To read the full-text of this research, you can request a copy directly from the authors.
Abstract
Research on the factors that may affect the future uptake of autonomous Mobiliyt on Demand (MoD) services are today more than ever relevant. In this paper, we attempt to investigate all aspects of acceptability of a proposed Autonomous-MoD service (AMoD). More specifically, we develop mode choice models and identify the factors that affect it in both sunny and rainy weather conditions, using state-of-the-art Machine Learning models and interpretation techniques, such as the permutation feature importance and partial dependence. Furthermore, we estimate the willingness of the service’s potential users to pay for reduced travel time and propose an on-board negotiation scheme of the travel time and cost for sharing one’s ride. For the above purposes, we conducted a questionnaire survey with 1600 participants in the city of Athens, Greece just after alleviation of the lockdown and measures related with the COVID-19 first wave. The models developed are capable of predicting the mode choice and acceptability of the negotiation scheme with an accuracy of over 80%. Except for the cost, travel and walking time of each alternative mode, the users’ mobility profile, attitude towards autonomous vehicles and demographic characteristics are identified as the most important factors affecting the respondents’ choices. Moreover, the willingness to pay for reduced travel time varies from 0.18 to 0.62€, depending on the mode and about 0.53€ for the on-board negotiation.
To read the full-text of this research, you can request a copy directly from the authors.
... Age and gender are most often considered. Overall, younger individuals seem to have a higher acceptability of SAVs (Bansal et al. 2016;Cartenì 2020;Fafoutellis et al. 2021;Guo et al. 2020;Haboucha et al. 2017;König and Grippenkoven 2020;Polydoropoulou et al. 2021;Sener and Zmud 2019;Wang et al. 2020;Wicki et al. 2019). However, studies that used age categories observed an inverted U-trend reflecting less interest of younger and older age groups (Berrada et al. 2020;Clayton et al. 2020;Krueger et al. 2016;Rosell and Allen 2020). ...
... Other studies found an insignificant influence of age, suggesting a more equal acceptability across different age groups (Alfonsi et al. 2018;Bansal et al. 2016;Feys et al. 2020;Guo et al. 2020;Haboucha et al. 2017;Hamadneh and Esztergár 2021;Krueger et al. 2016;Lavieri and Bhat 2019;Müller 2019;Paddeu et al. 2020;Rosell and Allen 2020;Wang et al. 2021;Yuen et al. 2020). Men seem to have a higher acceptability of SAVs (Alfonsi et al. 2018;Bansal et al. 2016;Cartenì 2020;Fafoutellis et al. 2021;Haboucha et al. 2017;König and Grippenkoven 2020;Lavieri and Bhat 2019;Nathanail et al. 2020;Rosell and Allen 2020;Sener and Zmud 2019;Wang et al. 2020) but women seem to be more interested in a ride-sharing system (Fafoutellis et al. 2021;Nathanail et al. 2020;Rosell and Allen 2020). Yet, several studies found an insignificant influence of gender (Berrada et al. 2020 Compared to the studies including socio-economic and demographic variables, the number of publications focusing on the acceptability of SAVs by vulnerable groups is more limited and recently published. ...
... Other studies found an insignificant influence of age, suggesting a more equal acceptability across different age groups (Alfonsi et al. 2018;Bansal et al. 2016;Feys et al. 2020;Guo et al. 2020;Haboucha et al. 2017;Hamadneh and Esztergár 2021;Krueger et al. 2016;Lavieri and Bhat 2019;Müller 2019;Paddeu et al. 2020;Rosell and Allen 2020;Wang et al. 2021;Yuen et al. 2020). Men seem to have a higher acceptability of SAVs (Alfonsi et al. 2018;Bansal et al. 2016;Cartenì 2020;Fafoutellis et al. 2021;Haboucha et al. 2017;König and Grippenkoven 2020;Lavieri and Bhat 2019;Nathanail et al. 2020;Rosell and Allen 2020;Sener and Zmud 2019;Wang et al. 2020) but women seem to be more interested in a ride-sharing system (Fafoutellis et al. 2021;Nathanail et al. 2020;Rosell and Allen 2020). Yet, several studies found an insignificant influence of gender (Berrada et al. 2020 Compared to the studies including socio-economic and demographic variables, the number of publications focusing on the acceptability of SAVs by vulnerable groups is more limited and recently published. ...
The increasing world population in urban areas and intensive use of private vehicles have several negative impacts, such as congestion, air pollution, and traffic crashes, affecting public health and well-being, especially for vulnerable groups (e.g., children, elderly, women), further increasing social inequality. Current sustainable transportation options like active and public transportation are not always alternatives for these groups. One of the most promising sustainable transportation options is shared autonomous transportation. However, it is rather unclear if vulnerable groups accept shared autonomous vehicles or SAVs, questioning their social sustainability. Based on the definitions for “social acceptability” and “social acceptance” of transportation options, the Transport Acceptancy-Vulnerability or TAV model is presented. The model combines the “acceptancy,” i.e., “acceptability,” “acceptance,” with the “vulnerability” referring to the four conditions to be met or the 4As, i.e., “availability,” “accessibility,” “affordability,” “attractability,” toward transportation options. This model can be used to address the “social acceptability” or “social acceptance” of sustainable transportation options by vulnerable groups. It allows the structuring and evaluation of literature as well as data helping to determine whether the factors of “acceptancy” or the conditions of “vulnerability” for a transportation option need to be improved. The usability is illustrated by examining the scientific literature on the acceptability of SAVs by elderly, women, households with children, and people with disabilities, and by an example of the social acceptability of SAVs by different potentially vulnerable groups. The model can help transportation authorities, operators, and practitioners to improve socially sustainable urban transportation and overall social inclusion.
