Journal of Advanced Transportation

Journal of Advanced Transportation / 2022 / Article
Special Issue

Conference Issue: Modelling and Evaluation of New Paradigms of Mobility

View this Special Issue

Research Article | Open Access

Volume 2022 |Article ID 9833689 | https://doi.org/10.1155/2022/9833689

Andrea Baldassa, Riccardo Ceccato, Federico Orsini, Riccardo Rossi, Massimiliano Gastaldi, "MaaS Bundling and Acceptance in the Pandemic Era: Evidence from Padua, Italy", Journal of Advanced Transportation, vol. 2022, Article ID 9833689, 10 pages, 2022. https://doi.org/10.1155/2022/9833689

MaaS Bundling and Acceptance in the Pandemic Era: Evidence from Padua, Italy

Academic Editor: Alessandro Severino
Received18 Mar 2022
Accepted07 Jul 2022
Published02 Aug 2022

Abstract

Given the benefits both individuals and collectivity have achieved over the past few years thanks to Mobility-as-Service (MaaS) systems, various studies were conducted to predict the level of acceptance of MaaS bundles from different territorial scales and in different countries. Results obtained are in some cases contradictory. Literature is lacking in the study of small-to-medium-sized urban contexts and in the effects of the ongoing COVID-19 pandemic. This paper aims to understand (1) what factors influence respondents’ preferences between their usual transportation means and a possible MaaS alternative and (2) what leads a user to prefer one MaaS bundle to another. A logistic regression and a mixed logit model were developed to reach the two aims, respectively. These models were calibrated using questionnaires administered to employees of the Municipality of Padua, a medium-sized city in Italy. Aspects concerning the perception of health safety in relation to the COVID-19 pandemic were included in the analyses. In 37% of the cases, users stated they would be willing to adopt at least one of the proposed MaaS bundles. The results suggest that MaaS solutions can be a useful tool for managing mobility even in medium-sized cities, provided users’ biosecurity concerns are addressed by appropriate countermeasures.

1. Introduction

Mobility-as-a-Service (MaaS) seeks to integrate different transportation modes to facilitate seamless intermodal planning, booking, and payment through a single interface. According to one of the earliest definitions available in the literature [1], MaaS is a system in which mobility operators provide to customers a comprehensive range of mobility services. In fact, it can be regarded as a set of transportation services that different operators provide, typically bundled in packages and available within a single integrated platform [2], with the aim of providing more efficient mobility solutions [3, 4]. MaaS could potentially reduce the ownership and costs of private vehicles and promote less resource-intensive modes of transportation, such as public transportation or shared services [57]. In this context, the present work focuses on understanding which factors influence bundle choice and its acceptance in a medium-sized city; aspects related to the COVID-19 pandemic were also considered in the analyses. As shown in Section 2, these aspects are still scarcely explored; in this sense, the present study can significantly contribute to the growing body of literature on MaaS bundling.

As the concept of MaaS is gaining increasing interest in both industry and research, there is a large body of literature that analyzes both its capabilities and its possible barriers/risks [8]. In the last few years, the use of intelligent transportation systems has been growing rapidly in modern day cities to promote safer and more organized mobility [9]. MaaS could be applied in the context of smart road management and could be functional toward implementing autonomous vehicles in transportation systems [10].

Typically, in a MaaS scheme, users can pay for each trip (pay-as-you-go, PAYG) or subscribe to mobility bundles/packages. The concept of bundles is receiving increasing attention (see Section 2), and many stated preferences (SP) surveys investigated the effect of sociodemographic characteristics and the transportation modes included in the bundle on users’ preferences and their willingness to pay. In many cases, these SP survey-based approaches were the only viable means of investigating the problem because only a limited number of real-world pilot studies have been completed (e.g., the UbiGo field trial in Gothenburg [11] or the Sydney MaaS trial [1214]), and only few commercial products are available (e.g., Whim). Although, alternative simulative agent-based approaches have been proposed [15, 16].

Recently, Reck et al. [17] reviewed the literature on MaaS bundles, highlighting that there is a lack of research on how to build MaaS bundles and that fundamental differences in survey design limit comparative learning from previous studies, which may explain some contradictory results. Mixed results may also be due to differences in the characteristics of the samples surveyed. Regarding this, Section 2 highlights that virtually all previous SP studies on MaaS bundling gathered data from large urban areas or even entire nations; conversely, there is a lack of studies investigating the potential of MaaS bundling in the context of small/medium-sized cities.

In the last few years, the COVID-19 pandemic has produced profound changes in everyday life worldwide, particularly in the way people move, with changes in travel frequency and preferred transportation modes [1820]. There has been much speculation on what mobility could look like after COVID-19, with one possible scenario being a return to the situation before the virus outbreak and another being a “new normal” that the pandemic experience has strongly influenced [21]. In this context, MaaS faces both opportunities and challenges [22]: on the one hand, an impulse toward the digital transition and sustainable transportation modes [23]; on the other hand, a reduction in travel demand and the use of public transportation [24].

Here, an SP survey was carried out during fall 2020 in Padua (Italy), a city with about 200,000 inhabitants within its administrative limits and about 400,000 when considering the wider urban area. The main objective of this paper is to contribute to the body of knowledge on MaaS bundles, as several studies have previously requested, including [17] and to investigate its potential within a medium-sized Italian city. In particular, the objectives are to understand (1) what factors influence respondents’ preferences between their usual transportation means and a possible MaaS alternative and (2) what leads a user to prefer one MaaS bundle to another.

Furthermore, the survey was administered within the COVID-19 pandemic context, and the results of the present study provide significant insights on how the pandemic may affect MaaS bundling and its acceptance.

