within an acceptable parking is a solution based on cars informing of the availability
walking range, and parking lots being open to the public. The of vacant spaces via a dedicated cloud [4]. Noticeable is
model maximizes the utilisation rate of the parking lots and the fact that the information about the sender is anonymous
minimises the walking distance between the parking lots and ensuring privacy to members of the community.
destination points. The model supports the use of parking
lots in residential areas to serve the users having nearby III. S YSTEM D ESIGN
commercial buildings as destination points.
The work proposed employs smart gate hardware (with
the same capabilities as an RSU) in place of a common
B. Large scale mobility models and simulation scenarios
RSU. The objective is to reduce the costs incurred due to
Large scale mobility models and simulation scenarios are the provisioning of separate hardware units that generally add
necessary in assessing the routine traffic and formulating steps up the expenses of a smart city. Conceptually, the hardware
required to manage varying traffic scenarios. These models is said to be in place of an electrical garage device to sense
also serve as a base to assess different concepts of mobility the spots, communicate and actuate the mechanical parts of
such as parking and intermodal mobility. the parking lot when required. In the work, each RSU or
1) MoST and PyPML: Simulation scenarios based on the smart gate hardware represent a parking lot in the simulation.
city-state Monaco are proposed in [6]. The MoST scenarios are In the requirement analysis phase, we distributed a survey to
based on a synthetic traffic demand rather than on actual data 34 employees from various German locations of the company
[6]. The demand presents a peak hour in the morning which Itemis AG. The survey was conducted in December 2020. The
could possibly comprise intermodal traffic. The scenario can goal was to assess the mindset of car drivers and residential
help to optimise individual planning of mobility and related parking lot owners. The results of the survey are in line with
mobility demands like parking. Such planning in turn helps the the conclusions of the works [9, 17] and help us derive
to study the impact on traffic congestion based on the used the mathematical model at the heart of our proposal, illustrated
optimization model. The use cases are implemented using next.
Eclipse Simulation of Urban MObility (SUMO) [20].
PyPML is a Parking Monitoring Library designed using A. Mathematical model
SUMO and using the Monaco SUMO Traffic (MoST) scenario The mathematical model of the work is designed consider-
[7]. The library assumes that optimal placement is already ing the concept of overtime parking and drivers’ preference to
defined. The core functionalities of the library are related to walk for a certain distance. Utilisation rate Um in Equation 1
monitoring, optimization, and control of the parking condi- is the result of matching time preferences and additional
tions. preferences of a vehicle with a parking lot. In a parking lot
2) Smart parking using Vehicular Ad Hoc Network m, the variable xmn (satisfying time preferences and additional
(VANET) in the Dublin scenario: The authors of [12] anal- preferences) equal to 1 multiplied by duration of parking gives
yse vehicular communication aspects such as the model for the utilisation rate of parking lot m by a vehicle n. The sum
placement of Road Side Units (RSUs) and illustrate a process of utilisation by vehicles n1 ....nN at a particular hour gives
for the gathering a data from the websites of the Admin- the overall utilisation of a parking lot m at a particular hour.
istration department of Inner city of Dublin. In their work, Equation 1 is an extended version of the mathematical model
only commercial parking lots have been considered for the in [9]. The extension comes by combining the model with
analysis. Details on the choices of software components and the utilisation rate, as defined in Equation 3. The revenue
methods to generate the datasets required for the work are generated by a parking lot is also part of this extended model.
given. Vehicle to Vehicle (V2V) communication frameworks Finally, the model considers a distribution of the traffic to lots
like VEhicles In Network Simulation (VEINS) [25], Objective by considering for how long a parking was not utilized for
Modular Network Testbed in C++ (OMNET++) [24] are used a threshold duration (say last 5 minutes). The distribution is
along with Eclipse SUMO [20] to realize the idea of “Smart ensured by the introduction of tlr,m variable.
parking using VANET.” Simulation results comparing CO2
emissions per vehicle between a model without using vehicular N
tndur · xmn
P
communication and a model using vehicular communication
conclude the work. Um = n=1 (1)
Tm (H)
scenarios. We chose this framework as it supports aspects of
connected mobility in the form of simulators. Also, support for
1 if, (Dm ≤ Dmax,n ) & (Pm = Pn ) &
applications development using a high level language (JAVA)
(tstart,m < t + tndur < tend,m ) &
xmn = (2) made this framework a good choice for the work proposed.