... In general, age seems to show a more negative trend, with younger individuals showing a higher acceptability of STSs (Table A1). This negative effect is particularly noticeable for SAVs (Bansal et al., 2016;Cartenì, 2020;Fafoutellis et al., 2021;Guo et al., 2020;Haboucha et al., 2017;König & Grippenkoven, 2020;Polydoropoulou et al., 2021;Sener & Zmud, 2019;Wang et al., 2020;Wicki et al., 2019). This trend may follow an inverted U-shape, as Berrada et al. (2020), Clayton et al. (2020), Krueger et al. (2016), and Rosell and Allen (2020) applied age cohorts and noted that young adults (respectively 31-40, 20-29, 24-29 and 35-44 years old) show a higher acceptability compared to younger and older age groups. ...
... Regarding car-sharing, some authors argue that the level of acceptability is higher among women (Ohta et al., 2013;Ullah et al., 2019;Zheng et al., 2009), while other authors claim that it is higher among men (Burghard & Dütschke, 2019;Ohta et al., 2013;Olaru et al., 2021;Te & Lianghua, 2019;. The influence on SAVs is more evident with a higher acceptability by men (Alfonsi et al., 2018;Bansal et al., 2016;Cartenì, 2020;Fafoutellis et al., 2021;Haboucha et al., 2017;König & Grippenkoven, 2020;Lavieri & Bhat, 2019;Nathanail et al., 2020;Rosell & Allen, 2020;Sener & Zmud, 2019;Wang et al., 2020) compared to women (Fafoutellis et al., 2021;Nathanail et al., 2020;Rosell & Allen, 2020;Wicki et al., 2019). According to Fafoutellis et al. (2021), Nathanail et al. (2020), and Rosell and Allen (2020), acceptability depends on the scheme with men preferring private and women public AVs. ...
... Regarding car-sharing, some authors argue that the level of acceptability is higher among women (Ohta et al., 2013;Ullah et al., 2019;Zheng et al., 2009), while other authors claim that it is higher among men (Burghard & Dütschke, 2019;Ohta et al., 2013;Olaru et al., 2021;Te & Lianghua, 2019;. The influence on SAVs is more evident with a higher acceptability by men (Alfonsi et al., 2018;Bansal et al., 2016;Cartenì, 2020;Fafoutellis et al., 2021;Haboucha et al., 2017;König & Grippenkoven, 2020;Lavieri & Bhat, 2019;Nathanail et al., 2020;Rosell & Allen, 2020;Sener & Zmud, 2019;Wang et al., 2020) compared to women (Fafoutellis et al., 2021;Nathanail et al., 2020;Rosell & Allen, 2020;Wicki et al., 2019). According to Fafoutellis et al. (2021), Nathanail et al. (2020), and Rosell and Allen (2020), acceptability depends on the scheme with men preferring private and women public AVs. ...
It is believed that shared transport services (STSs) can reduce transport poverty and social exclusion. This paper proposes a definition of “social acceptability” and “social acceptance” and examines whether vulnerable groups accept STSs. The notions “acceptability” and “acceptance” were distinguished and four necessary conditions, especially for vulnerable groups, or the 4As were identified: “availability”, “accessibility”, “affordability”, and “attractability”. In the context of STSs, “social acceptability” is defined as the degree to which an individual intends to use a STS before experiencing it in everyday travel based on the expected availability, accessibility, affordability, and attractability of the service, while “social acceptance” also incorporates the use of a STS after experiencing it in everyday travel based on a minimum level of perceived availability, accessibility, affordability, and attractability. This paper further reviews the scientific literature in transport research regarding the “acceptability” or “acceptance” of STSs by vulnerable groups. While several studies include socio-economic and demographic variables (e.g. age, gender) to explain the “acceptability” of STSs, only a few studies specifically focus on vulnerable groups. More research on the “social acceptance” of STSs, especially shared scooters, ride-sharing, and apps and Mobility as a Service (MaaS), by vulnerable groups is needed.
... Previous research has highlighted the importance of mode choice prediction in the design of strategies to promote the use of public transport and improve urban development (Hawas, Hassan, and Abulibdeh 2016). Recently, this approach has also been applied to address modern challenges, including smart city traffic management (Andrade and Gama 2022), active travel mode choice (Ali et al. 2022), the impact of introducing new travel modes (Abulibdeh 2023) and autonomous Mobility-on-Demand (MOD) services (Fafoutellis et al. 2022). ...
... In travel mode choice analysis, efforts to enhance transparency in 'black-box' SL-based models have led to various interpretability approaches. While some studies confine inference to discrete choice model results (Naseri et al. 2022), others employ model-agnostic techniques like permutation feature importance (PFI) (Hagenauer and Helbich 2017;Kim 2021;Zambang, Jiang, and Wahab 2021) and Partial Dependence (PD) plots (Fafoutellis et al. 2022;Zhao et al. 2020). PFI measures variable importance by assessing prediction error changes after feature permutation, while PD plots visualize average marginal effects of single features on the model's predictions (Molnar et al. 2023). ...
