Smart Parking Assistance Services and User Acceptance: A European

Model

Eleni G. Mantouka

, Foteini Orfanou

Martin Margreiter

, Eleni I. Vlahogianni

and Javier J. Sanchez Medina

Dep. of Transportation Planning and Engineering, National Technical University of Athens,

5 Iroon Polytechneiou Str, Zografou Campus, 157 73 Athens, Greece

Chair of Traffic Engineering and Control, Technical University of Munich, Arcisstrasse 21, 803 33 Munich, Germany

CICEI, DIS, ULPGC, Las Palmas de Gran Canaria, 35017 Las Palmas, Spain

KeyWords: Parking Sensors, Parking Assistance, Survey, User Constraints, User Acceptance.

Abstract: Technologies and systems assisting drivers to locate free on street parking space and/or inform on parking

availability may significantly reduce the traffic induced from cruising for parking space in cities. This paper

attempts to reveal the factors that may affect the acceptability of parking assistance systems in different

European cities, based on data collected through a questionnaire survey. The respondents are presented with

a real world parking assistance system based on in-vehicle ultrasonic sensors, which detects free parking space

in real time, and are, then, asked to respond to a set of questions in relation to their parking choice preferences.

The results of the survey are presented and modelled using a genetically optimized Logistic Regression Model.

Findings indicate that the proposed system would be useful for people who are not willing to spend too much

time in order to find an available parking space as well as to those who are not willing to walk long distances

from the parking place to their final destination. Moreover, results revealed that the certainty of the provided

recommendation significantly influences the effect of the other parameters on the acceptability of the

application. Finally, some further research steps are discussed.

1 INTRODUCTION

With the increase in demand for traveling in the cities,

the demand for parking areas increases leading drives

to circulate inside urban area (cruising) in search for

a parking space. Cruising for a parking space leads to

increased fuel consumption and induced traffic

congestion (Shoup, 2006, Arnott and Inci, 2006). A

study revealed that in specific time periods of a day

cruising for parking space may account for 50% of

the traffic (Shoup, 2006). In a different study,

researchers estimated that cruising for parking may

increase commuting by approximately 20% (Van

Ommeren et al., 2011). Some researchers estimated

the share of traffic that is cruising for parking using

https://orcid.org/0000-0002-3471-5966

https://orcid.org/0000-0002-3503-592X

https://orcid.org/0000-0002-0428-0914

https://orcid.org/0000-0002-2423-5475

https://orcid.org/0000-0003-2530-3182

data from video sensors (Hampshire and Shoup,

2018). They resulted that 15% of the traffic was

cruising for parking. On the other hand, other

researchers estimated that vehicles searching for

parking consist of approximately 30% of the overall

traffic (Dowling et al., 2017).

In order to tackle the negative impacts of cruising

for a parking space, several attempts based on

theoretical urban economics to regulate parking

prices in and out of city centres (Arnott and Rowse,

1999) have been conducted. Several studies

investigated the effect of parking policies on traffic

management using simulation of real-world data

(Shiftan and Burd-Eden, 2001, Chatman and

Manville, 2014).

Mantouka, E., Orfanou, F., Margreiter, M., Vlahogianni, E. and Medina, J.

Smart Parking Assistance Services and User Acceptance: A European Model.

DOI: 10.5220/0007727404910497

In Proceedings of the 5th International Conference on Vehicle Technology and Intelligent Transport Systems (VEHITS 2019), pages 491-497

ISBN: 978-989-758-374-2

491

Literature for long has indicated that the time

spent by drivers searching for a parking space is

related to the availability of parking information

(Ahangari et al., 2018). This availability is related to:

i. real time parking space monitoring infrastructure,

ii. the ability to produce crowdsourced information on

free street parking space and iii. the provision of

advanced parking assistance systems. Some

prominent examples of monitoring parking spaces

using underground magnetic sensors, such as the

FASTPRK project in Moscow (50000 sensors)

(Worldsensing, 2019) and SFpark project in San

Francisco (8200 sensors in 2014) (Chatman and

Manville, 2014). A dynamic approach, entailing the

collection of data from a moving unit either vehicle

or drone can be also a potential solution (D’Aloia et

al., 2015, Golias and Vlahogianni, 2018). These

approaches are established on deep learning and other

advanced algorithms for parking space identification

and availability prediction (Vlahogianni et al., 2015,

Monteiro and Ioannou, 2018, Golias and

Vlahogianni, 2018).