(Fm > 0) & (t − tlr,m > tth )
A. Simulators
0 Otherwise
Simulation of Urban Mobility (SUMO) is a free, open-
where Um is the utilisation rate of the parking lot ’m’, source simulation software for modelling inter-modal traffic
tdur n represents the duration for which the vehicle ’n’ needs systems [20]. Simple Network Simulator (SNS) is the network
parking. Tm (H) = tend,m -tstart,m represents the total duration of simulator which is a part of Eclipse MOSAIC framework. The
availability of a parking lot. tstart,m is the time at which parking simulator has all the basic communication attributes required
lot ’m’ starts to being offered. tend,m is the time at which for vehicular communication. The role of the environment
parking lot ’m’ is no longer available. But in the work, the simulator is to simulate events such as obstacles and weather
value of Tm (H) is equal to 1 as we consider the simulation conditions. Apart from the mapping simulator, the Application
scenarios on a per hour basis. lr,m is the time at which a parking Simulator emulates the necessary application that needs to be
lot is last reserved. tth is the threshold time before which a deployed in the components such as RSU or vehicles.
parking lot can be utilized. Dm is the actual walking distance
between the parking lot ’m’ and the destination. Dmax,n is the B. Setup
preferred walking distance by the driver of vehicle ’n’ to the The Scenario-Convert tool is a part of Eclipse MOSAIC
destination. Pm is the set of service offering parameters from [11]. The tool is used to generate cleaned SUMO network
the residential parking lot ’m’ and Pn is the set of services from an Open Street Map (OSM) file. It also generates
demanded by the driver of vehicle ’n’. Fm is the number of the configurations required for other simulators. In the work
free spots available in the parking lot ’m’. m = m1 ,.......mM , proposed, shared parking lots are to be provisioned at different
Each of the values of m represent the response i.e. by a parking geographic coordinates using Google maps [14]. The cleaned
owner offering a parking. n = n1 ,.......nN , Each of the values network obtained through this step is given by Figure 1. The
of n represent the demand i.e. a driver of vehicle ’n’ looking basic building blocks of such a SUMO network are edges,
for a parking. lanes and junctions which can be seen in Figure 1.
In Equation 3, Revenue Rm is the collective sum of parking
charges and overtime parking charges incurred by vehicles
n1 ,.......nN using the parking lot m. The decisions xmn and
xod.mn (given by Equations 2 and 4) coordinate the decisions
of considering the parking usage of vehicles whose timings
and preferences have been matched with the parking lot.
N
X
Rm = [(tndur · xmn · cm ) + (tnod · xod.mn · cod.m )] (3)
n=1
(
1 if, (t > tndur ) & (xmn = 1)
xod.mn = (4)
0 Otherwise Fig. 1. SUMO road network extracted using Scenario-Convert tool
where Rm is the revenue generated by a shared residential We generate parking lot data preserving essential details
parking lot ’m’. tdur n is the duration for which the vehicle ’n’ such as geographic coordinates, the number of free spots, the
occupies a parking lot. xmn is the decision variable that denotes presence of charging spots. Using the coordinates where the
whether a vehicle ’n’ occupies a parking lot ’m’ or not. cm parking lots are provisioned, we create a database of parking
is the fixed cost to occupy a spot in parking lot ’m’ for an lot entries and schema for reservations and overtime estimates
hour. tod n is the duration for which the vehicle ’n’ is parked (given in Figure 2).
overtime. xod.mn is the decision variable that denotes whether
a vehicle ’n’ occupies a parking lot ’m’ overtime or not. cod.m C. Demand data and integration into the scenario
is the cost to occupy a spot in parking lot m overtime for an The real demand is used in the simulation. We do so by
hour. resorting to the demand data from the official website of
Dublin administration [10]. The data is collected by means
IV. I MPLEMENTATION of a survey taken at 33 locations around the city centre. The
The base framework used to implement the proposed model survey locations are strategically planned so that the number
is Eclipse MOSAIC [11]. MOSAIC is a combination of several of inbound and outbound vehicles are tracked accurately. The
simulators coupled together to demonstrate connected mobility survey locations are marked in Figure 3. We consider the
and to assess their effects on large scale city-based traffic number of vehicles moving in and out of the city at entry
Fig. 3. Survey locations where inbound and outbound vehicles of Dublin
City Centre are tracked [10]
module, a communication module and an Operating System
(OS).
E. Data flows
To help understanding the operation of the system, let us
Fig. 2. Class diagram to explain the schema of the generated dataset consider the two most relevant data flows, that is, the parking
request and the resume signal.
1) Parking request: The parking request from a vehicle
point 31. Routes denote the paths taken by the vehicles in flows through RSU to the web service. The vehicle or car
SUMO network. Although the scenario-convert tool of Eclipse driver initiates the demand for a parking request by sending a
MOSAIC is capable of converting a set of geographic coordi- geocast message consisting of a request for parking. Geocast
nates into SUMO routes, the newly generated routes cannot be message communication is a form of message communication
directly merged into the route navigation database which is a where the message is sent to all the potential receivers in a
small shortcoming of the tool (observed until December 2020). geographical location [23]. A RSU receives the message and
To overcome this, a custom script is developed to use the passes it to the web service. The web service makes use of the
existing route generation functionalities of Eclipse MOSAIC. shared private parking model to allocate a suitable parking lot
to the vehicle/car driver. The smart gate hardware units serve
D. Applications both the vehicle parked in the parking lot and the vehicles that
The components involved in the simulation are vehicles and request . At the same time, they represent the shared residential
RSU. We developed applications that need to be deployed in parking lots in the system. The hardware unit receives the
each component. The application represents the behaviour of response from cloud-based web service regarding the parking
the component (say parking functionality in a vehicle). The lot allocated in the destination of the vehicle. Figure 4 illsu-
modules that a vehicle in simulation possesses are a navigation trates how a parking request is initiated by a vehicle, served
module, an in-built operating system that coordinates with by the web service and how the communication from and to
internal components and a communication module capable of the vehicle through the Smart gate hardware units (or RSU)
Vehicle to Everything (V2X) communication. The vehicle’s occurs. Finally, the hardware (or RSU) communicates the
application in the work proposed involves the usage of these parking lot availability through a unicast message. A unicast
three modules. RSU forwards the request from the vehicle to message is a message intended for a particular receiver in a
cloud-based web services. In the shared parking model, RSU geographical location. In the simulation, the IP address of the
represents the smart gate hardware which can be equipped vehicle uniquely identifies the receiver [19].
