With the increasing frequency of natural hazards worldwide, developing effective evacuation strategies is crucial to reducing their adverse impacts. This study examines the determinants of evacuation mode choice in the event of an imminent volcanic eruption in Auckland, New Zealand, using data from a stated preference survey. Key influencing factors, including socio-demographics, risk perception, evacuation timing, and destination choice, are analysed using logistic regression and machine learning models (Random Forest, Support Vector Machine, Probabilistic Neural Network, and Classification and Regression Tree). The results identify vehicle ownership, evacuation timing, and route/destination preferences as critical determinants. Among the models, Random Forest demonstrates the highest predictive accuracy, outperforming logistic regression. The findings provide valuable insights for emergency planners and policymakers to design more resilient and adaptive evacuation strategies.
Pedestrian safety perception significantly influences their route choice decisions, especially among those who rely solely on walking due to financial constraints. These pedestrians, often referred as captive pedestrians, are widely seen in many low- and middle-income countries. Despite their prevalence, little is known about how safety perceptions influence their route choice decisions. This study investigates the influence of safety perceptions on the route choice decisions of captive pedestrians in New Delhi, India. Using intercept surveys, respondents were classified as captive or choice pedestrians and their safety perceptions assessed across 21 walking and crossing scenarios. While both groups showed similar perceptions for most scenarios, significant differences were seen with respect to preference for company while walking, presence of hawkers, and footpath buffers to avoid carriageway spillovers. Differences were also observed across gender and age groups. To understand how these perceptions influence route choice, route data was collected from captive pedestrians. Using reported and alternative routes (modelled on ArcGIS), a route choice model was estimated. Results show that route length, presence of footpaths, and green buffers significantly influenced their choices. These findings highlight the distinct safety concerns and preferences of captive pedestrians, underlining the need for targeted infrastructure designs that address their distinct safety concerns and boost route selection.
The transition toward autonomous vehicles (AVs) is expected to substantially alter on-street parking demand and traffic conditions in urban areas, particularly due to increased pick-up and drop-off (PUDO) activities. This study investigates how different on-street parking configurations influence parking choices and overall travel efficiency under mixed traffic conditions involving both AVs and human-driven vehicles (HVs). A Grey Entropy-based model is developed to model the choice between regular and PUDO spaces. Specifically, AVs can either choose regular spaces like HVs or adopt self-driving as a substitute for long-term parking while relying on PUDO areas for passenger exchange. An agent-based simulation framework is employed to capture interactions between vehicle behaviors, space utilization, and system performance across varying AV penetration levels. To examine alternative management strategies, a simulated annealing (SA) algorithm is applied to optimize configuration of PUDO spaces with the objective of minimizing total vehicle travel time (VTT). Two configuration strategies are compared through numerical experiments. The proportional correspondence strategy (PCS) is used as a baseline strategy, in which the share of PUDO spaces is configured in proportion to the AV penetration rate. The static configuration (SC) strategy uses the simulation-optimization framework to determine a fixed regular/PUDO space configuration for each scenario. The results indicate that the SC strategy substantially reduces VTT compared with PCS, with reductions of more than 38
Residential relocation can disrupt habitual behaviours, offering a window of opportunity for encouraging sustainable lifestyle changes. This study tested the effectiveness of a personalised, web-based intervention aimed at reducing travel- and energy-related emissions among recent movers in Norway (n = 212). The intervention, delivered three months post-relocation, provided tailored feedback on participants’ carbon footprints and interactive behavioural suggestions. Outcomes included weekly car use, overnight car trips, flight frequency, and electricity consumption. Participants who received the intervention made significantly fewer weekly car trips (− 20
Disruptive events, ranging from extreme weather events to global pandemics, inevitably reshape the way people live, travel, and interact with their environments. Understanding both temporary shifts and longer-term behavioral adaptations in response to disruptions is critical for planning resilient transportation systems that can adapt to evolving mobility patterns. This study leverages longitudinal Point of Interest (POI) data from 15,720 unique panelists in the United States between February 2020 and May 2022 to uncover the multi-year temporal dynamics of individual activity-travel patterns in response to the COVID-19 pandemic. Applying spectral clustering techniques to a comprehensive set of mobility indicators, we identify six distinct mobility styles: anchored dweller, off-peak mover, morning commuter, home-based traveler, dynamic explorer, and frequent trip-maker. The mobility styles approach acts as a dimensionality reduction of complex patterns into interpretable profiles that can be tracked and compared over time, capturing variations in activity patterns, spatial habits, and schedule habits. Our findings reveal that the temporary shift toward the anchored dweller mobility style peaked during the height of pandemic restrictions but gradually weakened over time despite remaining higher than pre-pandemic levels. By mid-2022, higher activity styles have rebounded beyond pre-pandemic levels, while high levels of at-home dwell times persist, suggesting structural changes to activity-travel behavior. We also observe heterogeneity in transitions between mobility styles, with higher-income individuals more likely to transition to lower-mobility styles during the peak of pandemic restrictions. These results underscore the uneven impacts of disruptions across different segments of the population. The study introduces an approach that simplifies complex activity-travel patterns into interpretable mobility styles, providing a structured framework for studying resilience and adaptation in response to disruptions.