2. SP Studies on MaaS Bundling

In the scientific literature on MaaS, one of the first studies investigating the concept of bundles as a mobility management tool was that of Matyas and Kamargianni [25], using an SP survey specifically designed for the task [26, 27]. Their study area was Greater London and involved people over 18 years with a sample that was partially representative of the whole population. Respondents had to choose between three bundles, which could include a combination of subscriptions to public transportation, bike sharing, and taxi and car sharing; subsequently, they were asked about their likelihood to subscribe to their chosen plan. Methodologically, the authors applied two mixed multinomial logit (MNL) models to analyze their results, the first one including data from the whole sample, and the second one including only the responses of those who were likely to subscribe to the chosen bundle. The results of the models were closely aligned, and they highlighted public transportation’s role as the backbone of MaaS. A similar SP experiment was carried out in the Netherlands to investigate the potential role of MaaS bundles as a tool to influence the choice of the commute mode [28]. The study focused exclusively on employees who commute to work using private/leased car and public transportation. Mobility packages included different combinations of subscription levels for trains, public transportation, car sharing, and e-bike sharing, and also the use of private cars, with an increased parking tariff. The result of their mixed MNL model suggested that the presence of trains, car sharing, price, and the parking tariff were all influential factors for the choice of employees’ mode, showing the advantages of combining both incentives and deterrents in transportation demand management strategies.

Concurrently, Ho et al. [29] adopted a face-to-face stated choice study to investigate the potential uptake and willingness to pay for MaaS. Their study involved 252 Sydney-based people who were asked to choose between their current travel pattern and other MaaS bundles generated based on previous responses. The MaaS alternatives included both bundles and PAYG plans. Their results indicated that nearly half of the survey respondents would accept the MaaS solution and that the levels of potential uptake varied considerably in each segment of the population, with casual car users being the most likely adopters and car nonusers being the least. Furthermore, the willingness to pay was much lower compared to the current daily limit (according to a subsequent case study located in Australia [30]), suggesting a careful segmentation of the market and the need for cross-subsidy strategies. Kim et al. [31], who investigated the potential of the MaaS package in Seoul using a web-based interview, followed a similar approach and design. In their study, the binary choice task between the current travel pattern and a proposed mobility package was repeated five times. A latent class model was developed and showed that latent classes corresponded to public transportation-oriented and balanced-mode people, with different influencing factors. Their study’s results emphasized the importance of considering the characteristics of users and building environment in planning MaaS systems.

Guidon et al. [32] adopted a survey design in which Zurich citizens were asked to consider MaaS bundles and standalone subscriptions to transportation services. Many transportation modes were found to be significantly more attractive within bundles, with the exception of taxi and bike/e-bike sharing. For this reason, the authors concluded that mobility providers should not include all transportation modes in composing the bundles.

Another survey-based study investigated MaaS bundles in Greater Manchester [33]. Unlike previously discussed works, respondents had to “build” their ideal MaaS package from a pool of mobility services, including bike sharing, car sharing, taxi, and public transportation (an approach also adopted in another SP choice survey administered in the Netherlands [34]). The price was then calculated based on the chosen features. Consistent with previous studies [29], MaaS could particularly benefit casual car users, while frequent bus users prefer MaaS bundles that only include public transportation. They also discussed the importance of sociodemographic characteristics and the willingness to pay. Within the same research project, another paper addressed the question of preference heterogeneity and developed a latent class choice model [35]. From this, three latent classes emerged, each with a different MaaS package preference and individual characteristics, with current travel habits, gender, age, education, and income all playing significant roles.

According to this literature overview on MaaS bundles, several studies have been conducted over the last few years. However, it is relevant to stress that even if they all shared the same broad objective (i.e., to study the demand for MaaS bundles), their conclusions strongly depend on how the surveys were designed. In particular, it is possible to identify four types of discrete choice designs:(1)Respondents choose among different MaaS bundles (e.g., [25])(2)Respondents choose between the current travel pattern and one or more MaaS bundles (e.g., [29])(3)Respondents choose between MaaS bundles and standalone subscriptions (e.g., [32])(4)Respondents build their own MaaS bundle (e.g., [33])

This led to somewhat contradictory results in certain cases (e.g., Guidon et al. [32] found a positive willingness to pay for a MaaS bundle with car sharing, whereas Matyas and Kamargianni [27] found negative preferences for MaaS bundles that included this service), and several papers (e.g., [25, 36]) highlighted the need for further studies to confirm/extend their findings. Furthermore, all of the above-discussed works dealt with case studies that considered entire nations or large urban areas. The potential of MaaS bundling in small/medium-sized cities is still relatively unexplored. Finally, only a few published papers discussed SP surveys on MaaS bundling conducted during the COVID-19 pandemic [37], and virtually none discussed its implications on users’ preferences.

Given this background in the literature, the present work contributes to exploring MaaS bundling and its acceptance within the specific context of a medium-sized city, including COVID-19 factors and their implications. The SP design adopted here was inspired by [25] and can be regarded as a combination of designs 1 and 2, considering that respondents were asked to choose both between different bundles and between the selected bundle and their current travel plan (more details in Section 3).

3. Materials and Methods

To investigate preferences in MaaS bundling and to analyze factors that could affect its adoption, an online questionnaire was administered between October 2020 and January 2021 to employees of the Municipality of Padua. Note that Padua is an Italian densely populated medium-sized city; currently, around 200,000 people live within its administrative limits and 400,000 within the wider urban area, with a density of about 2,300 inhabitants per square kilometer.

Considering the information needed to apply a stratified random sampling approach was not available, all people working in the municipality were asked to participate in the survey (including office workers, technicians, police officers, and teachers). In particular, while workplaces of potential respondents were within the municipality but not in the same building, individuals could live in any area of the Veneto region.

Specifically, 142 out of 255 respondents declared to live within the city limits, 230 within its province, and all of them within the Veneto region. The city offers several transportation services: urban bus, tram, bike sharing, e-bike sharing, e-scooter sharing, taxi, and car sharing. Padua is also connected to external areas through suburban bus and train services; in addition, there are two Park & Ride areas (one in the north and one in the south) at the city’s entrance.