with a vehicular communication stack, enabling the sensing of 2) Resume signal: The resume signal originates from a
the arrival of vehicles and also the actuation of the gate, when vehicle when the parking is completed. The parking duration
required. The smart gate hardware is thus the edge component varies randomly. In order to resume the journey, the resume
that ultimately enables the smart sharing of the parking lots. signal is sent to a RSU nearby, which forwards the signal
Similar to a vehicle, RSU is also equipped with a navigation with all necessary details to the cloud. The aim of this signal
and shared residential parking lots. An important objective
is to match the vehicle’s preferred walking distance and the
distance between the parking lot and the destination of the
vehicle. Open Source Routing Machine (OSRM) provides
routing and navigation facilities offline and is based on OSM
data [21]. We deploy the service as a docker container [8].
Then comes the matching strategies in which one of them
matches the current time and duration of parking with the
availability of parking lots. Another strategy is to match the
preferences of vehicle with the services offered by a parking
lot. Finally, an allocation module is present to distribute
the demand. In a demand-supply management application,
the important focus is to distribute the demand evenly. This
is applicable to satisfying the different parking demands of
the car drivers by the parking lots. Hence, there is a need
for a decision based allocation module that takes care of
even distribution. In the work proposed, we use a Cellular
Automaton model for this purpose.
Fig. 4. Sequence of the simulation for parking request A Cellular Automaton (CA) (Plural: Cellular Automata) is
a decision making module that work on a group of cells [1].
Using a CA, each cell’s next state is decided by it’s current
transfer is to assess the overtime parking revenue of each state and the nearby cells’ current state [1]. A rule takes care
parking lot and to make parking lots available for vehicles of the transformation of cell’s states into new state [1]. In the
looking for parking. The RSU communicates the overtime work proposed, a one-dimensional CA is used for the purpose
estimation calculated in the web service to the vehicle for of distributing the traffic in the microscopic area chosen. The
future acceptance of vehicles by the parking lots. In short, proposed work uses rule 51. In the work proposed, mapping is
as shown in Figure 5, the resume signal flow resembles the done in such a way that occupancy of parking lots in circular
parking request but differs by the input signal’s content and area with destination of the car driver as centre and with
the overtime response content. preferred walking distance of the driver as radius are mapped
to lattice of parking lots as current state and then transformed
into new state. This realisation is the implementation of a
one-dimensional or linear CA as we only take occupancy (0
or 1) of parking lots in an area into account but not other
parameters like area. The reason is that distance manipulation
module takes care of sensing parking lots in the neighbourhood
area considered. The efficiency of using rule 51 in the work
proposed is proven by utilisation analysis of the shared private
parking lots which mainly requires the parking lots to be
distributed with the demand evenly.
In this work, by lattice we refer to an ordered group of
individual cells [1]. Each cell represents a parking lot. Figure 6
illustrates the working of Rule 51 in graphical form. Three
contiguous cells on the top indicate the current state of cells
where the middle one is the considered cell and cells on
Fig. 5. Sequence of the simulation for resume signal either side indicate its neighbours [1]. Cells are binary: a black
cell represents an occupied lot, while a white one an empty
lot [1]. The cell on the bottom indicates the transformed state.
F. Web Service and Shared Parking model Figure 7 illustrates the process of how the mapping illustrated
The cloud-based web service incorporates both the base and before is realised in the work proposed. The allocation module
shared parking model. Application interfaces act as the access ensures that for a reservation, the closest parking lot that was
to the web service, since it differentiates and redirects the not reserved in the threshold time (tth ) duration is used. The
request from RSU. The evaluation module receives the request calculation of the previous state of the parking lots starts by
from the application interface and proceeds to gather and getting the state of the parking lots nearest to the destination.
elaborate the data and then finally responds to the requester Obtaining parking lots nearby is taken care of by distance
with the appropriate response. A database is also a part of the manipulation module. Then the field last reserved time helps
web service. The database consists of details about commercial to calculate the previous state. The next step is to filter the
parking lots on the basis of the threshold time. The step 1) Emission logs help to assess vehicular emission output
calculation of previous state ensures this. Rule 51 transforms of all the vehicles.
the cells of the lattice into the next state in the transformation 2) Database records (Reservation and Overtime Estimate)
step. The vehicle demanding parking is allocated the parking help to extract the following data:
lot enabled by the allocation module. In the case of multiple • Reservation data (to estimate utilisation)
enabled parking lots in the lattice, the first and foremost entry • Revenue and Overtime Estimates (to estimate over-
in the lattice satisfies the demand generated by a vehicle. all revenue of each parking lot).