Various parking assistance systems have been

developed and used worldwide in order to reduce the

existing parking problems and to improve the

efficient use of the existing parking supply. Smart

parking systems are already used for defining parking

occupancy, parking guidance information and,

parking facility management while different

technologies are used for parking space detection like

inductive loops, infrared sensors, magnetometers,

ultrasonic detectors, radar sensors, etc. (Revathi and

Dhulipala, 2012, Fraifer and Fernström, 2016,

Faheem et al., 2013, Lin et al., 2017). All the parking

systems aim at assisting drivers while cruising for

parking and make the parking search less time

consuming and easier.

But, how influential is the provision of parking

information to the users’ parking behaviour? Past

studies have underlined that the time seeking for a

parking space appears to be critical in the parking

choice behaviour (Ibeas et al., 2014). Other

researchers found that, in addition to parking fee,

search time and access time, a risk-averse attitude and

a positive car care (maintenance) attitude are

determinants for parking choice (Soto et al., 2018).

The importance of parking assistance systems has

been studied using a stated preference survey and a

driving simulator to evaluate the effects of different

types of information related to parking space

availability on parking choice and circulation

behaviour (Ahangari et al., 2018). The study revealed

that age and parking availability information affect

parking choice behaviour.

The scope of this paper is to evaluate the

acceptance of parking assistance systems in a

European level based on a questionnaire survey. The

survey is based on a real-world novel parking

assistance technology based on ultrasonic sensors

installed in the vehicle, which detects free parking

spaces and provides information to the drivers about

parking space availability near their destination. The

survey is conducted in order to reveal users’

expectations and needs from such a tool and also to

present and analyse their willingness to use it. Finally,

a genetically optimized Logistic Regression Model is

applied in order to reveal the factors affecting users’

acceptability of such a parking assistance tool and

their willingness to integrate it in their everyday life.

2 ACCEPTANCE OF SMART

PARKING ASSISTANCE

SYSTEMS

2.1 The Reference System

The proposed parking assistance system uses

ultrasonic sensors installed in the vehicle and enables

the detection of free gaps on the right and left side of

a street when an equipped vehicle passes by. The

ultrasonic system is able to detect the complete scene

with its limiting vehicles, curb stone information,

length and depth of the detected gap and other

additional attributes. The concept of the proposed

technology is illustrated in Figure 1.

Figure 1: Ultrasonic parking place detection principle.

The data is transmitted to a backend server and

based on historic information and a parking area map

the detected gap is classified either as a parking space

or a non-parking space (driveways, exits, etc.). More

specifically, the parking place detections resulting out

of driveways etc. are filtered out whereas other

parking places are validated as real parking spaces

where parking is indeed allowed (Margreiter et al.,

VEHITS 2019 - 5th International Conference on Vehicle Technology and Intelligent Transport Systems

492

2017). The proposed technology aims not only at

detecting free parking spots but also at predicting

parking space availability around the destination of

the driver and for a certain arrival time. This

challenge is solved by using historic occupancy data

which is updated by real time data collected from

available transmitter vehicles in the region. Through

this system the users can be informed about free

parking space availability and existence at any time

near their destination and be guided towards the free

parking spot without having to drive around the same

blocks. More details about the technology and its

services can be found in Margreiter et al. (2015),

Margreiter et al. (2017) and Orfanou et al. (2017).

2.2 The Survey

Beside the continuously increase of smart parking

assistance systems, the success of such a system is

based on its acceptance from the potential users

(drivers). For this reason, a survey was conducted to

get more detailed information about driver’s parking

behaviour and habits. Moreover, the respondents had

to give answers related to their needs and their

expectations from a parking assistance system as well

as if and under which circumstances, they are willing

to accept and use such a system.

The National Technical University of Athens and

the Technical University of Munich conducted a

survey in order to investigate how willing the drivers

are to use such an application described in the

previous section. The two research teams collected

more than 500 answers and the respondents were

coming from different European cities developing a

cultural and habitual diversity among the participants.

Due to the fact that parking assistance technologies

should address the needs of drivers coming from

different countries and cities it is important to detect

the differences in their parking habits, needs and

constraints so that the system can meet and satisfy the

expectations of all potential users.