Shared autonomous vehicles (SAVs) can facilitate socially sustainable transport. Pilot projects with SAVs are steadily increasing, but it remains unclear which individuals accept SAVs and why. This research investigates the ‘social acceptability’ of SAVs through an online survey that accompanied an automated shuttle pilot on a university campus in Hannover, Germany (September–November 2022). A total of 140 respondents completed all the 41 social acceptability statements, which were evaluated using an exploratory factor analysis to identify five factors defining social acceptability: ‘social acceptability’, ‘effort expectancy’, ‘self-efficacy’, ‘safety expectancy’, and ‘performance expectancy’. A subsequent cluster analysis based on these factors suggests four social acceptability groups: ‘avoiders’, ‘resisters’, ‘self-doubters’, and ‘innovators’, though generally there is a high social acceptability towards SAVs. Significant differences between groups based on gender are identified, but not based on age, residential area, education, work, or income. By identifying the factors contributing to social acceptability and distinguishing how different groups might react to SAVs, a better understanding of social acceptability is obtained that will help prepare authorities and providers for the arrival and implementation of SAVs.
The dramatic experience due to COVID-19 spread has reshaped travel preferences of public transport (PT) users worldwide, especially in urban areas. As the PT is expected to recover its major role in such areas, it is important to understand the factors influencing PT users’ willingness to pay (WTP) for onboard safety measures, in the event of future pandemic scenarios. Furthermore, both individual latent traits (e.g. concern for the pandemic, trust/distrust in city services and national government actions) and perceived entity of the pandemic are expected to influence preferences for PT users under such a post-pandemic scenario. This paper analyses the preferences and attitudes of PT users in the Naples metropolitan area (Italy) through a hybrid choice model (HCM). First, WTPs for onboard service features are assessed in three hypothetical pandemic alert scenarios, which are explicitly introduced in the model as context variables. Second, the model allows for assessing the relative importance of onboard characteristics as the pandemic scenario evolves. Third, the model incorporates psycho-attitudinal variables and shows how they impact WTPs. Finally, several policy implications for policymakers and transport companies operating in the study area are derived. In particular: (a) WTPs for increased/reduced occupancy rate and green pass check at the entrance significantly depend upon the latent traits investigated; (b) relative importance of safety measures varies significantly between the pandemic alert scenarios; (c) possible ticketing strategies for PT users have been investigated based on the HCM findings, searching for the configuration of safety measures to ensure that users accept a 100% allowed capacity on board during moderate/high pandemic scenarios without varying the price, as well as the price variations needed to stay in an indifference range of the utility in restricted conditions of the service; (d) the acceptability of safety measures has been assessed through a simulation exercise, finding that non-vaccinated travellers are 2.6 and 2.1 times more willing to accept a full capacity of the buses/trains on board than vaccinated people if subscribers or not, respectively.
Predicting and understanding travellers’ mode choices is crucial to developing urban transportation systems and formulating traffic demand management strategies. Machine learning (ML) methods have been widely used as promising alternatives to traditional discrete choice models owing to their high prediction accuracy. However, a significant body of ML methods, especially the branch of neural networks, is constrained by overfitting and a lack of model interpretability. This study employs a neural network with feature selection for predicting travel mode choices and Shapley additive explanations (SHAP) analysis for model interpretation. A dataset collected in Chengdu, China was used for experimentation. The results reveal that the neural network achieves commendable prediction performance, with a 12% improvement over the traditional multinomial logit model. Also, feature selection using a combined result from two embedded methods can alleviate the overfitting tendency of the neural network, while establishing a more robust model against redundant or unnecessary variables. Additionally, the SHAP analysis identifies factors such as travel expenditure, age, driving experience, number of cars owned, individual monthly income, and trip purpose as significant features in our dataset. The heterogeneity of mode choice behaviour is significant among demographic groups, including different age, car ownership, and income levels.
New mobility-on-demand services together with the emerging technology of autonomous vehicles (AV) aim to revolutionize urban transportation systems, by introducing autonomous driving and sophisticated sharing and routing schemes for efficiently serving individual’s needs and requirements. On the other hand, the COVID-19 pandemic has disrupted travel patterns due to the emerging trends of social distancing and teleworking. In this paper, we aim at investigating users’ perception on autonomous vehicles, mobility on demand schemes as well as on the future transportation landscape using data collected through a questionnaire survey in the Metropolitan Area of Athens, Greece conducted after the first COVID-19 pandemic wave. First, a statistical analysis of the responses is performed and, then, a clustering approach is followed to identify user profiles based on daily mobility patterns and attitudes towards autonomous vehicles. Subsequently, the identified profiles are exploited in the development of a Bayesian Network to reveal interrelations between user profiling, attitudes and perceptions for future mobility services. Regarding the acceptance of Autonomous Mobility on Demand (AMoD) services, as well as travelers’ level of happiness concerning future scenarios of urban transportation, results have shown that the majority of travelers in Athens will be more than happy in the case where the entire transportation system is served with AMoD services.
Sweden’s strategy to manage the spread of Covid-19 has not included any form of lockdown, in contrast to the approaches adopted by most other countries. Instead, the strategy has been largely based on strong recommendations for society. Even though Sweden has not had any form of lockdown, the Covid-19 pandemic has during a relatively short period of time brought changes for society, significantly disrupting everyday life. The pandemic poses both challenges and opportunities for sustainable future transport, not least public transport provision, supply and use. The purpose of this study is to investigate how changes for society have translated into changes for mobility as an element of everyday life during the early stages of a pandemic. This study draws on a map-based online survey (public participatory GIS) which was purposefully designed to allow people to contribute with their experiences in order to capture how the current situation has affected several different facets of people’s everyday life. Results suggest that effects on mobility, such as the possibility to telework, affect different groups differently and may exacerbate existing differences in terms of gender, geography and mobility. In order to mitigate negative effects, transport policy needs to be tailored in order to take these heterogeneities into account. Both spatio-temporal adjustment and modal adjustment were dominant themes for most activities, although the dominance of these themes varied among the activities. Our findings give an indication of both the short and long-term impacts on everyday mobility in the Swedish context, for groups of inhabitants in the city of Malmö. Through deepening our understanding of the processes at play, we suggest eight possible policy responses that can be carefully tailored, both in the interim and into the future.