Wolfram’s Rule 51 helps to distribute traffic evenly. A single
CA represents a lattice of parking lots.
Fig. 6. Definition of Rule 51 [1]
Fig. 8. Simulation-based testbed setup
Fig. 7. Transformation of cell states
B. Benefits of employing a communication strategy between
V. R ESULTS AND E VALUATION vehicles and parking lots
We evaluate the model and its implementation on the data of We consider a scenario consisting of 4 parking lots each
the city of Dublin. The evaluation is run along five dimensions: with 2 spots as capacity and 76 vehicles, 20 of which are
(A) Realisablity of a simulation-testbed, (B) Utility of Vehicle actively look for parking. The metrics that are used to illustrate
to Infrastructure communication in terms of emissions, (C) the benefits of the proposed model are ”Mean Vehicular CO2
Correlation between emissions and parking ease, (D) Eco- Emission” and ”Reroutes.” The base scenario is realized in
nomic benefits for the parkign owners, and (E) Effects of the SUMO. The scenario reflects the realistic behaviour of a driver.
number of available parking lots. When the free spots are not available in a parking lot, a driver
looks for another parking lot in the vicinity. This information
A. Realisation of a simulation-based testbed to assess the comes to the knowledge of the driver when the vehicle is near
shared private parking model the parking lot. Searching for another parking lot is interpreted
To arrive at a stable infrastructure required for the proposed as a reroute and this, in turn, leads to more CO2 emissions
work, the performance and the applicability of various V2X by each of the vehicles. In scenario with a communication
frameworks have been analysed. According to the College strategy i.e. proposed model, vehicles receive information
of Computing, Georgia Institute of Technology and School about services offered through V2X messages. The vehicle’s
of Civil and Environmental Engineering, Georgia Institute On-Board Unit (OBU) processes the received V2X message
of Technology, a simulation-based testbed can be defined as and instructs the navigation module of the vehicle to go to the
a testbed being capable of simulating different components parking lot with free spots. In both the scenarios, vehicular
required for the experiment, operating the components si- emissions and reroutes are extracted using the inbuilt logging
multaneously as in real-world, and observing the proposed features of SUMO and MOSAIC. These values are obtained
system design changes through logs [13]. Here, we intend after running the simulation.
to set up a simulation-based testbed for a microscopic area. The results show that a communication system makes the
The testbed must be scalable in terms of number of vehicles drivers aware of the services offered and redirects the vehicle
and parking lots. Figure 8 illustrates the setup we realized. to a corresponding parking lot with lesser vehicular CO2
After each simulation run, the scripts facilitate getting logs for emissions i.e. substantially reduced reroutes.
each of the research questions (denoted by the processing and The results are presented in Table I and Figure 9 and they
consolidation step in the figure). The logs are common to both highlight that the rate of decrease in mean vehicular emission
the base model and the shared model. The logs extracted are when equipping the scenario with a communication strategy
then used as inputs for comparison and correlation analyses. is 46.75%. The number of reroutes in the proposed model
The logs that are useful for the analyses are listed next. scenario is 0 as the services offered are communicated to the
vehicle before it reaches the destination whereas, in the case of and demand. The demand data varies each hour. For each hour,
the base model scenario, it is 16.The value 16 is obtained as a the entire simulation is run for 3 varying number of shared
result of running the base simulation using SUMO. Reroute is parking lots i.e. 24, 28 and 32 in addition to the commercial
a standard metric related to parking. The value increases with parking lots (with capacity ranging between 500 and 1100).
the number of times vehicles are getting redirected in search To estimate emissions, we resort to the model proposed in
of a parking spot., which is part of SUMO simulations. [15] which gives the guidelines to estimate emission output
by the vehicles in the simulation. We quantify the benefits
TABLE I achieved by the proposed model through vehicular emissions
R ESULTS OF EVALUATING THE BENEFITS OF EMPLOYING A in terms of CO2 . Emission logs are recorded for every time
COMMUNICATION STRATEGY - I
step of the simulation for all vehicles involved, i.e., both those
Metric Unit Base Proposed
looking for parking and the other ones.
Scenario model 1) Relationship between CO2 emission and distance trav-
scenario elled: to establish the relationship between CO2 emission
Mean Vehicular CO2 milligrams 1.909 × 106 1.016 × 106 value and distance, we perform a correlation analysis. The
Emission (mg)
Reroutes (No unit) 16 0
results, presented in Figure 10), show that there is a strong
linear correlation between CO2 emission values and distance
i.e. with the increase in emission values, distance increases.
2) Linear regression model: The data coming from this
setup is amenable to fit a linear regression model. This is
useful to estimate distances for the emission data coming from
the simulations.