The distributed questionnaire was divided in three

parts (Table 1). In the first part, the participants had

to give answers related to their parking behaviour. In

the second part, an overview and brief description of

the new developed technology was given, and the

potential users revealed their expectations from such

a technology, as well as their willingness to use it. The

last part contained some personal information about

the gender and age of the respondents.

Table 1: Questionnaire’s part description.

Part

Description

Questions Content

Parking

Behaviour

Time searching, Distance

searching, detour,

parking search strategy,

walking distance, parking

facilities

Proposed

system:

Characteristics

and

Acceptance

Type of information,

willingness to use,

desired accuracy of

information, evaluation

of system’s features

Socio-

demographics

Gender, age, Car

ownership, city of

residence

The age and gender distribution in the sample is

presented in Figures 2 and 3. Overall, 66% of the

responders are men and 34% women, while the

majority of them belong to the age group 26-35 years

old.

Figure 2: Gender distribution in the sample.

Figure 3: Age distribution in the sample.

As mentioned before, here it is attempted to

analyse parking behaviour and the willingness to use

a parking assistance system of people living in several

European cities. To this end, the sample includes

citizens of the biggest cities of Greece, Germany,

Austria and Switzerland, while there are some

respondents from Italy, Portugal and Cyprus but they

are underrepresented. In addition, several

characteristics of each city were included with their

corresponding levels as shown in the Table 2.

66%

34%

Male Female

16%

62%

11%

18-25 26-35 36-45 46-55 >55

Smart Parking Assistance Services and User Acceptance: A European Model

493

Table 2: City attributes and the corresponding levels.

Attributes

Levels

Population

<100000

<500000

>500000

Area

<50km

<200 km

>200 km

Access to the

internet

High

(>75%)

Medium

(>50%)

Low

(<50%)

Usage of

goods and

services that

are obtained

through the

internet

High

(>75%)

Medium

(>50%)

Low

(<50%)

Frequency of

internet use

High

(>75%)

Medium

(>50%)

Low

(<50%)

Data processing procedures included the exclusion of

incomplete questionnaires as well as other fault or

malicious answers, such as household size greater

than 20. The final sample size includes a total number

of 374 questionnaires.

3 FINDINGS

3.1 Preliminary Assessment of

Responses

A preliminary assessment of the responses was

conducted first. Results showed that 51% of the

sample spent less than 5 minutes to find a free parking

space while only 4% of the sample have to spend

more than 20 minutes. As far as it concerns the time

drivers are willing to spend on finding a free parking

space results are almost the same, as shown in Figure

Figure 4: Time spent and time drivers are willing to spend

on finding a free parking space.

Moreover, the vast majority of the sample are not

willing to walk more than 1km from their car to their

final destination, while in addition 54% of the sample

stated that their tolerable distance between the

parking place and their final destination is less than

400m.

Furthermore, most of the drivers (70 %) prefer to

park on the street, rather than to park in a public

(18%) or private (12 %) parking garage. By means of

parking strategies, most of the drivers (74 %) prefer

to search for a free parking space between several city

blocks while only 19% of the drivers try to find a

parking space by driving within the same block. The

rest of the sample prefers to either wait passively at

the same place for a space to be free or drive only

along a particular road with intermediate reversals.

These results also highlight the need for an

application which provides information on the

availability of parking places on the street as the one

described in the present work.

Figure 5 presents the importance of several

aspects of the proposed application as they were rated

by the respondents. Findings revealed that the two

most important aspects of the application for the

majority of the sample are the provision of

information about the type of the available parking

place (e.g. place for habitants only, place for people

with disabilities) and the saving of time when

searching for a parking space. In addition, of great

importance is the depiction of the exact location of

the available parking space as well as the ability of

the app to recommend the parking places which are

close enough to the final destination. On the other

hand, results revealed that the least important aspect

of the application is the ability to book a parking

space on the road.

Figure 5: Distribution of answers concerning the

importance of the aspects of the application.