Background
The COVID-19 crisis has meant a significant change in the lifestyle of millions of people worldwide. With a lockdown that lasted almost three months and an impulse to new normality, transport demand has suffered a considerable impact in the Spanish case. It is mandatory to explore the effect of the pandemic on changes in travel behaviour in post-COVID-19 times.
Methodology
A nationwide survey was carried out during the lockdown in Spring 2020 to overview the recent changes. The survey collected both stated preferences (socio-demographic characteristics and mobility-related attributes), and revealed preferences (individuals’ habits, especially in the frequency of the trips according to the trip purpose, and opinions regarding the willingness and acceptability of these changes, and which actors would have to drive them, and how) of individuals. This paper aims to study and understand the willingness to adopt a set of measures to improve the safety conditions of public transport and shared mobility services against possible contagion from COVID-19 and the willingness to pay for them.
Results
The results obtained show that some measures, such as the increase of supply and vehicle disinfection, result in a greater willingness to use public transport in post-COVID-19 times. Similarly, the provision of covers for handlebars and steering wheels also significantly increases individuals’ willingness to use sharing services. However, respondents expect that these measures and improvements would be implemented but maintaining the same pre-COVID-19 prices. The results of this research might help operators deploy strategies to adopt their services and retain users.
Demand for different modes of transportation clearly interacts. If public transit is delayed or out of service, customers might use mobility on demand (MoD), including taxi and carsharing for their trip, or discard the trip altogether, including a first and last mile that might otherwise be covered by MoD. For operators of taxi and carsharing services, as well as dispatching agencies, understanding increasing demand, and changing demand patterns due to outages and delays is important, as a more precise demand prediction allows for them to more profitably operate. For public authorities, it is paramount to understand this interaction when regulating transportation services. We investigate the interaction between public transit delays and demand for carsharing and taxi, as measured by the fraction of demand variance that can be explained by delays and the changing OD-patterns. A descriptive analysis of the public transit data set yields that delays and MoD demand both highly depend on the weekday and time of day, as well as the location within the city, and that delays in the city and in consecutive time intervals are correlated. Thus, demand variations must by corrected for these external influences. We find that demand for taxi and carsharing increases if the delay of public transit increases and this effect is stronger for taxi. Delays can explain at least 4.1% (carsharing) and 18.8% (taxi) of the demand variance, which is a good result when considering that other influencing factors, such as time of day or weather exert stronger influences. Further, planned public transit outages significantly change OD-patterns of taxi and carsharing.
We present a brief history of the field of interpretable machine learning (IML), give an overview of state-of-the-art interpretation methods and discuss challenges. Research in IML has boomed in recent years. As young as the field is, it has over 200 years old roots in regression modeling and rule-based machine learning, starting in the 1960s. Recently, many new IML methods have been proposed, many of them model-agnostic, but also interpretation techniques specific to deep learning and tree-based ensembles. IML methods either directly analyze model components, study sensitivity to input perturbations, or analyze local or global surrogate approximations of the ML model. The field approaches a state of readiness and stability, with many methods not only proposed in research, but also implemented in open-source software. But many important challenges remain for IML, such as dealing with dependent features, causal interpretation, and uncertainty estimation, which need to be resolved for its successful application to scientific problems. A further challenge is a missing rigorous definition of interpretability, which is accepted by the community. To address the challenges and advance the field, we urge to recall our roots of interpretable, data-driven modeling in statistics and (rule-based) ML, but also to consider other areas such as sensitivity analysis, causal inference, and the social sciences.
Various measures were recommended or imposed by the governments to control the spread of COVID-19. Travel behaviors are significantly influenced due to such measures. However, people have various travel needs ranging from grocery shopping to work. This study examines the changes that occurred in travel behavior due to the COVID-19 pandemic. Data were collected through an online questionnaire survey that included questions on trip purpose, mode choice, distance traveled, and frequency of trips before and during COVID-19. 1,203 responses were collected from various countries around the world.
Results explained that trip purpose, mode choice, distance traveled, and frequency of trips for the primary travel were significantly different before and during the pandemic. Further, the majority of trips were made for shopping during the pandemic. There was a significant shift from public transport to private transport and non-motorized modes. People placed a higher priority on the pandemic related concerns while choosing a mode during the pandemic as compared to the general concerns. Gender, car ownership, employment status, travel distance, the primary purpose of traveling, and pandemic-related underlying factors during COVID-19 were found to be significant predictors of mode choice during the pandemic.
Outcomes of this study could be useful in transport planning and policymaking during pandemics based on the travel needs of people. In particular, government authorities could utilize such knowledge for planning smart and partial lockdowns. Service providers, e.g., taxi companies and retailers, could use such information to better plan their services and operations.
Simulation studies suggest that pooled on-demand services (also referred to as Demand Responsive Transport, ridesharing, shared ride-hailing or shared ridesourcing services) have the potential to bring large benefits to urban areas while inducing limited time losses for their users. However, in reality, the large majority of users request individual rides (and not pooled rides) in existing on-demand services, leading to increases in motorised vehicle miles travelled. In this study, we investigate to what extent fare discounts, additional travel time, and the (un)willingness to share the ride with (different numbers of) other passengers play a role in the decision of individuals to share rides. To this end, we design a stated preference study targeting Dutch urban individuals. In our research, we (1) disentangle the sharing aspect from related time–cost trade-offs (e.g. detours), (2) investigate preference heterogeneity regarding the studied attributes and identify distinct market segments, and (3) simulate scenarios to understand the impact of the obtained parameters in the breakdown between individual and pooled services. We find that less than one third of respondents have strong preferences against sharing their rides. Also, we find that different market segments vary not only in their values of the willingness to share, but also in how they perceive this willingness to share (per-ride or proportional to the in-vehicle time). Further, the scenario analysis demonstrates that the share of individuals who are willing to share rides depends primarily on the time–cost trade-offs, rather than on the disutility stemming from pooling rides per se.