Mobility as a Service (MaaS) is a promising approach to addressing urban transportation challenges by promoting the use of public transportation. For the sustainability of MaaS, beyond retaining existing users, it is more important to attract potential users who shift from private vehicle travel. To scale up MaaS adoption, it is important to identify the heterogeneity among potential users who are aware of an operating MaaS scheme but have not yet adopted it. Using data from 498 non-users of Taiwan’s TPASS, a nationwide operational public transport pass scheme, this study employs latent profile analysis to identify distinct segments based on five psychological constructs: perceived economic benefits, relative attitude toward private versus public transport, habitual travel behavior, environmental concern, and adoption intention. Four heterogeneous profiles emerged: flexible multimodal commuters, economic-oriented commuters, private transport enthusiasts, and sustainable transport supporters. The profiles differ substantially across the five constructs, with habitual vehicle use and relative attitude most clearly separating car-oriented from transit-receptive segments. Current commuting mode, vehicle ownership, and commuting distance systematically predict profile membership. These findings underscore the importance of recognizing user diversity in operational MaaS contexts and offer practical implications for developing segment-specific marketing and policy interventions to promote MaaS uptake.
Once considered “wasted”, travel time has increasingly evolved into a time–space locus for activities. In this context, travel time use may influence how people organise daily activities and impact how individuals travel. Yet, travel time use has been investigated in isolation from the broader time-use context, so we remain unaware of how travel-based and out-of-trip activities complement or substitute for one another. This research investigates the relationships between travel-based and out-of-trip activities, taking “activity transfer” (i.e., engaging in travel-based activities to free up time outside the journey) into consideration. To model time use while accounting for such relationships, we calibrate an extended multiple discrete continuous (eMDC) model using revealed-preference data from a sample that represents the adult population living in Greater Melbourne and Geelong (Australia). Our results suggest that travel-based and out-of-trip mandatory and maintenance activities tend to complement each other. By contrast, out-of-trip and travel-based discretionary activities exhibit a substitution relationship. Activity transfer seems to weaken complementarity and strengthen substitution relationships between travel-based and out-of-trip activities. These results suggest that relationships that hold in the broader time-use context also hold within trips, given that travel time accommodates excess demand from high-priority activities and creates a space–time locus for those experiencing leisure-time scarcity. Our findings have implications for transport practice and policy, as they highlight the need for tailored interventions to cater for the activity needs of specific population segments.
Despite growing interest in mobility and well-being in older age, the relationship between public transport attributes, travel satisfaction, and quality of life has not been sufficiently examined within one integrated analytical framework. This study examines how public transport attributes influence travel satisfaction and quality of life (QoL) among older adults in urban environments. Data were collected from 478 individuals aged 60 + across five Polish cities using online and telephone surveys. The analysis focused on five dimensions of the travel experience: transit stop amenities, service provision, in-vehicle experience, fare affordability, and comfort. Travel satisfaction was measured using the multidimensional Satisfaction with Travel Scale (STS), integrating both cognitive and affective aspects, while QoL was assessed using subjective life satisfaction indicators. Partial Least Squares Structural Equation Modeling (PLS-SEM) revealed that comfort and the travel environment were the strongest predictors of travel satisfaction, which in turn was positively associated with QoL Socioeconomic factors, such as education and income, were also positively associated with QoL. The findings extend existing models by incorporating age-specific attributes and highlighting the importance of experiential dimensions of travel. Practical recommendations include upgrading stop infrastructure, reducing crowding, enhancing vehicle conditions, and fostering respectful driver–passenger interactions. Although the study focuses on Polish contexts, the results have broader implications for aging societies worldwide, particularly in regions facing similar socio-demographic and infrastructural challenges. The study underscores the need for inclusive, user-centered transport policies that move beyond efficiency and functionality, ensuring that mobility fosters active, fulfilling, and sustainable lives for older adults.