51%

45%

54%

40%

0-5 minutes 6-20 minutes More than 20

minutes

Actual time Time willing to spent

50%

100%

Not

important at

all

Relatively

unimportant

Neutral Important Very

important

Comparison between the length of the free

parking space and the length of user's car

Allow the user to book the free position that has

been detected

Depiction of the exact location of the free parking

space

Shortest distance between parking space and final

destination

Less time searching for a parking space

VEHITS 2019 - 5th International Conference on Vehicle Technology and Intelligent Transport Systems

494

Interestingly, only 23% of the drivers require that

the information on the availability of parking is at

least 90% certain in order to be provided through the

application, while the rest of them would like to

receive this information although may be not so valid.

More specifically, almost 60% of the sample wants to

receive this information when this information is 70

– 90% valid that the detected gap corresponds to a

free parking space instead of an exit or a driveway.

Finally, respondents were asked to select between

three alternatives concerning the kind of information

that they want to receive. In terms of the

recommendation of a free parking space the

application may:

1. Suggest an eventually longer route to the

destination with a higher availability of free

parking places (suggestion of the appropriate

route).

2. Depict the availability of free parking places

for all city blocks in the area around the

destination.

3. Recommend one specific free parking place

close to the destination.

The corresponding results are shown in Figure 6. The

responses show that users will prefer the much more

realistic alternative for the system, which is to depict

the availability of free parking places for all city

blocks in the area around the destination. This may be

due to the fact that they are used to rely on experience

and do not necessarily need the system to point to a

specific free parking space.

Figure 6: Distribution of the sample between the three

alternatives concerning the type of provided information on

the availability of parking spaces.

3.2 A European Model of Parking

Assistance Systems Acceptance

In this section, the most significant variables that

affect the acceptance of the proposed system are

being discussed. For this purpose, a Genetically

Optimized Logistic Regression model was developed

taking into consideration the interactions between

independent variables. The search for the best model,

especially in cases where there are many predictors to

be considered of different form, is exhaustive and

time consuming. Using nature-inspired meta-

heuristics, such as genetic algorithms, swarm

optimization etc., for big size problems makes the

searching for the near optimum solution much more

computationally efficient and flexible when

compared to exact optimization algorithms

(Vlahogianni et al., 2014). In this paper, the

genetically based searching algorithm of Calcagno

and de Mazancourt (2010) is implemented, which is

based on the Yang’s enhanced genetic search operator

with an immigration function to improve

convergence for complex problems (Yang, 2004).

The dependent variable in the model is a binary

variable describing the acceptability of the system,

taking the value 1 if the respondent is willing to use

the application and the value 0 otherwise. As

independent variables are used the demographics of

the users (gender, age, car ownership) as well as

drivers’ habits and strategies when searching for a

free parking space. Regarding the search algorithm,

the fitness function w

for the i

model is given by:





 















(1)

where QAICc

is the information metric for the i

model and QAICc

best

the information metric for the

best model in the population of models (Burnham and

Anderson, 2003). As a note, higher QAICc

means

lower fit.

Results are summarized in the table below (Table

3). The accuracy of the model is 76%, while the

precision is 73.7%.

The estimated coefficients of the regression model

indicate to what extent the acceptance of the

application changes when different parameters

(independent variables) change. In the case where

pairwise interactions are being considered, the

estimated coefficients describe the simultaneous

influence of the two variables to the dependent

variable.

According to the results, male drivers are more

likely to accept a smart parking assistance system and

more specifically younger ones. Furthermore, the

actual time that the driver usually spends to find a free

parking space has a positive impact on the

acceptability of the system. The interaction effect

indicates that this effect is greater for people who are

willing to walk longer from the parking to their final

destination and is lower for those who usually park in

a public parking area.

Moreover, drivers who are willing to either drive

longer in order to find a free parking space or walk

longer from this point to their final destination are not

35%

56%

Alternative 1 Alternative 2 Alternative 3

Smart Parking Assistance Services and User Acceptance: A European Model

495

willing to use the proposed system. The willingness

to use the app is even lower when the application

provides low-certainty feedback and

recommendations. On the other hand, if the certainty

of the system’s recommendation about the parking

availability is greater those who usually spend more

time when searching for a parking space are more

likely to use the application.

Table 3: Estimates of the Logistic Regression model of the

acceptance of the application.

Variables

Estimate

Pr(>|t|)

Sign.