The spread of the COVID-19 virus has resulted in unprecedented measures restricting travel and activity participation in many countries. Social distancing, i.e., reducing interactions between individuals in order to slow down the spread of the virus, has become the new norm. In this viewpoint I will discuss the potential implications of social distancing on daily travel patterns. Avoiding social contact might completely change the number and types of out-of-home activities people perform, and how people reach these activities. It can be expected that the demand for travel will reduce and that people will travel less by public transport. Social distancing might negatively affect subjective well-being and health status, as it might result in social isolation and limited physical activity. As a result, walking and cycling, recreationally or utilitarian, can be important ways to maintain satisfactory levels of health and well-being. Policymakers and planners should consequently try to encourage active travel, while public transport operators should focus on creating ways to safely use public transport.
The rapid development of autonomous vehicles (AV) in recent years has drawn the attention of numerous countries in terms of its feasibility for use and deployment as individually-owned vehicles or for large-scale fleet planning and deployment as a mobility-on-demand (MOD) service. Singapore is no exception to this global trend and in her pursuit to be smart and car-lite, numerous efforts are made to have AV trials in place and test out their potential deployment in the city state. As one of the many prerequisites of AV planning, public perception on AV plays a vital role when designing any potential AV deployment scheme. As such, a stated preference survey comprising both online survey and field interviews/surveys, was performed island-wide to understand how commuters in Singapore perceive about different AV-based MOD modes. The logit kernel model is adopted to determine how different preference attributes and key demographic indicators can affect the use of AV-based MOD services over other existing first- and last-mile connection modes. The model results have identified how demographics such as gender, age, housing type, education level and income level can influence the travel mode choice. Also, the impacts brought by individuals’ stated preferences over convenience, privacy and familiarity of ride-hailing apps are also investigated. Such findings can provide useful insight in planning future car-lite towns and implementing AV-based MOD services in these towns.
On-demand ridesplitting is a form of ridesourcing where riders with similar origins and destinations are matched to the same driver and vehicle in real time, and the ride and costs are split among users. With the convenience of all kinds of ridesourcing services, the number of ridesplitting passengers increases, which may have a great impact on the urban mobility. In this paper, we analyze ridesplitting behavior and its impact on multimodal mobility, e.g., vehicle kilometers traveled (VKT) and transportation modal shift, using real-world ridesourcing data extracted from an on-demand ride service platform in Hangzhou, China, and questionnaires filled by on-demand ridesplitting passengers. With the consideration of the VKT shifted from non-passenger/private vehicles, this paper uses the saved VKT of two ridesplitting types, e.g., DiDi Hitch and DiDi Express ridesplitting, to quantify the ridesplitting impact. For the whole ridesourcing ecosystem, ridesplitting is estimated to decrease 58,124 VKT per day in Hangzhou, of which Hitch and Express ridesplitting contribute 2175 km and 55,949 km per day, respectively. The saved VKT of Hitch is much smaller than Express ridesplitting for the following two reasons: (1) Hitch orders are fewer than Express ridesplitting; (2) more than half of the Hitch passengers shift modes from bus/metro transit or other non-passenger/private cars. This paper shines some lights on understanding the emerging on-demand ridesplitting behavior and quantifying its impact on multimodal urban mobility.
Inspired by the success of private ridesourcing companies such as Uber and Lyft, transit agencies have started to consider integrating ridesourcing services (i.e. on-demand, app-driven ridesharing services) with public transit. Ridesourcing services may enhance the transit system in two major ways: replacing underutilized routes to improve operational efficiency, and providing last-mile connectivity to extend transit’s catchment area. While an integrated system of ridesourcing services and public transit is conceptually appealing, little is known regarding whether and how consumers might use a system like this and what key service attributes matter the most to them. This article investigates traveler responses to a proposed integrated transit system, named MTransit, at the University of Michigan Ann Arbor campus. We conducted a large-sample survey to collect both revealed preference (RP) and stated preference (SP) data and fit a RP-SP mixed logit model to examine the main determinants of commuting mode choice. The model results show that transfers and additional pickups are major deterrents for MTransit use. We further applied the model outputs to forecast the demand for MTransit under different deployment scenarios. We find that replacing low-ridership bus lines with ridesourcing services could slightly increase transit ridership while reducing operations costs. The service improvements offered by ridesourcing mainly come from reductions in wait time. Though relatively small in our study, another source of improvement is the decrease of in-vehicle travel time. Moreover, we find that when used to provide convenient last-mile connections, ridesourcing could provide a significant boost to transit. This finding verifies a popular notion among transit professionals that ridesourcing services can serve as a complement to public transit by enhancing last-mile transit access.
There is considerable interest in modeling and forecasting the impacts of autonomous vehicles
on travel behavior and transportation network performance. In an autonomous vehicle future,
individuals may privately own such vehicles, or use mobility-on-demand services provided by
transportation network companies that operate shared autonomous vehicle fleets or adopt a
combination of these two. This paper presents a comprehensive model system of autonomous
vehicle adoption and use. A Generalized Heterogeneous Data Model system is estimated on data
collected as part of the Puget Sound Regional Travel Study. Results show that lifestyle factors
play an important role in shaping autonomous vehicle usage. Younger urban residents who are
more educated and tech-savvy are more likely to be early adopters of autonomous vehicle
technologies, favoring a sharing-based service model over private ownership. Models such as
that presented in this paper can be used to predict adoption of autonomous vehicle technologies,
which will in turn help forecast autonomous vehicle impacts under alternative future scenarios.