Understanding how everyday mobility relates to subjective well-being remains an important issue in urban and transport research. This study examines the relationship between mobility exposure, travel experience, and life satisfaction through the integrative analytical lens of Urban Mobility Burden and Opportunity (UMBO). Using a combination of survey data and passive mobility tracking from Google Maps Timeline, this study analyzes the mobility patterns of 693 urban youth, most of whom were university students, with several senior high school students also included in the sample. The dataset includes 441,649 observed trips recorded over a longitudinal observation period of more than 150 days. Partial Least Squares Structural Equation Modeling (PLS-SEM) is employed to examine the relationships among mobility exposure variables, including distance, in-vehicle travel time, and daily trip frequency; travel experience variables, including travel stress, time pressure, social connectedness, and travel satisfaction; and life satisfaction. The results provide partial support for the UMBO framework. The burden pathway is clearly reflected in the associations between mobility exposure, Travel Stress, Time Pressure, Travel Satisfaction, and Life Satisfaction. The opportunity pathway is mainly reflected in the positive association between Daily Trips and Travel Satisfaction, while the expected pathway from Social Connectedness to Travel Satisfaction was not statistically supported. Notably, higher daily trip frequency is associated with greater Travel Satisfaction and indirectly with higher Life Satisfaction, suggesting that more frequent short trips may reflect greater activity participation rather than increased mobility burden. These findings highlight the importance of considering travel experiences when evaluating the relationship between urban mobility and well-being, particularly in rapidly motorizing Global South contexts where public transport provision is limited and daily mobility remains strongly dependent on motorized individual travel.
Flex-route transit service provides a hybrid alternative between conventional fixed-route transit and demand-responsive service, particularly in low-density suburban areas where demand is spatially dispersed. A key planning challenge is determining how much slack time should be added to the schedule to accommodate deviations while limiting negative impacts on existing fixed-stop users. This paper develops an analytical framework for early-stage flex-route service planning that links slack-time allocation, operator costs, fixed-stop passenger costs, and mitigation policies. The framework derives a segment-based slack-time expression under idealized rectilinear and uniform-demand assumptions and embeds this expression in a stakeholder impact model. Two service-transition scenarios are analyzed: maintaining all existing fixed stops and reducing the number of fixed stops. The framework further identifies the fare reductions and/or frequency improvements required to offset the additional generalized cost imposed on fixed-stop users. A hypothetical numerical study and sensitivity analysis illustrate how service area width, demand intensity, slack time, and passenger value of time affect request accommodation, user burden, and mitigation requirements. Additional robustness experiments quantify the trade-off between segment-based and pooled slack, demonstrate the sensitivity of required slack to near-route and outer-edge demand clustering, and show that limited backtracking provides modest acceptance gains at higher request levels. The results show that wider service areas and higher on-demand request volumes require greater slack time, but additional slack also increases costs for operators and fixed-stop users. The proposed framework provides transit planners with a transparent tool for evaluating flex-route feasibility and balancing service flexibility with schedule reliability and passenger equity.
Electric vehicles (EVs) are rapidly being adopted as part of the global shift toward sustainable transport. Although they provide environmental and technological advantages, their safety impacts compared to internal combustion engine vehicles (ICEVs) remain unclear. This paper uses a Social Justice lens to examine how the EV transition shapes equity and safety outcomes. Through a systematic literature review, 18 studies were identified from 2761 records, analysing crash risk, injury severity, and related harms across population groups and spatial contexts. Findings indicate that while EVs may offer certain safety gains, they also pose greater risks to vulnerable road users (e.g., pedestrians, cyclists, and low-income communities) due to design features such as increased weight, rapid acceleration, and silent operation, alongside unequal access to safe infrastructure. Marginalised populations face higher exposure and limited benefits. Addressing these inequities requires inclusive safety metrics, improved data collection, and equitable infrastructure planning to prevent reinforcing transport injustice.