Private parking areas

-0.846

0.007

Gender

(Male=0/Female=1)

-0.228

0.076

Actual time searching

for parking * Max

walking distance

-0.141

0.030

Max walking distance

* Maximum distance

willing to travel

0.200

0.001

***

Actual time searching

for parking *

Certainty

0.083

0.05

Maximum distance

willing to travel *

Certainty

-0.134

0.001

Area* Maximum

distance willing to

travel

-0.078

0.048

Internet access*

Frequency of detour

-0.085

0.006

Internet access *

Goods & Services

0.198

0.001

Internet access * Age

-0.172

0.000

***

Gender * Age

0.099

0.07

Public parking areas *

Actual time searching

for parking

-0.239

0.09

Private parking areas

* Actual time

searching for parking

0.384

0.03

Private parking areas

* Area

0.186

0.033

0 ‘***’ , 0.001 ‘**’ , 0.01 ‘*’ , 0.05 ‘.’ , 0.1 ‘ ’

Additionally, people who are willing to travel

large distances in order to find a free parking space

are less likely to use such an assistance system. The

impact of this factor is lower when the city where they

live is larger.

Finally, as expected, the level of modernization of

the city and the familiarity of its population with the

internet and other smartphone applications is of great

importance for the acceptance of smart assistance

systems as the one proposed. More specifically,

people who live in cities where the majority of the

population has access to the internet are willing to use

the proposed application. The impact of this factor is

greater when most of the population uses goods and

services through the internet and is lower for the

elderly.

4 CONCLUSIONS

In this paper, we investigated the factors that affect

the acceptance of smart parking assistance systems.

The analyses were conducted on data collected

through a questionnaire survey, which referred to a

real-world parking collection data scheme and

associated assistance services. The answers of over

500 respondents were modelled using a genetically

optimized Logistic Regression model, which was

trained to predict the probability of accepting the

specific parking assistance system by taking into

account a large set of predictors and their interactions.

Results indicate that the proposed application

would be useful for people who are not willing to

spend too much time in order to find an available

parking space as well as to those who are not willing

to walk long distances from the parking place to their

final destination. Furthermore, findings revealed that

the certainty of the provided recommendation

significantly influences the effect of the other

parameters on the acceptability of the application.

Finally, results show that younger males are more

likely to use such an application.

The above findings may provide a comprehensive

view on the characteristics of potential users of the

proposed system. Nevertheless, future research

should focus on a more in-depth analysis of users

parking behaviour and city’s attributes on the

acceptance of such a system. Finally, further research

should also investigate the characteristics of the users

who are willing to pay for such a system.

REFERENCES

Ahangari, S., Chavis, C., Jeihani, M., Moghaddam, Z. R.

(2018). Quantifying the Effect of On-Street Parking

Information on Congestion Mitigation using a Driving

Simulator. Transportation Research Record,

0361198118773893.

Arnott, R., Rowse, J. (1999). Modeling Parking. Journal of

Urban Economics, 45(1), pp. 97-124.

Arnott, R., Inci E. (2006). An Integrated Model of

Downtown Parking and Traffic Congestion. Journal of

Urban Economics, 60(3), pp.418–442.

Burnham, K. P., Anderson, D. R. (2003). Model selection

and multimodel inference: a practical information-

VEHITS 2019 - 5th International Conference on Vehicle Technology and Intelligent Transport Systems

496

theoretic approach. Springer Science & Business

Media. Second edition. Springer: New York.

Calcagno, V., de Mazancourt, C. (2010). glmulti: an R

package for easy automated model selection with

(generalized) linear models. Journal of statistical

software, 34(12), pp. 1-29.

Chatman, D. G., Manville, M. (2014). Theory Versus

Implementation in Congestion-priced Parking: An

Evaluation of SFpark, 2011-2012. Research in

Transportation Economics, 44, pp. 52–60.

D’Aloia, M., Rizzi, M., Russo, R., Notarnicola, M. N.,

Pellicani, L. (2015). A marker-based image processing

method for detecting available parking slots from

UAVs, New Trends Image Analysis Processing. Cham,

Switzerland: Springer, pp. 275–281.

Dowling, C., Fiez, T., Ratliff, L., Zhang, B. (2017). How

Much Urban Traffic is Searching for Parking?

Simulating Curbside Parking as a Network of Finite

Capacity Queues. arXiv preprint arXiv:1702.06156.