3) Comparison between base and shared models : the two
setups, one not including private parking and one including
them, are compared on the basis of CO2 emissions (mil-
ligrams) and distance travelled (metres) by vehicles in the
simulation. At each time of the day, the parking demand
changes. The results of the comparison are presented by
Figure 11 and Table II. Figure 11 represents the comparison
analysis between base model (without shared private parking
lots) and shared model (with shared private parking lots) in
terms of vehicular emissions and distance travelled by vehicles
in the simulation. Three sets of plots are present in Figure 11.
From top to bottom, each set is plotted for base model and
shared model at different hours of a day. For each hour of
Fig. 9. Results of evaluating the benefits of employing a communication a day, mean vehicular emission and mean distance travelled
strategy - II by vehicles in the simulation are plotted. Each blue bar and
blue stem represents vehicular emission and distance travelled
by vehicles respectively for the base model. Each orange bar
C. Benefits for car drivers and orange stem represents vehicular emission and distance
The vehicles moving in the simulation might have abnormal travelled by vehicles respectively for the shared model.
routes which eventually change the emission values. Hence, Table II is constructed by considering vehicular emissions
in order to remove this abnormality, we run the simulations in all hours of a day. In Table II, the mean value of vehicular
several times to consolidate the result. The hours during which emissions of all the vehicles in base model and shared model
the simulation is run is between 7:00 to 18:00. The reason are compared. Base model and shared model are compared and
behind this decision is that the demand data obtained from evaluated by considering the rate of decrease in emissions in
[10] gives the vehicle volume within this time period. The Table II. The base parking model remains the same as part of
simulation scenario is run for each hour and the logs are the comparison for all the ratios. But based on the demand at
deserialized. The process is repeated for the base model and a particular hour of a day, the output emissions and distance
shared model scenarios for each hour of the day separately values change for the shared parking model due to the varying
and the results are consolidated for further analyses. The number of shared parking lots. The results show that as the
infrastructure for the simulation of the base model scenario number of shared private parking lots increases, the emission
has a V2X framework simulation scenario and Python-Bottle value decreases. The rate of decrease in vehicular emissions
based Web Service [16]. In particular, we consider two cases: (when compared to the base parking model) is 2% linearly
a base model and a shared model scenarios. The first one with an increase in shared parking lots of 4. Since distance
consists of 4 commercial parking lots ranging from 500 to travelled is linearly dependent on emission values, there is a
1100 spots. The second one has varying number of parking lots decrease in distance travelled by the vehicles in the shared
Fig. 10. Correlation analysis between CO2 emission and distance travelled by vehicle
Fig. 11. Results of analysing the CO2 emission and distance travelled by vehicles in the base model and the shared model
private parking model with respect to that in the base parking over time, then there is an entry as over time estimate asso-
model. ciated with the reservation. As per the mathematical model,
reservation of a spot is done only when all the conditions are
D. Benefits for parking owners satisfied, the most important of which is the preferred distance
The utilisation rate of a parking lot is the sum of all the to travel by foot for a car driver. The distance by foot from
reservation time of a parking lot over the unit of time. The the parking lot must be lower than the maximum distance
expression for utilisation rate is given by Equation 1. The preferred by the car driver to walk. Apart from the walking
records from the database are interpreted for estimating the distance condition, we also consider free spots availability
utilisation rate of each parking lot. The reservation entry is also and plan to consider additional preferences such as disability
associated with over time estimates. If the vehicle is parked preferences and presence of charging spots.
TABLE II by the parking lots is proportional to the utilisation rate. As
C OMPARISON BETWEEN BASE PARKING MODEL AND SHARED PARKING the utilisation rate increases, revenue also increases. But the
MODEL IN TERMS OF VEHICULAR CO 2 EMISSION
randomized overtime parking behaviour of the vehicles helps
the parking lots to earn more.
Number Mean CO2 Mean CO2 Rate of de-
of shared emission in emission crease in ve- It is also important to note that the utilisation rate and
parking lots base parking in shared hicular emis- revenue of shared parking lots are constant for all graphs
model (mg) parking sion (%) of Figure 12. This is because we denote the changes in
model (mg) the utilisation rate of shared parking lots collectively and
24 1.016 × 106 9.65 × 105 5.01 not individually. The constant number of shared parking lots
28 1.016 × 106 9.44 × 105 7.08 each with 2 spots satisfy the maximum demand possible.