Both the existing parking allocation models and pricing algorithms might over-allocate parking flow to the high-utility parking lots, which become quickly saturated by the early-birds and consequently the more qualified latecomers have to switch to a low-utility parking lot. To solve the spatiotemporal imbalance of area demand and supply, this study proposes a proportional pricing algorithm with an announced parking rate to regulate parking flow proportionally to the number of available spaces within each parking lot. For each parking lot at each allocation step, the announced parking rates are iteratively optimized to minimize the parking deviation between the current parking flow (obtained from the ordinary allocation model with the latest announced parking rates) and target values obtained from the proportional allocation model with the predetermined permitted ratios. At each optimization iteration, a parking lot with the largest parking deviation is selected with a certain increment in the corresponding announced parking rate, which is iteratively increased until satisfying the termination conditions. Then, a SUMO-based simulation was carried out to investigate the performances of the proportional pricing algorithm, using an empirical case study in Wujiaochang central business district (CBD), Shanghai. The results demonstrate that the proportional pricing algorithm outperforms the static parking rate scheme and online pricing strategy in decreasing both the nonmonetary costs and parking disutility, and has a potential to perform as better as the laborious offline pricing method. The proposed proportional pricing algorithm may assist in the design and operation of urban parking reservation systems.
With the world’s ageing population and increasing transport equity issues, shared autonomous vehicles (SAVs) represent a potential solution to the persistent mobility barriers faced by older adults and people with reduced mobility. These vulnerable populations often have limited access to transport due to poor access to public transit and limited paratransit service availability. Despite a growing interest in autonomous mobility, there is little evidence on how older adults and people with reduced mobility perceive SAVs, particularly in southern European contexts and under varying mobility and digital access conditions. This study analyses perceptions referring to SAV benefits and barriers, and the intended use of SAVs among older adults and people with reduced mobility in the Athens metropolitan area. It specifically identifies the main factors underpinning the willingness of these vulnerable populations to employ SAV services. A cross-sectional survey was conducted with 267 participants aged 60 + and mobility-impaired adults using paper-based questionnaires. The study used descriptive statistics, chi-square tests and ordinal logistic regression modelling to examine the relationships between demographic characteristics and perceptions of SAVs. Road safety was found to be the most important factor for SAV acceptance, followed by service reliability and vehicle accessibility. Concerns about human intervention in emergencies and accident risks were the biggest acceptance barriers. The regression model showed that perceived vehicle accessibility, autonomy benefits, cost savings and service reliability significantly predicted positive attitudes towards the societal value of SAVs. The results suggest that for SAVs to be popular among older adults and people with reduced mobility, service safety, affordability and reliability must be championed and emergency response concerns should be addressed.
This paper investigates the factors affecting household vehicle transactions, including additions, removals, and replacements, in the past (spring 2020 to fall 2023) and in the anticipated future (fall 2023 to fall 2026). Using a two-wave panel dataset (n = 1612) from surveys conducted in the U.S., we explore a broad set of hypotheses concerning vehicle transaction dynamics over time. An integrated choice and latent variable model identifies the effects of latent attitudes, sociodemographic characteristics, life events, work arrangements, and COVID-related health concerns in shaping vehicle transaction decisions. Results show that novelty-seeking individuals were more likely to engage in future vehicle transactions, whether by increasing, decreasing, or replacing vehicles. Younger adults, households with children, and those experiencing an increase in the number of children or adults showed a higher likelihood of acquiring vehicles, likely in response to a growing travel needs. Transitions into the workforce and rising household income further increased the likelihood of vehicle acquisition. An increase in commute frequency reduced the likelihood of vehicle shedding during the pandemic and also increased the likelihood of post-pandemic vehicle acquisition, likely due to a rebound in demand for non-commuting trips. COVID-related health concerns discouraged vehicle shedding during the pandemic. Importantly, past vehicle transactions strongly predicted future behaviors. Households that added vehicles during the pandemic were likely to either increase, decrease, or replace vehicles again. In contrast, those that shed vehicles tended to reacquire vehicles, and those who replaced vehicles were more inclined to do so again in the future.