The survey was divided into four sections. In the part related to revealed preferences (RP), information was collected regarding travel habits, such as the number of weekly trips, frequency of use of different vehicles, and average travel time for working trips. Respondents were asked to select the most frequently used mode of commute in the prepandemic phase, among the following 14 modes: walking, private bike, bike sharing, e-scooter sharing, motorbike, car as a driver alone, car as a driver with passengers, car as a passenger, carpooling, car sharing, taxi, urban public transportation (bus or tram), suburban bus, and train.

In the second part, subjects were asked about aspects related to the COVID-19 pandemic, such as the perceived level of health risk with respect to the current spread of the virus, the perceived level of safety from the health point of view for different travel modes, individual concern about contagion, and the importance of specific measures to contain the spread of the virus onboard vehicles and at public transit stops and stations.

In the third part, panel-structured SP experiments were administered. In particular, each respondent faced choice tasks in which they had to choose between two MaaS bundles [25]. Each bundle’s attributes were defined as a mix of existing mobility services in Padua. The cost of each mobility option in the bundles was derived from what real operators provide in the area. In this way, realistic SP experiments were obtained. Furthermore, following the approach that [25] proposed, not to increase the response burden, neither the status quo option nor the PAYG option were included among the alternatives. Five different experiments were defined, with different attributes of the alternatives, consistently with information from the RP part of the survey (e.g., main transportation mode used and driving license). A D-optimal design procedure was implemented, generating 16 choice scenarios. These scenarios were grouped in blocks so that each unique respondent had to answer four choice tasks. After each SP experiment, respondents had to indicate whether they were willing to adopt the chosen MaaS package, as opposed to their current travel mode. This twofold question structure allowed investigating the MaaS bundling as well as the potential adoption of the service. Table 1 shows the potential components of the proposed MaaS bundles as well as their subscription levels and associated costs. Each of the proposed bundles was defined monthly.


AttributesLevelsStandard cost (€/month)

Bike sharingUnlimited, not included10, 0
E-scooter sharingUnlimited, not included25, 0
Car sharing5 hours included, pay-as-you-go, not included50, 1, 0
Local public transportUnlimited, unlimited rides for 10 days, not included39, 20, 0
Train/suburban busUnlimited, not includedAs stated in RP, 0
Park & RideUnlimited, not included10, 0
Night bus10 rides, 5 rides, not included15, 7.5, 0

Cost0.8, 1, 1.2∗ standard cost

Socioeconomic information was collected in the last part of the survey, both at the household level (members, number of cars, bikes and motorbikes, and total income) and at the individual level (age, gender, postcode of residence, subscription to public transportation, car sharing, and bike sharing).

Data collected were used to calibrate two different models. First, a mixed logit model was implemented to understand factors that influence the choice of one bundle rather than another. Secondly, logistic regression was developed to analyze factors influencing the potential adoption of a proposed MaaS package. Methodologically, with respect to the four discrete choice designs identified in the literature review in Section 2, the present approach can be considered as a combination of designs 1 and 2 because respondents were first asked to choose among different bundles (leading to the MaaS bundle choice model, Section 3.1) and later between the selected bundle and their current travel pattern (leading to the MaaS adoption model, Section 3.2). In this way, a comprehensive analysis was performed, with the aim to analyze not only the choice of a specific MaaS bundle but also the future use of the service.

3.1. MaaS Bundle Choice Model

Mixed logit is a discrete choice model that overcomes some of the standard logit model’s limitations by considering random variations in taste between selectors, unrestricted substitution patterns across choices, and correlation in not observed factors over time [38]. In this paper, a mixed logit model was calibrated using answers to SP experiments, where respondents had to choose between the two MaaS bundles. The model allowed accounting for the survey’s panel structure. The model was calibrated using Biogeme software [39].

Table 2 shows the list of exogenous variables used for model specification.


NameDescriptionTypeLevel

HH_CARNumber of household carsMetricHousehold
HH_MOTONumber of household motorbikesMetricHousehold
INCOMEHousehold income (1000€)MetricHousehold
F_CARFrequency of private car use (times/week)MetricIndividual
PDLiving inside Padua municipalityMetricIndividual
BSBike sharing service (unlimited) in the MaaS packageDummyMaaS
COSTCost of the MaaS packageMetricMaaS
CS_HOURCar sharing service (5 hours) in the MaaS packageDummyMaaS
CS_RATECar sharing service (pay-as-you-go) in the MaaS packageDummyMaaS
NBNight bus service in the MaaS packageMetricMaaS
PARKPark & Ride service (unlimited) in the MaaS packageDummyMaaS
SCE-scooter sharing service (unlimited) in the MaaS packageDummyMaaS
SUBSuburban (train/bus) public service (unlimited) in MaaS packageDummyMaaS
TPLUrban public transportation service (unlimited) in the MaaS packageDummyMaaS
TPL_10DAYSUrban public transportation service (unlimited for 10 days) in the MaaS packageDummyMaaS
RP_ACTIVEUsed active mobility (by foot or bike) in the past to commuteDummyTrip
RP_CARUsed car (driver or passenger) in the past to commuteDummyTrip
RP_MOTOUsed motorbike in the past to commuteDummyTrip
RP_PTUsed public transportation in the past to commuteDummyTrip

3.2. MaaS Adoption Model

Logistic regression is a generalized linear regression model used when the dependent variable is dichotomic. In this paper, the model was used to investigate the probability of adopting MaaS, considering the characteristics of both the bundles and the respondents.

Significant variables were selected manually and by applying an automated stepwise procedure. The model was built with R statistical software [40]. Table 3 shows the list of exogenous variables used in the model’s final version. Note that the variables related to the COVID-19 pandemic were included in the model specification; in this way, potential effects of the pandemic on future MaaS diffusion were investigated.