In this paper, we study clustering with respect to the k-modes objective function, a natural formulation of clustering for categorical data. One of the main contributions of this
paper is to establish the connection between k-modes and k-median, i.e., the optimum of k-median is at most the twice the optimum of k-modes for the same categorical data clustering problem. Based on this observation, we derive a deterministic algorithm that
achieves an approximation factor of 2. Furthermore, we prove that the distance measure in k-modes defines a metric. Hence, we are able to extend existing approximation algorithms for metric k-median to k-modes. Empirical results verify the superiority of our method.
Random forests are becoming increasingly popular in many scientific fields because they can cope with "small n large p" problems, complex interactions and even highly correlated predictor variables. Their variable importance measures have recently been suggested as screening tools for, e.g., gene expression studies. However, these variable importance measures show a bias towards correlated predictor variables.
We identify two mechanisms responsible for this finding: (i) A preference for the selection of correlated predictors in the tree building process and (ii) an additional advantage for correlated predictor variables induced by the unconditional permutation scheme that is employed in the computation of the variable importance measure. Based on these considerations we develop a new, conditional permutation scheme for the computation of the variable importance measure.
The resulting conditional variable importance reflects the true impact of each predictor variable more reliably than the original marginal approach.
In this paper, we study clustering with respect to the k-modes objective function, a natural formulation of clustering for categorical data. One of the main contributions of this paper is to establish the connection between k-modes and k-median, i.e., the optimum of k-median is at most twice the optimum of k-modes for the same categorical data clustering problem. Based on this observation, we derive a deterministic algorithm that achieves an approximation factor of 2. Furthermore, we prove that the distance measure in k-modes defines a metric. Hence, we are able to extend existing approximation algorithms for metric k-median to k-modes. Empirical results verify the superiority of our method.
The COVID-19 outbreak led to significant changes in daily commuting. As lockdowns were imposed to metropolitan areas throughout the globe, travelers refrained heavily from using public transport, to maintain social distancing. Based on data from Athens, Greece, this paper investigates the anticipated, post-pandemic behavior of travelers with respect to public transport use. Focus is given on analyzing those factors that affect post-pandemic recovery time of public transport users, i.e. the time travelers would refrain from using public transport, following a gradual exit from the pandemic outbreak and relaxation of lockdowns. The analysis is performed using both a clustering algorithm and a discrete duration model. Both methodologies highlighted the fact that the frequency of using public transport before the pandemic along with the travelers’ age, influence their behavior in terms of recovery time. Results from the discrete duration model suggest also that self-employed and travelers who mostly use private vehicles, are less likely to use public transport after the outbreak. Concerning the psychological factors that shape COVID-19 safety-related perceptions that affect public transport use, travelers who would be willing to use protection gear when traveling with are also less likely to return to public transport. Findings of this study could be useful for policy making, suggesting that efficient marketing strategies toward promoting public transport usage in a post-pandemic era should focus on travelers with specific socio-demographic and travel characteristics.
This paper presents a novel deep learning-based travel behaviour choice model. Our proposed Residual Logit (ResLogit) model formulation seamlessly integrates a Deep Neural Network (DNN) architecture into a multinomial logit model. Recently, DNN models such as the Multi-layer Perceptron (MLP) and the Recurrent Neural Network (RNN) have shown remarkable success in modelling complex and noisy behavioural data. However, econometric studies have argued that machine learning techniques are a ‘black-box’ and difficult to interpret for use in the choice analysis. We develop a data-driven choice model that extends the systematic utility function to incorporate non-linear cross-effects using a series of residual layers and using skipped connections to handle model identifiability in estimating a large number of parameters. The model structure accounts for cross-effects and choice heterogeneity arising from substitution, interactions with non-chosen alternatives and other effects in a non-linear manner. We describe the formulation, model estimation, interpretability and examine the relative performance and econometric implications of our proposed model. We present an illustrative example of the model on a classic red/blue bus choice scenario example. For a real-world application, we use a travel mode choice dataset to analyze the model characteristics compared to traditional neural networks and Logit formulations. Our findings show that our ResLogit approach significantly outperforms MLP models while providing similar interpretability as a Multinomial Logit model.
The paper investigates the choice of the transport mode to access the city center of Trieste, Italy. We have collected revealed and stated choice data and attitudinal data in order to investigate the mode of transport used before and during the Covid-19 pandemic. We studied the main choice determinants jointly with respondents’ psychological attitudes using an integrated choice/latent variable modelling framework. The model incorporates choice data, generalized cost variables, random parameters, and latent variables. The latter capture the concern for the environment, both at global and local level, the attitudes towards physical exercise and the Covid-19-related risks. The model is applied to assess the impact of extending the density of the cycling lanes on active mobility. Our estimates indicate a potential increase in cycling which, however, would not translate into an increase in active mobility, since the bike would substitute some trips currently made by foot. We also detect a high substitutability between the bike and the bus, while the motorcycle and the car are much less affected. It is confirmed that the Covid-19 pandemic altered significantly the transport mode choices, having a strong negative impact on bus and shifting bus users towards private modes, both motorized and non-motorized.