32 1.016 × 106 9.16 × 105 9.80 But when there are no vacant spots within the range of the
driver’s preference, the rerouting occurs. This rerouting also
contributes to maintaining the constant utilisation of parking
1) Description of comparison analysis plots: In Figure 12 lots.
and Figure 13, utilisation and revenue generated are com- 3) Change in utilisation of shared parking lots: Let us now
pared between the scenarios with base model (without shared take the perspective of the single parking lot. For this purpose,
parking lots) and shared model (with shared parking lots) 3 shared parking lots are considered. The pattern by which
in Figure 12 and Figure 13. Due to varying number of these parking lots are utilised varies for each simulation run,
shared private parking lots, there are three sets of comparisons thus generating many possible configurations. In Figure 14 we
between same base model scenario but different shared model show few patterns of the utilization of the three parking lots.
scenarios (i.e. from top to bottom in each comparison set, From the plots in Figure 14, the important inference is that
shared model with 24, 28 and 32 shared private parking lots). each of the parking lots is guaranteed to get utilized at least
2) Comparison analysis between commercial and shared twice (Parking Lot 3 in (a) and Parking Lot 1 in (c)). This is
parking lots in terms of utilisation rate and revenue: Figure 12 guaranteed by the allocation module using Rule 51. In the case
compares the the base model and the shared model in terms of of other parking lots, at some time of the day, the utilisation
utilisation rate. In Figure 12, each block has a set of line plots exceeds 2 and ranges between 2 to 3. This is due to their
with utilisation rate in y-axis and time of day in x-axis. The location and also to the destination of the vehicles at that
set of plots include utilisation of different commercial parking time. This contributes evidence to the fact that shared private
lots along with combined utilisation of many shared private parking lots are utilized evenly. Also, the presence of such
parking lots. ILAC parking lot is a commercial parking lot in shared parking lots decreases the dependency on commercial
North-Western part of Dublin. From Figure 12, ILAC parking parking lots as visible in Figures 12 and 14).
lot (plotted in green in each block of Figure 12) is utilized
more than that of the other commercial and shared parking E. Effects of the number of available parking lots
lots in the proximity. This is due to its location being in a Lastly, we evaluate the effects of the ratio between shared
strategic point that makes it more preferred than any other. parking lots and commercial parking lots available. The results
The allocation module (equipped with CA) is intended are from the driver’s and the parking lot owner’s perspective.
to make sure that the demand is evenly distributed among This collection helps us to correlate how the variation of the
the parking lots in the shared parking model (including the ratio could benefit the shared private parking model. From the
commercial parking lots). But in the process, due to the simulation one can identify the optimal number of parking
strategic location of some parking lots and the vehicle driver’s lots needed for the benefit of the stakeholders. Also, it further
end destination, the strategically located parking lots are helps to understand how to extend the model from a part of
preferred more than others. This is evident from the maximised a city to a full city-based scenario such as for the MoST.
utilisation of ILAC parking lot. We analyse how shared parking The parameter Vehicular Emission denotes the effects on the
lots are utilized when compared to commercial parking lots. vehicles to changes in the number of shared parking lots from
With the increase in the number of shared private parking lots, a driver’s perspective. The parameters Redirects and Supply to
the demand satisfied by the shared parking lots increases. This demand denote how the model with a certain number of shared
in turn decreases the utilisation of the commercial parking private parking lots react to the demand (number of vehicles)
lots. The presence of shared parking lots contributes to a at a particular hour. Figure 15 shows the results of correlation
considerable drop in the utilisation of commercial parking lots, analysis in the form of a heatmap.The gradient of correlation
especially ILAC. This is evident from Figure 12. of different parameters with each other is represented by
In Figure 13, each block has a set of line plots with heatmap. The gradient is represented by numerical variation
revenue in y-axis and time of day in x-axis. The set of plots between -1 and 1 and also, in the form of coloring scheme
include revenue generated by different commercial parking from dark green to dark red in Figure 15.
lots along with combined revenue generated by many shared 1) Relationship between ratio and vehicular emission:
private parking lots both in base model and shared model. Figure 15 shows that the ratio and the mean vehicular emission
When considering collectively, the overall revenue generated of all the vehicles in the simulation are moderately correlated
Fig. 12. Results of analysing the utilisation of different commercial parking lots and shared parking lots, collectively
Fig. 13. Results of analysing the revenue of different commercial parking lots and shared parking lots, collectively
with a correlation of -0.32. This correlation can be best shared parking lots, more vehicles utilize the shared parking
described as with the increase in the number of shared parking lots rather than resorting to the commercial parking lots.
lots, there is a moderate decrease in vehicular CO2 emissions.
From the driver’s perspective, this means that the vehicular 3) Relationship between ratio and redirects: The term
emission due to additional cruise time to search for parking redirects indicates the number of vehicles that are not served
spot reduces considerably. by the shared parking model. The correlation heat map shows
2) Relationship between ratio and supply to demand: the that the correlation coefficient is -0.5. The conclusion is that
term supply to demand denotes the number of reservations as the number of shared parking lots increases, the number
made by the vehicles in the shared parking lots instead of the of vehicles not served decreases. In the present simulation,
commercial parking lots. There is a strong positive correlation varying the number of shared parking lots, the best-suited
of 0.82 that shows that with the increase in the number of value is 32.