NameDescriptionTypeLevel

HH_CARNumber of household carsMetricHousehold
INCOMEHousehold income (1000€)MetricHousehold
AGEAgeMetricIndividual
F_BUSFrequency of urban bus use (times/week)MetricIndividual
F_PBFrequency of private bike use (times/week)MetricIndividual
F_TRAINFrequency of train use (times/week)MetricIndividual
GENDER_FFemaleDummyIndividual
MEAS_E_4Importance of automatic access control via web, phone, or app booking system for public transportation to reduce the risk of infection (important)DummyIndividual
MEAS_E_5Importance of automatic access control via web, phone, or app booking system for public transportation to reduce the risk of infection (very important)DummyIndividual
MEAS_F_4Importance of daily sanitization of vehicles and stations to reduce the risk of infection (important)DummyIndividual
MEAS_H_5Importance of body temperature checks with thermal scanners at station entrance to reduce the risk of infection (very important)DummyIndividual
MEAS_L_2Importance of walk a few more minutes to reach less crowded stations/stops to reduce the risk of infection (not important)DummyIndividual
MEAS_L_5Importance of walk a few more minutes to reach less crowded stations/stops to reduce the risk of infection (very important)DummyIndividual
PDLiving inside Padua municipalityDummyIndividual
SAFE_BUS_1Perceived level of travelling health risk on transportation means (not at all safe)DummyIndividual
SAFE_BUS_2Perceived level of travelling health risk on transportation means (not safe)DummyIndividual
COSTCost of the MaaS packageMetricMaaS
CS_HOURCar sharing service (5 hours) in the MaaS packageDummyMaaS
CS_RATECar sharing service (pay-as-you-go) in the MaaS packageDummyMaaS
NBNight bus service in MaaS packageMetricMaaS
PARKPark & Ride service in the MaaS packageDummyMaaS
SCE-scooter sharing service (unlimited) in the MaaS packageMetricMaaS
TPLUrban public transportation service (unlimited) in the MaaS packageDummyMaaS
TPL_10DAYSUrban public transportation service (unlimited for 10 days) in the MaaS packageDummyMaaS
RP_BUSUsed urban bus in the past for the commuting tripDummyTrip
RP_DRIVEPASSUsed car (as driver with passengers) in the past for the commuting tripDummyTrip
RP_MOTOUsed motorbike in the past for the commuting tripDummyTrip
TTTravel time for the commuting trip (min)MetricTrip

4. Results and Discussion

There were 255 complete questionnaires. There are 1,650 employees of Padua’s municipality, which corresponds to a sampling rate of about 16%; the sample size was considered appropriate for a margin of error of 10% at 95% confidence level [41].

Table 4 provides a summary of the respondents’ socioeconomic characteristics.


N%

Totals255100
Household members
15321
25622
36425
46525
More than 4177

Licensed drivers
052
16827
211545
35120
More than 3166

Available cars
062
110943
211846
3208
More than 321

Available bikes
0104
13213
27027
35522
More than 38835

Available motorbikes
016565
17529
2135
321
More than 300

Household income (€/month)
Less than 100062
1000–15007027
1500–20003112
2000–30008533
3000–40004618
4000–6000114
6000–1000021
More than 10,00042

Gender
Female16866
Male8734

Age
18–2400
25–2910
30–3452
35–444016
45–548935
55–6411645
More than 6542

Bus pass
Yes3915
No21685

Train pass
Yes229
No23391

Bike sharing pass
Yes21
No25399

Living in Padua
Yes14256
No11344

Table 5 shows the main travel means subjects used to reach their workplace; driving own car was the most common travel mode, followed by private bikes. Indeed, most respondents (56%) live in Padua’s municipality and need to cover relatively short distances to commute. A quarter of the interviewees adopted public transportation modes (bus and train). At the time of survey administration, shared mobility services (bike, e-scooter, and car) seem not to be appreciated (1% of the sample selected bike sharing).


N%

Total255100
Walking104
E-scooter sharing00
Private bike5923
Bike sharing21
Motorbike2510
Car (alone)8634
Car (driver, with passengers)73
Car (as passenger)52
Car sharing00
Bus2711
Suburban bus125
Train229

4.1. MaaS Bundle Choice-Model Estimation

Results from the calibration phase of the mixed logit model that predicts the choice between two MaaS bundles are reported in Table 6. In the model’s final version, three random parameters were adopted. Constants were included with a normal distribution, as were the cost of packages; conversely, the variable related to the presence of e-scooter sharing service in the bundle followed a uniform distribution. Moreover, 1,000 Halton intelligent draws were used.


Variable nameCoefficientSEz value value

ASC_Ast.dev.1.0207.0007.140<0.001∗∗∗
ASC_B mean0.3500.1931.8100.070†
ASC_Bst.dev.0.9830.1426.900<0.001∗∗∗
B_BS0.4060.1462.7800.005∗∗
B_COSTmean−0.0140.004−3.420<0.001∗∗∗
B_COSTst.dev.−0.0210.006−3.620<0.001∗∗∗
B_CS_RATE0.3050.1571.9500.051
B_F_CARxCS_HOUR−0.2020.069−2.9400.003∗∗
B_HH_CARxNB−0.0460.012−3.860<0.001∗∗∗
B_HH_CARxTPL_10DAYS0.6820.1853.680<0.001∗∗∗
B_HH_MOTOxSC−0.5350.239−2.2400.025
B_INCOMExSC0.1180.0751.5700.116
B_PDxCOST−0.0060.003−2.1100.036
B_PDxTPL_10DAYS0.9180.3352.7400.006∗∗
B_RP_ACTIVExSC1.4600.3224.530<0.001∗∗∗
B_RP_CARxPARK1.2700.2784.560<0.001∗∗∗
B_RP_CARxTPL1.4300.3044.710<0.001∗∗∗
B_RP_MOTOxCS_RATE1.4100.4772.9500.003∗∗
B_RP_PTxSUB3.1800.9203.4600.001∗∗
B_SCmean−1.6800.524−3.2000.001∗∗
B_SCst.dev.2.6400.9152.8800.004∗∗
B_TPL0.6560.2632.4900.013
B_TPL_10DAYS−0.3230.223−1.4500.147
Significance codes: ∗∗∗ value <0.001; ∗∗ value <0.01; ∗ value <0.05; † value <0.10
Statistics
N. of observation1,020.00
N. of draws1,000.00
Null log likelihood−1,072.59
Final log likelihood−563.99
Likelihood ratio test1,017.20
Rho-square-bar0.45
Akaike criterion (AIC)1,173.99