As a result of the coronavirus pandemic, in spring 2020 numerous protective measures were taken in Germany and all over the world. This has changed our everyday life and our mobility considerably. It is in question whether and how the pandemic and the lockdown have impacted transport mode use, attitudes towards transport modes and the ownership of individual mobility options during the lockdown period. In order to shed light on these essential aspects of transport policy, we carried out a representative travel survey in Germany during the strictest period of lockdown in the beginning of April. We have analysed overall and individual changes in transport mode usage and attitudes towards transport modes, focussing on the bicycle, the car and public transport. Also, the changes in the perception of individual mobility options with a focus on car-free households were investigated. Our results indicate that public transport lost ground during the particularly restricted period of lockdown while individual modes of transport, especially the private car, became more important. Our findings are highly relevant for transport policy when developing measures for expanding the possibilities for sustainable individual transport and developing concepts that strengthen public transport. These aspects are key for achieving a sustainable transport system in the medium- and long-term despite the the coronavirus pandemic.
To better understand the young people’s ridesplitting behavior characteristics and behavioral impacts on the emerging ride-sourcing platform, we present a survey-based comparison of DiDi Hitch and DiDi Express Pool that provide ridesplitting services in Hangzhou, China. Hitch is a social carpooling platform that helps commuters share rides. Express Pool is a mobility network providing pooled rides as the ridesplitting option of Express. In this paper, we explore why ridesplitting is popular, which mode is the primary, and how ridesplitting affects young people’s travel behavior. A total of 607 investigated responses were collected via the DiDi ride-sourcing platform. Based on the survey data, a variety of behavioral aspects were analyzed, including the trip purpose, temporal and spatial characteristics of trips, travel time/cost variation, and modal shift frequency. Behavioral impacts on modal shift were further analyzed in a binary logit framework. The findings show that Hitch and Express Pool have both similarities and differences. Hitch is intended for long trips, while Express Pool is mainly intended for short trips as a supplement of multimodal mobility. It also provides evidence for ridesplitting user identification and the market share loss of bus and taxi. Some suggestions are raised for discussions in future research.
With the era of fully automated vehicles (AVs) quickly approaching, ridesharing services could have an important role in increasing vehicle occupancy, reducing vehicle miles traveled, and improving traffic conditions. However, the extent to which these potentials can be achieved depends on consumers’ disposition to sharing rides. From a travel behavior perspective, two essential elements to the adoption of shared rides are individuals’ acceptance of increased travel times associated with pick-up/drop-off of other passengers and their approval of strangers sharing the same vehicle. The current study develops the notion of willingness to share (WTS), which represents the money value attributed by an individual to traveling alone compared to riding with strangers, to investigate the adoption of shared rides. Using a multivariate integrated choice and latent variable approach, we examine current choices and future intentions regarding the use of shared rides and estimate individuals’ WTS as well as their values of travel time for two distinct trip purposes. Results show that users are less sensitive to the presence of strangers when in a commute trip compared to a leisure-activity trip. We also observe that the travel time added to the trip to serve other passengers may be a greater barrier to the use of shared services compared to the presence of a stranger. However, the potential to use travel time productively may help overcome this barrier especially for high-income individuals.
This paper presents the preliminary results of a recent nationwide survey that focuses on mode choice behavior in view of emerging mobility options such as ride-sourcing and automated vehicles (AV). The survey provides a comprehensive scan of current mode choice patterns and the influencing factors. Then it presents stated preference (SP) choices to understand how travelers measure the trade-offs among different mode alternatives. The choice experiments focused on four potential user markets: drivers who usually drive for daily activities, passengers who depend on other household members or friends, transit users or users who do not have access to a private vehicle on a regular basis, and visitors or who do not have access to a private vehicle occasionally. The results suggest that on-demand services incorporating AV technologies (with lower operating costs) may become a viable option for many travelers. Most drivers and passengers preferred single ride than shared ride regardless of whether it is on a daily or occasional basis. However, for transit users, shared rides showed higher potential than exclusive services, which may indicate that cost is a primary consideration in the mode choice decisions for these users.
Enabled by mobile technologies and fueled by the economic downturn, ridesharing has emerged in recent years as a private transportation facet of the shared economy. Our study investigates the motives for participation in situated ridesharing. We propose a theoretical model that includes economic benefits, time benefits, transportation anxiety, trust, and reciprocity either as direct antecedents of ridesharing participation intention, or mediated through attitude towards ridesharing. We conduct a scenario-based survey, with 300 participants. Our findings indicate that, in situations where transportation anxiety is high (e.g. construction on the road), if people can trust the ridesharing service providers and participants, in the presence of economic and time benefits, they will have a strong intention to participate in ridesharing.
This study gains insight into individual motivations for choosing to own and use autonomous vehicles and develops a model for autonomous vehicle long-term choice decisions. A stated preference questionnaire is distributed to 721 individuals living across Israel and North America. Based on the characteristics of their current commutes, individuals are presented with various scenarios and asked to choose the car they would use for their commute. A vehicle choice model which includes three options is estimated:
(1) Continue to commute using a regular car that you have in your possession.
(2) Buy and shift to commuting using a privately-owned autonomous vehicle (PAV).
(3) Shift to using a shared-autonomous vehicle (SAV), from a fleet of on-demand cars for your commute.
A factor analysis determined five relevant latent variables describing the individuals’ attitudes: technology interest, environmental concern, enjoy driving, public transit attitude, and pro-AV sentiments. The effects that the characteristics of the individual and the autonomous vehicle have on use and acceptance are quantified through random utility models including logit kernel model taking into account panel effects.Currently, large overall hesitations towards autonomous vehicle adoption exist, with 44% of choice decisions remaining regular vehicles. Early AV adopters will likely be young, students, more educated, and spend more time in vehicles. Even if the SAV service were to be completely free, only 75% of individuals would currently be willing to use SAVs. The study also found various differences regarding the preferences of individuals in Israel and North America, namely that Israelis are overall more likely to shift to autonomous vehicles.