Faheem, F., Mahmud, S.A., Khan, G.M., Rahman, M.,

Zafar, H. (2013) A Survey of Intelligent Car Parking

System, Journal of Applied Research and Technology,

11(5), pp. 714-726.

Fraifer, M., Fernström, M. (2016). Investigation of smart

parking systems and their technologies. 37

International Conference on Information Systems. IoT

Smart City Challenges Applications (ISCA 2016),

Dublin, Ireland, pp. 1-14.

Golias, K., Vlahogianni, E. I. (2018). Convolutional Neural

Networks for On-Street Parking Space Detection in

Urban Networks. IEEE Transactions on Intelligent

Transportation Systems.

Hampshire, R. C., Shoup, D. (2018). What share of traffic

is cruising for parking?. Journal of Transport

Economics and Policy (JTEP), 52(3), 184-201.

Ibeas, A., dell’Olio, L., Bordagaray, M., Ortúzar, J.D.

(2014). Modelling parking choices considering user

heterogeneity. Transportation Research Part A: Policy

and Practice, 70, pp. 41–49.

Lin, T., Rivano, H., Le Mouël, F., Baron, B., Spathis, P.,

Rivano, H., Voge, M. E. (2017). A survey of smart

parking solutions, IEEE Transactions on Intelligent

Transportation Systems, 18(12), pp. 3229-3253.

Margreiter, M., Mayer, P., Orfanou, F. (2015). A Concept

for Crowdsourcing of In-vehicle Data to Improve

Urban On-street Parking, mobil.TUM International

Scientific Conference on Mobility and Transport

“Technologies, Solutions and Perspectives for

Intelligent Transport Systems”, June 2015.

Margreiter, M., Orfanou, F., Mayer, P. (2017).

Determination of the parking place availability using

manual data collection enriched by crowdsourced in-

vehicle data, Transportation research procedia, 25, pp.

497-510.

Monteiro, F.V., Ioannou, P. (2018) On-Street Parking

Prediction Using Real-Time Data, 21

International

Conference on Intelligent Transportation Systems

(ITSC), November 4-7, Maui, Hawaii, USA.

Orfanou, F., Margreiter, M., Mayer, Ph., Alpas, M.,

Vlahogianni E. (2017). Driver’s Willingness to Use

Parking Assistance Tools and their Expectations: A

Case Study for the Cities of Munich and Athens, 8

International Congress on Transportation Research in

Greece “Transportation by 2030: Trends and

Perspectives”, Thessaloniki, Greece.

Revathi, G., Dhulipala, V. S. (2012). Smart parking systems

and sensors: A survey. Computing, Communication and

Applications (ICCCA), 2012 International Conference,

pp. 1-5, IEEE.

Shiftan, Y., Burd-Eden, R. Modeling Response to Parking

Policy (2001). Transportation Research Record:

Journal of the Transportation Research Board, 1765,

pp. 27–34.

Shoup, D. C. (2006). Cruising for Parking. Transport

Policy, 13(6), pp. 479–486.

Soto, J. J., Márquez, L., Macea, L. F. (2018). Accounting

for attitudes on parking choice: An integrated choice

and latent variable approach. Transportation Research

Part A: Policy and Practice, 111, pp. 65-77.

Van Ommeren, J., Wentink, D., Dekkers, J. (2011). The

Real Price of Parking Policy. Journal of Urban

Economics, 70(1), pp. 25–31.

Vlahogianni, E. I., Karlaftis, M. G., Golias, J. C. (2014).

Short-term traffic forecasting: Where we are and where

we’re going. Transportation Research Part C:

Emerging Technologies, 43, 3-19.

Vlahogianni, E. I., Kepaptsoglou, K., Tsetsos, V., Karlaftis,

M.G. (2015). A Real-Time Parking Prediction System

for Smart Cities, Journal of Intelligent Transportation

Systems, 2450(November), pp. 1–13.

Worldsensing (2019). [Online]. Available at:

http://www.fastprk.com/ [Accessed: 3 Jan. 2019]

Yang, W.X. (2004). An Improved Genetic Algorithm

Adopting Immigration Operator. Intelligent Data

Analysis, 8, pp. 385–401.

Smart Parking Assistance Services and User Acceptance: A European Model

497