The bundle’s cost had a negative effect, as expected (B_COST). MaaS packages with unlimited bike sharing (B_BS), car sharing with PAYG option (B_CS_RATE), and unlimited urban public transportation (B_TPL) were likely to be chosen. On the contrary, the presence of a 10-day urban public transportation pass (B_TPL_10DAYS) had a negative effect on users’ choice of the bundle; however, this effect was reversed for people living in Padua (B_PDxTPL_10DAYS).

Mobility habits were found to play a significant role on the bundle choice. Those travelling to work using a private car appreciated Park & Ride (B_RP_CARxPARK). Moreover, the coefficient of the variable representing the interaction between the private car frequency and the presence of 5-hour car sharing was negative, indicating that this service could not foster substituting private vehicles (B_F_CARxCS_HOUR). Surprisingly, the car did not reduce the likelihood of choosing MaaS bundles with public transportation. In particular, commuters travelling via car appreciated bundles containing urban transit more than others did (B_RP_CARxTPL), and the number of cars available in the family positively affected the adoption of bundles with a 10-day public transportation pass (B_HH_CARxTPL_10DAYS); this aspect will certainly need to be verified with future studies, but it could be due to the proposed service’s flexibility. However, the coefficient of the variable indicating the interaction between the night bus service and household cars was negative (B_HH_CARxNB); this was an expected result because people belonging to households with a large number of vehicles are likely to use one of them, instead of public transportation, during the night.

As expected, commuters using public transportation showed a strong propensity toward MaaS packages containing suburban transportation services (B_RP_PTxSUB). In addition, people adopting active modes seemed to prefer bundles with the e-scooter sharing service (B_RP_ACTIVExSC).

4.2. MaaS Adoption-Model Estimation

The calibration of the logistic regression parameters and model statistics are reported in Table 7. The model predicts the probability of adopting the MaaS plan chosen in the previous SP experiment. Concerning general bundle composition, results showed that the car-sharing service (CS_HOUR, CS_RATE) prompted respondents to reject the MaaS (according to the negative coefficient reported in [25]). Users who took more time to travel from home to work declared they were more willing to adopt the MaaS system (TT). People who lived in Padua, on the other hand, were less interested in the service (PD), except for bundles that included Park & Ride (PDxPARK) and the e-scooter sharing service (PDxSC). Women were less attracted to MaaS in general (GENDER_F), but they seemed to appreciate bundles containing unlimited urban transportation services (GENDER_FxTPL). This is somewhat consistent with recent studies reporting that women do not well receive shared mobility regarding safety, reliability, and user-friendliness, even though women are known to use public transportation more than men do [42]. This also highlights the need for additional studies on the gender gap in the transportation sector [43]. As the weekly use of city bus, private bike, or train increased, the attractiveness of MaaS decreased (F_BUS, F_PB, F_TRAIN). The older respondents appreciated MaaS more (AGE), which opposed what was observed in other studies [34, 44], except for the packages that contained the e-scooter sharing service (AGExSC). Understandably, as income per household increased, there was a greater propensity to accept higher cost bundles (INCOMExCOST), but at the same time, those with greater economic resources did not appreciate the night bus service (INCOMExNB). As the number of cars available to the family increased, the declared propensity for the adoption of MaaS decreased (HH_CAR). Users who used the bus, the car as a passenger or the motorcycle as the main means of commute declared they would be in favor of adopting the proposed MaaS bundles (RP_BUS, RP_DRIVEPASS, RP_MOTO).


Variable nameCoefficientOddsSEz value value

AGE0.0261.0270.0141.8740.061†
AGExCS_HOUR0.0621.0640.0321.9200.055†
AGExSC−0.0120.9890.005−2.3410.019
CONSTANT−2.9660.0520.908−3.2650.001∗∗
CS_HOUR−3.4550.0321.816−1.9030.057†
CS_RATE−0.5340.5860.205−2.6080.009∗∗
F_BUS−0.1460.8640.088−1.6590.097†
F_PB−0.1430.8670.059−2.4390.015
F_TRAIN−0.1950.8230.078−2.4920.013
GENDER_F−0.6660.5140.232−2.8770.004∗∗
GENDER_FxTPL0.8942.4450.3582.4950.013
HH_CAR−0.4170.6590.143−2.9050.004∗∗
INCOMExCOST0.0011.000<0.0014.787<0.001∗∗∗
INCOMExCS_HOUR−0.0011.000<0.001−2.0610.039
INCOMExNB−0.0011.000<0.001−3.2680.001∗∗
MEAS_E_41.0472.8490.3952.6540.008∗∗
MEAS_E_51.2593.5210.4093.0810.002∗∗
MEAS_F_41.4204.1380.3024.697<0.001∗∗∗
MEAS_H_50.5411.7170.2242.4090.016
MEAS_L_20.9722.6430.4782.0350.042
MEAS_L_50.5811.7870.2022.8750.004∗∗
PD−1.1260.3240.293−3.849<0.001∗∗∗
PDxPARK0.9172.5010.3492.6270.009∗∗
PDxSC0.6151.8490.3631.6940.09†
RP_BUS0.7282.0710.5041.4450.149
RP_DRIVEPASS0.8882.4300.5151.7230.084†
RP_MOTO0.8422.3200.3232.6090.009∗∗
SAFE_BUS_1xTPL−1.3230.2660.331−4.000<0.001∗∗∗
SAFE_BUS_1xTPL_10DAYS−0.6370.5290.259−2.4580.014
SAFE_BUS_2xTPL−1.1520.3160.382−3.0160.003∗∗
SAFE_BUS_2xTPL_10DAYS−0.5630.5700.353−1.5970.11
TT0.0211.0220.0063.3720.001∗∗
Significance codes: ∗∗∗ value <0.001; ∗∗ value <0.01; ∗ value <0.05; † value <0.10
Statistics
N. of observation1,200
Null log likelihood−507.63
Final log likelihood−420.61
Cragg and Uhler’s rho-squared0.25