Methods to encourage SAV use include increasing the costs for regular cars as well as educating the public about the benefits of shared autonomous vehicles.
In the recent years many developments took place regarding automated vehicles (AVs) technology. It is however unknown to which extent the share of the existing transport modes will change as result of AVs introduction as another public transport option. This study is the first where detailed traveller preferences for AVs are explored and compared to existing modes. Its main objective is to position AVs in the transportation market and understand the sensitivity of travellers towards some of their attributes, focusing particularly on the use of these vehicles as egress mode of train trips. Because fully-automated vehicles are not yet a reality and they entail a potentially high disruptive way on how we use automobiles today, we apply a stated preference experiment where the role of attitudes in perceiving the utility of AVs is particularly explored in addition to the classical instrumental variables and several socio-economic variables. The estimated discrete choice model shows that first class train travellers on average prefer the use of AVs as egress mode, compared to the use of bicycle or bus/tram/metro as egress. We therefore conclude that AVs as last mile transport between the train station and the final destination have most potential for first class train travellers. Results show that in-vehicle time in AVs is experienced more negatively than in-vehicle time in manually driven cars. This suggests that travellers do not perceive the theoretical advantage of being able to perform other tasks during the trip in an automated vehicle, at least not yet. Results also show that travellers’ attitudes regarding trust and sustainability of AVs are playing an important role in AVs attractiveness, which leads to uncertainty on how people will react when AVs are introduced in practice. We therefore state the importance of paying sufficient attention to these psychological factors, next to classic instrumental attributes like travel time and costs, before and during the implementation process of AVs as a public transport alternative. We recommend the extension of this research to revealed preference studies, thereby using the results of field studies.
Shared autonomous vehicles (SAVs) could provide inexpensive mobility on-demand services. In addition, the autonomous vehicle technology could facilitate the implementation of dynamic ride-sharing (DRS). The widespread adoption of SAVs could provide benefits to society, but also entail risks. For the design of effective policies aiming to realize the advantages of SAVs, a better understanding of how SAVs may be adopted is necessary. This article intends to advance future research about the travel behavior impacts of SAVs, by identifying the characteristics of users who are likely to adopt SAV services and by eliciting willingness to pay measures for service attributes. For this purpose, a stated choice survey was conducted and analyzed, using a mixed logit model. The results show that service attributes including travel cost, travel time and waiting time may be critical determinants of the use of SAVs and the acceptance of DRS. Differences in willingness to pay for service attributes indicate that SAVs with DRS and SAVs without DRS are perceived as two distinct mobility options. The results imply that the adoption of SAVs may differ across cohorts, whereby young individuals and individuals with multimodal travel patterns may be more likely to adopt SAVs. The methodological limitations of the study are also acknowledged. Despite a potential hypothetical bias, the results capture the directionality and relative importance of the attributes of interest.
In the field of transportation, data analysis is probably the most important and widely used research tool available. In the data analysis universe, there are two ‘schools of thought’; the first uses statistics as the tool of choice, while the second – one of the many methods from – Computational Intelligence. Although the goal of both approaches is the same, the two have kept each other at arm’s length. Researchers frequently fail to communicate and even understand each other’s work. In this paper, we discuss differences and similarities between these two approaches, we review relevant literature and attempt to provide a set of insights for selecting the appropriate approach.Research highlights► In the field of transportation, data analysis is probably the most important and widely used research tool available. ► Differences and similarities between two ‘schools of thought’ – Statistics and Computational Intelligence – are revealed and discussed. ► Relevant literature in transportation research is reviewed and critically analyzed. ► A set of insights for selecting the appropriate approach for transportation applications is provided.
The adoption of congestion pricing depends fundamentally upon drivers’ willingness to pay to reduce travel time during the congested morning peak period. Using revealed preference data from a congestion pricing demonstration project in San Diego, we estimate that willingness to pay to reduce congested travel time is higher than previous stated preference results. Our estimate of median willingness to pay to reduce commute time is roughly $30 per hour, although this may be biased upward by drivers’ perception that the toll facility provides safer driving conditions. Drivers also use the posted toll as an indicator of abnormal congestion and increase their usage of the toll facility when tolls are higher than normal.
Long, medium, and short-term effects of COVID-19 on mobility and lifestyle
Jan 2021
de Palma
Shared automated vehicles: Review of business models
A Stocker
S Shaheen
Greedy Function Approximation : A Gradient Boosting Machine Author (s): Jerome H . Friedman Source : The Annals of Statistics
Oct 2001
ANN STAT
1189-1232
J Friedman
Friedman, J. (2001). Greedy Function Approximation : A Gradient Boosting Machine Author (s): Jerome H. Friedman Source : The Annals of Statistics, Vol.
29, No. 5 (Oct., 2001), pp. 1189-1232 Published by : Institute of Mathematical Statistics Stable URL : http://www. The Annals of Statistics, 29(5), 1189-1232. https://doi.org/10.1214/009053606000000795.
Long, medium, and short-term effects of COVID-19 on mobility and lifestyle. THEMA Working Papers
Jan 2021
A De Palma
S Vosough
de Palma, A., Vosough, S., 2021. Long, medium, and short-term effects of COVID-19 on mobility and lifestyle. THEMA Working Papers. https://ideasrepec.
org/p/ema/worpap/2021-06.html.
Jan 2022
752-766
P Fafoutellis
E G Mantouka
E I Vlahogianni
P. Fafoutellis, E.G. Mantouka, E.I. Vlahogianni et al.
International Journal of Transportation Science and Technology 11 (2022) 752-766