Fig. 14. Utilisation of individual parking lots
Fig. 15. Results of analysing the effects of change in number of shared parking lots
VI. C ONCLUSIONS vate parking model, required number of framework researches
and trial runs. The design, implementation and evaluation
Smart cities try to optimize the way in which the shared via simulations show the high potential benefits of the smart
infrastructure is used by its citizens, opening to the possibility shared parking model in terms of CO2 emission reductions,
of people actively participating resource utilization. The work decrease of driven distance, and economic compensation for
proposed here provides a fully worked solution on how to private parking owners. Finally, an important outcome of the
address the problem of parking scarcity. Additional benefits work is that the results demonstrate how the change in the
of sharing private parking spots in terms of environmental and number of shared parking lots affects the model itself and the
economic benefits are also identified. We proposed a setup of a participants of the shared parking model.
simulation-based testbed to realize and analyze the shared pri-
A. Limitations transportation. This could further reduce the congestion in the
Limitations of the present study include the following ones. considered area. Accordingly, there will be some liabilities
The scalability of the setup ensures that the applicability to for the car drivers as they use the property of the owner.
different demand data varying by each hour. But the simulation The mathematical models used in the work proposed could
has to be re-initiated for each hour. This is due to the fact that be extended in order to evaluate the effects of the change in
the retention of the number of free spots (of each parking terms of metrics discussed in the work or by adding additional
lot) after respective vehicles resumed is not certain. This metrics.
uncertainty is due to possible communication failures of the R EFERENCES
resume signal of the vehicles.
[1] A New Kind of Science. Last accessed: 2021-03-18.
The parking and overtime parking behaviour of each vehicle
Champaign, Ilinois, USA: Wolfram Media Inc., 2002.
is modelled in the simulation as random entry. The random
ISBN : 1579550088.
entry is chosen by each vehicle from the set of defined
[2] Muhammad Abdullah et al. “Exploring the impacts of
entries (taken care of by the application deployed in each
COVID-19 on travel behavior and mode preferences”.
vehicle involved in the simulation). For example, for vehicle
In: Transportation Research Interdisciplinary Perspec-
A, the parking duration and the resuming duration are chosen
tives 8 (2020), p. 100255. ISSN: 2590-1982. DOI: https:
randomly from the defined set as 30 minutes and 35 minutes,
/ / doi . org / 10 . 1016 / j . trip . 2020 . 100255. URL:
respectively. Though the duration pair is a fixed one from the
https : / / www . sciencedirect . com / science / article / pii /
defined set, the method of choosing a pair from the set is
S2590198220301664.
random. There is a need to model the driver’s behaviour to
[3] Muhammad Alam et al. “Real-Time Smart Parking Sys-
park overtime and incur additional charges due to overtime
tems Integration in Distributed ITS for Smart Cities”.
parking which has not been addressed in the work proposed.
Apart from modelling the parking behaviour externally, the In: Journal of Advanced Transportation 2018 (Oct.
vehicle resume activity needs to be triggered externally to re- 2018), p. 1485652. ISSN: 0197-6729. DOI: 10 . 1155 /
flect the externally modelled behaviour. There are few standard 2018 / 1485652. URL: https : / / doi . org / 10 . 1155 / 2018 /
metrics to analyse a shared parking model. In the work pro- 1485652.
posed, the beneficial perspectives of the model of a vehicle and [4] Bosch Mobility Solutions. Bosch Community Park-
a parking lot are quantified by vehicular emission, utilisation ing. Last accessed: 2021-03-18. 2016. URL: https : / /
rate and revenue. But, from a smart city’s perspective, some www . bosch - mobility - solutions . com / en / products -
additional standard metrics could be added. This will pave the and - services / mobility - services / connected - parking /
way for more standard ways of evaluating a shared parking community-based-parking/.
model (for example quantifying traffic congestion created due [5] Mitava Chaturvedi, M.H.P. Zuidgeest, and M.J.G. Brus-
to parking searches). sel. “Parking in balance: a geospatial analysis of effi-
Finally, the messages sent between various components in ciency of parking system”. English. In: 30 slides ; India
the simulation needs a standard. There are accepted standards Geospatial Forum 2013 ; Conference date: 22-01-2013
of messages based on priority. In the work proposed, the Through 24-01-2013. Jan. 2013, s1–s30.
prioritisation of the parking request and response messages [6] Lara Codeca and Jérôme Härri. “Monaco SUMO Traffic
is not addressed. Hence, an evaluation of how other messages (MoST) Scenario: A 3D Mobility Scenario for Cooper-
related to co-operative mobility get along with the messages ative ITS”. In: SUMO 2018, SUMO User Conference,
related to service requests and responses is in order. Simulating Autonomous and Intermodal Transport Sys-
tems, May 14-16, 2018, Berlin, Germany. B, May 2018.