Variables related to the COVID-19 pandemic could play a nonnegligible role on MaaS adoption. As previous authors reported [24], biosecurity concerns about public transportation was found to be one of the main affecting factors. Respondents who considered using urban public transportation very unsafe or unsafe, from a health point of view, declared that they would be more likely to reject a MaaS scheme containing the urban public transportation option (SAFE_BUS_1XTPL, SAFE_BUS_1xTPL_10DAYS, SAFE_BUS_2xTPL, SAFE_BUS_2xTPL_10GG). Regarding the countermeasures that could be adopted to contain the spread of COVID-19, those who rated as important or very important to control access to public means via the web, telephone, or app booking system declared they were more inclined to adopt proposed MaaS solutions (MEAS_E_4, MEAS_E_5); the same was true for those who believed that daily sanitizing vehicles and stations was important (MEAS_F_4) and for those who considered body temperature checks with thermal scanners at the station entrance as very important (MEAS_H_5). MEAS_L_2 and MEAS_L_5 were both positive, showing that MaaS can be appreciated by those who are unwilling and, similarly, by those who are willing to walk a few more minutes to reach less crowded stations/stops to reduce risks associated with COVID-19.

5. Conclusions

In this research activity, answers from a questionnaire administered to employees of the Municipality of Padua (Italy) were used to understand factors that (1) affect the adoption of MaaS and (2) most influence the choice of a specific package over another. Two models have been built: the first, mixed logit to interpret which MaaS bundle compositions were more attractive than others were; the second model, a logistic regression model to evaluate the factors that affected MaaS adoption replacing the transportation mode currently used. As this is an SP experiment, all the information collected is in hypothetical terms. From a socioeconomic point of view, age and gender had no effect on package composition preferences, but women were less likely to change their habits regarding adopting MaaS; high-income people and people with a large number of cars available in the household were not particularly interested in the services offered. Users who made very long trips from home to work (regarding time) were more likely to join MaaS, probably because it would represent an inexpensive/cheap solution in any case. People who lived near the workplace and those who were accustomed to riding their bike to get to work were not inclined to choose MaaS solutions (the two groups partially coincided). Furthermore, extending the potential catchment area of MaaS played a significant role in its future adoption; in particular, people living in the city center, although less willing to subscribe, were more likely to accept a bundle with Park & Ride services and e-scooter sharing. From this point of view, the present work contributes to shed light on the MaaS relevance in a typical medium-sized city. In about 20% of the cases (202 out of 1,020 choices), respondents stated they would be willing to change their travel habits by adopting the proposed MaaS bundle. More than 37% of respondents (95 out of 255) indicated they were inclined to accept more than one of the bundles proposed in the four choice exercises. Regarding proposed MaaS configurations, car-sharing services received the least appreciation. Users accustomed to using shared modes were most likely to join MaaS. Lastly, the biosecurity concern of travelling on public transportation could significantly reduce the likelihood to adopt MaaS. This represents a barrier to adopting the service because public transit is usually regarded as the backbone of MaaS. However, the analysis pointed out specific countermeasures, both onboard vehicles and at stations/stops, which could mitigate the negative effects of the COVID-19 pandemic on the future diffusion of MaaS.

In conclusion, the results showed that a good portion of commuters would be willing to adopt MaaS, demonstrating the potential of the latter in improving urban and nonurban mobility. The results obtained can help companies build MaaS bundles that are more attractive for users, thus promoting its adoption. Starting from these considerations, future studies could investigate the attractiveness of packages’ components in a similar context, in particular, focusing on packages’ prices and exploring different and/or larger samples. The MaaS system can be an excellent tool to improve traveler habits and simultaneously improve environmental and traffic conditions in the network, despite the situation and uncertainties related to the presence of the COVID-19 pandemic. Although the virus implies a general lower propensity to use shared means of transportation, using the right countermeasures could be a correct approach to return to normal modal split levels, and in fact, foster new sustainable mobility services such as MaaS in many cities.

Data Availability

The data used to support the findings of this study are available upon request from the corresponding author.

Conflicts of Interest

The authors declare that there are no conflicts of interest regarding the publication of this paper.

Authors’ Contributions

Conceptualization was conducted by MG, RR, RC, AB, and FO; methodology was carried out by AB and RC; software was performed by AB, RC, and FO; validation was conducted by RC and FO; formal analysis was carried out by AB; investigation was conducted by AB, FO, and RC; data curation was performed by AB, writing of the original draft was carried out by AB, FO, and RC; writing, review, and editing were performed by RC, MG, FO, and AB; supervision was conducted by MG and RR.

Acknowledgments

The authors thank the Municipality of Padua for providing support in survey administration and Andrea Colovini for his contribution in survey design.

References

  1. S. Heikkil, Mobility as a Service - A Proposal for Action for the Public Administration, Case Helsinki (Master of Science in Technology), Aalto University, Helsinki, Finland, 2014.
  2. J. Sochor, H. Arby, I. M. KarlssonKarlsson, and S. Sarasini, “A topological approach to Mobility as a Service: a proposed tool for understanding requirements and effects, and for aiding the integration of societal goals,” Research in Transportation Business & Management, vol. 27, pp. 3–14, 2018. View at: Publisher Site | Google Scholar
  3. N. C. J. Cox, Estimating Demand for New Modes of Transportation Using a Context-Aware Stated Preference Survey, Massachusetts Inst. Technol, United States, pp. 1–195, 2015.
  4. M. Kamargianni and M. Matyas, “The business ecosystem of mobility-as-a-service,” Transp. Res. board, vol. 96, pp. 1–14, 2017. View at: Google Scholar
  5. M. Kamargianni, W. Li, M. Matyas, and A. Schäfer, “A critical review of new mobility services for urban transport,” Transportation Research Procedia, vol. 14, pp. 3294–3303, 2016. View at: Publisher Site | Google Scholar
  6. P. Jittrapirom, V. Marchau, R. van der Heijden, and H. Meurs, “Future implementation of mobility as a service (MaaS): results of an international Delphi study,” Travel Behaviour and Society, vol. 21, pp. 281–294, 2020. View at: Publisher Site | Google Scholar
  7. E. Ko, H. Kim, and J. Lee, “Survey data analysis on intention to use shared mobility services,” Journal of Advanced Transportation, vol. 2021, Article ID 5585542, p. 10, 2021. View at: Publisher Site | Google Scholar
  8. L. Butler, T. Yigitcanlar, and A. Paz, “Barriers and risks of Mobility-as-a-Service (MaaS) adoption in cities: a systematic review of the literature,” Cities, vol. 109, Article ID 103036, 2021. View at: Publisher Site | Google Scholar
  9. F. Arena, G. Pau, A. Ralescu, A. Severino, and I. You, “An innovative framework for dynamic traffic lights management based on the combined use of fuzzy logic and several network architectures,” Journal of Advanced Transportation, vol. 2022, Article ID 1383349, 2022. View at: Publisher Site | Google Scholar
  10. S. Trubia, A. Severino, S. Curto, F. Arena, and G. Pau, “Smart roads: an overview of what future mobility will look like,” Infrastructure, vol. 5, no. 12, p. 107, 2020. View at: Publisher Site | Google Scholar
  11. J. Sochor, I. C. M. A. Karlsson, and H. Strömberg, “Trying out mobility as a service: experiences from a field trial and implications for understanding demand,” Transportation Research Record: Journal of the Transportation Research Board, vol. 2542, no. 1, pp. 57–64, 2016. View at: Publisher Site | Google Scholar
  12. C. Q. Ho, D. A. Hensher, D. J. Reck, S. Lorimer, and I. Lu, “MaaS bundle design and implementation: lessons from the Sydney MaaS trial,” Transportation Research Part A: Policy and Practice, vol. 149, pp. 339–376, 2021. View at: Publisher Site | Google Scholar
  13. C. Q. Ho, D. A. Hensher, and D. J. Reck, “Drivers of participant’s choices of monthly mobility bundles: k,” Transportation Research Part C: Emerging Technologies, vol. 124, Article ID 102932, 2021. View at: Publisher Site | Google Scholar
  14. D. A. Hensher, C. Q. Ho, and D. J. Reck, “Mobility as a service and private car use: evidence from the Sydney MaaS trial,” Transportation Research Part A: Policy and Practice, vol. 145, pp. 17–33, 2021. View at: Publisher Site | Google Scholar
  15. C. Cisterna, G. Giorgione, and F. Viti, “Explorative analysis of potential MaaS customers: an agent-based scenario,” Procedia Computer Science, vol. 184, pp. 629–634, 2021. View at: Publisher Site | Google Scholar
  16. C. Cisterna, F. Bigi, F. Tinessa, and F. Viti, “Analysis of MaaS membership attributes: an agent-based approach,” Transportation Research Procedia, vol. 62, pp. 483–490, 2022. View at: Publisher Site | Google Scholar
  17. D. J. Reck, D. A. Hensher, and C. Q. Ho, “MaaS bundle design,” Transportation Research Part A: Policy and Practice, vol. 141, pp. 485–501, 2020. View at: Publisher Site | Google Scholar
  18. C. Balbontin, D. A. Hensher, M. J. Beck et al., “Impact of COVID-19 on the number of days working from home and commuting travel: a cross-cultural comparison between Australia, South America and South Africa,” Journal of Transport Geography, vol. 96, Article ID 103188, 2021. View at: Publisher Site | Google Scholar
  19. T. Shibayama, F. Sandholzer, B. Laa, and T. Brezina, “Impact of covid-19 lockdown on commuting: a multi-country perspective,” European Journal of Transport and Infrastructure Research, vol. 21, no. 1, pp. 70–93, 2021. View at: Google Scholar
  20. R. Shortall, N. Mouter, and B. Van Wee, “COVID-19 passenger transport measures and their impacts,” Transport Reviews, vol. 0, no. 0, pp. 1–26, 2021. View at: Publisher Site | Google Scholar
  21. D. A. Hensher, “What might Covid-19 mean for mobility as a service (MaaS)?” Transport Reviews, vol. 40, no. 5, pp. 551–556, 2020. View at: Publisher Site | Google Scholar
  22. S. Paiva and F. Mourao, “Mobility-as-a-Service challenges and opportunities in the post-pandemic,” in Proceedings of the 2021 IEEE Glob. Conf. Artif. Intell. Internet Things, pp. 136–141, Dubai, United Arab Emirates, December 2022. View at: Google Scholar
  23. N. A. Megahed and E. M. Ghoneim, “Antivirus-built environment: lessons learned from Covid-19 pandemic,” Sustainable Cities and Society, vol. 61, Article ID 102350, 2020. View at: Publisher Site | Google Scholar
  24. A. Tirachini and O. Cats, “COVID-19 and public transportation: current assessment, prospects, and research needs,” Journal of Public Transportation, vol. 22, no. 1, pp. 1–34, 2020. View at: Publisher Site | Google Scholar