B. Future work [7] L. Codecá, J. Erdmann, and J. Härri. “A SUMO-
The work proposed has a microscopic area as the Based Parking Management Framework for Large-Scale
simulation-based testbed. One can extend it to a large city as Smart Cities Simulations”. In: 2018 IEEE Vehicular
the testbed for assessing the benefits of the smart shared private Networking Conference (VNC). 2018, pp. 1–8. DOI: 10.
parking model. The work proposed can be extended to Hard- 1109/VNC.2018.8628417.
ware In Loop (HIL) simulation or to a digital testbed. This [8] Docker. Docker. en. Last accessed: 2021-03-18. URL:
helps to address the realistic uncertainties that go unaddressed https://docker.com/.
due to the software simulation (e.g., hardware or firmware [9] Ma Dongfang et al. “A Distribution Model for Shared
issues of components involved in V2X communication). The Parking in Residential Zones that Considers the Uti-
shared parking model is present as a single, centralized web lization Rate and the Walking Distance”. In: Journal
service. This setup could be made as a distributed architecture of Advanced Transportation 2020 (2020). ISSN: 0197-
with proper synchronisation of the distributed data pushing the 6729. DOI: https://doi.org/10.1155/2020/6147974. URL:
logic to the periphery of the network in an Edge Computing https://www.hindawi.com/journals/jat/2020/6147974/.
fashion. Apart from considering the preferred walking distance [10] Dublin Traffic Counts. https : / / data . smartdublin . ie /
of the car driver, one could also consider adding other forms dataset/traffic-volumes. Accessed: 2021-03-18.
of travel options, such as a bike, an electric scooter, or public
[11] Fraunhofer FOKUS and DCAITI (Daimler Center for [24] OpenSIM. omnet++. en. Last accessed: 2021-03-18.
Automotive IT Innovations). Eclipse MOSAIC. Last URL : https://omnetpp.org/.
accessed: 2021-03-18. 2020. URL: https://www.eclipse. [25] Christoph Sommer, Reinhard German, and Falko
org/mosaic/. Dressler. “Bidirectionally Coupled Network and Road
[12] Wesley Fung. “VANET: Evaluating Smart Parking Per- Traffic Simulation for Improved IVC Analysis”. In:
formance in a Dublin Scenario”. MA thesis. Dublin: IEEE Transactions on Mobile Computing (TMC) 10.1
University of Dublin, Trinity College, May 2017. (Jan. 2011), pp. 3–15. DOI: 10.1109/TMC.2010.133.
[13] Georgia Institute of Technology. Georgia Institute of [26] Zenpark. Zenpark. Last accessed: 2021-03-18. 2016.
Technology. Last accessed: 2021-03-18. 2019. URL: URL : https://www.zenpark.com.
https : / / www . cc . gatech . edu / computing / pads /
transportation/testbed/description.html.
[14] Google. Google Maps. en. Last accessed: 2021-03-19.
URL: https://maps.google.com.
[15] HBEFA. Handbook of Emission Factors for Road
Transport (HBEFA). Last accessed: 2021-03-18. 2004.
URL: https://www.hbefa.net/e/index.html.
[16] Marcel Hellkamp. Bottle: Python Web Framework. en.
Last accessed: 2021-03-18. URL: https://bottlepy.org/.
[17] Xin Huang et al. “Research on parking sharing strate-
gies considering user overtime parking”. In: PLOS ONE
15.6 (June 2020), pp. 1–22. DOI: 10.1371/journal.pone.
0233772. URL: https://doi.org/10.1371/journal.pone.
0233772.
[18] Kanchan, Pragati et al. “Real-time Location Based
Shared Smart Parking System”. In: E3S Web Conf. 170
(2020), p. 03003. DOI: 10.1051/e3sconf/202017003003.
URL: https://doi.org/10.1051/e3sconf/202017003003.
[19] Lap Kong Law, Srikanth V. Krishnamurthy, and
Michalis Faloutsos. “Understanding the Interactions be-
tween Unicast and Group Communications Sessions in
Ad Hoc Networks”. In: Mobile and Wireless Commu-
nication Networks. Ed. by Elizabeth M. Belding-Royer,
Khaldoun Al Agha, and Guy Pujolle. Boston, MA:
Springer US, 2005, pp. 1–12. ISBN: 978-0-387-23150-
1.
[20] Pablo Alvarez Lopez et al. “Microscopic Traffic Sim-
ulation using SUMO”. In: The 21st IEEE Interna-
tional Conference on Intelligent Transportation Sys-
tems. IEEE, 2018. URL: https://elib.dlr.de/124092/.
[21] Dennis Luxen and Christian Vetter. “Real-time rout-
ing with OpenStreetMap data”. In: Proceedings of the
19th ACM SIGSPATIAL International Conference on
Advances in Geographic Information Systems. GIS ’11.
Chicago, Illinois: ACM, 2011, pp. 513–516. ISBN: 978-
1-4503-1031-4. DOI: 10.1145/2093973.2094062. URL:
http://doi.acm.org/10.1145/2093973.2094062.
[22] Mobypark B.V. Mobypark B.V. Last accessed: 2021-03-
18. 2016. URL: https://www.mobypark.com.
[23] Julio C. Navas and Tomasz Imielinski. “Geo-
Cast—Geographic Addressing and Routing”. In: Pro-
ceedings of the 3rd Annual ACM/IEEE International
Conference on Mobile Computing and Networking. Mo-
biCom ’97. Budapest, Hungary: Association for Com-
puting Machinery, 1997, pp. 66–76. ISBN: 0897919882.
DOI : 10.1145/262116.262132. URL : https://doi.org/10.
1145/262116.262132.