  25. M. Matyas and M. Kamargianni, “The potential of mobility as a service bundles as a mobility management tool,” Transportation, vol. 46, no. 5, pp. 1951–1968, 2019. View at: Publisher Site | Google Scholar
  26. M. Matyas and M. Kamargianni, “A stated preference experiments for mobility-as-a-service plans,” in Proceedings of the 5th IEEE Int. Conf. Model. Technol. Intell. Transp. Syst. MT-ITS 2017 - Proc, pp. 738–743, Naples, Italy, June 2017. View at: Publisher Site | Google Scholar
  27. M. Matyas and M. Kamargianni, “Survey design for exploring demand for Mobility as a Service plans,” Transportation, vol. 46, no. 5, pp. 1525–1558, 2019. View at: Publisher Site | Google Scholar
  28. Z. H. Farahmand, K. Gkiotsalitis, and K. T. Geurs, “Mobility-as-a-Service as a transport demand management tool: a case study among employees in The Netherlands,” Case Studies on Transport Policy, vol. 9, no. 4, pp. 1615–1629, 2021. View at: Publisher Site | Google Scholar
  29. C. Q. Ho, D. A. Hensher, C. Mulley, and Y. Z. Wong, “Potential uptake and willingness-to-pay for Mobility as a Service (MaaS): a stated choice study,” Transportation Research Part A: Policy and Practice, vol. 117, pp. 302–318, 2018. View at: Publisher Site | Google Scholar
  30. C. Mulley, C. Ho, C. Balbontin et al., “Mobility as a service in community transport in Australia: can it provide a sustainable future?” Transportation Research Part A: Policy and Practice, vol. 131, no. September 2019, pp. 107–122, 2020. View at: Publisher Site | Google Scholar
  31. S. Kim, S. Choo, S. Choi, and H. Lee, “What factors affect commuters’ utility of choosing mobility as a service? An empirical evidence from Seoul,” Sustainability, vol. 13, no. 16, p. 9324, 2021. View at: Publisher Site | Google Scholar
  32. S. Guidon, M. Wicki, T. Bernauer, and K. Axhausen, “Transportation service bundling - for whose benefit? Consumer valuation of pure bundling in the passenger transportation market,” Transportation Research Part A: Policy and Practice, vol. 131, pp. 91–106, 2020. View at: Publisher Site | Google Scholar
  33. I. Tsouros, A. Tsirimpa, I. Pagoni, and A. Polydoropoulou, “MaaS users: who they are and how much they are willing-to-pay,” Transportation Research Part A: Policy and Practice, vol. 148, pp. 470–480, 2021. View at: Publisher Site | Google Scholar
  34. V. Caiati, S. Rasouli, and H. Timmermans, “Bundling, pricing schemes and extra features preferences for mobility as a service: sequential portfolio choice experiment,” Transportation Research Part A: Policy and Practice, vol. 131, pp. 123–148, 2020. View at: Publisher Site | Google Scholar
  35. M. Matyas and M. Kamargianni, “Investigating heterogeneity in preferences for Mobility-as-a-Service plans through a latent class choice model,” Travel Behaviour and Society, vol. 23, pp. 143–156, 2021. View at: Publisher Site | Google Scholar
  36. M. Hasselwander, J. F. Bigotte, A. P. Antunes, and R. G. Sigua, “Towards sustainable transport in developing countries: preliminary findings on the demand for mobility-as-a-service (MaaS) in Metro Manila,” Transportation Research Part A: Policy and Practice, vol. 155, pp. 501–518, 2022. View at: Publisher Site | Google Scholar
  37. M. V. Corazza and G. Carassiti, “Mobility as a Service in the City of Rome: Potential and Feasibility,” in Proceedings of the 2021 IEEE Int. Conf. Environ. Electr. Eng. 2021 IEEE Ind. Commer. Power Syst. Eur. (EEEIC/I& CPS Eur, pp. 1–6, Bari, Italy, 2021. View at: Google Scholar
  38. K. E. Train, “Discrete choice methods with simulation,” Discret. Choice Methods with Simul, vol. 18, pp. 1–334, 2003. View at: Publisher Site | Google Scholar
  39. M. Bierlaire, “A Short Introduction to PandasBiogeme,” Tech. Rep., 2020, Technical report TRANSP-OR 200605. View at: Google Scholar
  40. R. Core Team, “R: A Language and Environment for Statistical Computing,” 2020. View at: Google Scholar
  41. D. A. Hensher, J. M. Rose, and W. H. Greene, “Applied choice analysis: a primer,” Cambridge University Press, pp. 1–717, 2005, Cambridge, UK. View at: Google Scholar
  42. M. Pirra, S. Kalakou, A. Carboni, M. Costa, M. Diana, and A. R. Lynce, “A preliminary analysis on gender aspects in transport systems and mobility services: presentation of a survey design,” Sustainability, vol. 13, no. 5, p. 2676, 2021. View at: Publisher Site | Google Scholar
  43. M. Pirra, A. Carboni, and M. Diana, “Assessing gender gaps in educational provision, research and employment opportunities in the transport sector at the european level,” Education Sciences, vol. 10, no. 5, p. 123, 2020. View at: Publisher Site | Google Scholar
  44. A. Vij, S. Ryan, S. Sampson, and S. Harris, “Consumer preferences for mobility-as-a-service (MaaS) in Australia,” Transportation Research Part C: Emerging Technologies, vol. 117, Article ID 102699, 2020. View at: Publisher Site | Google Scholar

Copyright © 2022 Andrea Baldassa et al. This is an open access article distributed under the Creative Commons Attribution License, which permits unrestricted use, distribution, and reproduction in any medium, provided the original work is properly cited.

Related articles

No related content is available yet for this article.
 PDF Download Citation Citation
 Download other formatsMore
 Order printed copiesOrder
Views76
Downloads51
Citations

Related articles

No related content is available yet for this article.

Article of the Year Award: Outstanding research contributions of 2021, as selected by our Chief Editors. Read the winning articles.