Previous studies on Autonomous Vehicles (AVs) have largely focused on users inside the vehicle, neglecting Vulnerable Road Users (VRUs) such as pedestrians and cyclists. This study addresses this gap by examining VRUs’ perceptions of AVs through a national survey of 1,165 respondents across the United States. While past research has mainly examined direct associations between explanatory factors (e.g., sociodemographic characteristics and prior experience) and AV perceptions, this study proposes an indirect link, suggesting that VRUs’ attitudes toward AVs may be influenced by their existing perceptions of human drivers encountered in everyday road environments. The results show that 43.5
Many cities suffer from traffic congestion, but physical expansion of roads is often infeasible in urban areas. A solution in such circumstances is to build underground roads, which involves adding new lanes beneath existing roads or other facilities. This study explores route choice behavior on a 10 km stretch of an urban expressway where both underground toll lanes and ground-level free lanes are available. Real-time travel time information for each route is provided via variable message signs (VMS) for informed choices. We collected revealed preference data by taking videos at the entrance of two alternative routes in Seoul. The study analyzed 142,045 trips and found that overall 35 percent of those trips chose underground toll lanes. The share of drivers choosing underground toll lanes tends to be higher during peak hours compared to non-peak hours or weekends. By using the mixed logit with interaction effects, we found that there are meaningful heterogeneities in the sensitivity to travel time and thus the value of travel time (VOT) affected by temporal characteristics. Drivers generally have higher VOTs on Fridays, followed by other weekdays, and then weekends. In addition, people are more willing to use underground toll lanes during peak hours than during non-peak hours. The traffic count variable (a proxy for congestion level) has a positive coefficient, indicating that higher congestion levels are associated with a stronger preference for underground routes. After controlling for travel time, cost, and temporal effects, the alternative-specific constant is negative, which signifies that, ceteris paribus, drivers tend to prefer the ground-level option over the underground option. We also explored optimal pricing by accounting for time-dependent traffic volume and VOT. Overall, this study deepens our understanding of underground toll lane choice behavior and potentially suggests implications for pricing and lane operations.
Battery electric vehicle (BEV) users must integrate charging into their daily travel routines in ways unlike conventional refueling, requiring a deliberate coordination of trip purposes, time constraints, and charging accessibility. Understanding these behavioral charging decisions is critical for supporting the long-term sustainability and usability of EV systems. Yet most prior studies focus on infrastructure placement or isolated aspects of charging decisions, without systematically integrating activity-based travel routines into behavioral models. This study addresses that gap by investigating how BEV users in Japan choose between home charging and fast charging, focusing on trip purposes, acceptable waiting time, and socio-demographic characteristics. Using survey data from 441 BEV users across the Chubu and Kanto regions of Japan, three structural equation models (SEMs) are estimated to examine the behavioral pathways underlying charging decisions. Model 1 captures the baseline effects of socio-demographics and time sensitivity. Model 2 introduces total trip frequency as a mediator, and Model 3 disaggregates trip purposes—commuting, shopping, and leisure—to examine their effects on charging behavior. The results indicate that trip activity patterns are significant determinants of charging behavior. Commuting trips are most consistently associated with home charging, reflecting predictable travel schedules that facilitate planned overnight residential charging. Shopping trips show the strongest association with fast charging, suggesting that commercial destinations serve as key locations for opportunistic charging. Leisure trips exhibit a flexible, context-dependent pattern depending on trip distance and time availability. Acceptable waiting time emerges as a critical behavioral constraint, with lower acceptance directly increasing fast charging use and indirectly shaping charging behavior through travel frequency. Socio-demographic heterogeneity further shapes charging behavior: higher-income households show greater reliance on home charging, while middle-income households depend more on public fast charging. BEV model type also influences charging strategy, with Nissan Leaf owners relying predominantly on residential charging, whereas imported BEV owners adopt more diverse charging approaches. Regional differences between Chubu and Kanto highlight the role of residential density and private parking availability in shaping charging behavior. These findings suggest that EV infrastructure planning should integrate activity-based travel patterns.
The rise of technological advancements has led to the commonplace practice of online shopping for retail, grocery, and food. However, little research has been conducted on the interplay of these components in burdened communities (BCs) that face issues of marginalization and limited access to digital resources. This study aims to provide a comprehensive understanding of travel behavior changes by analyzing the interconnectedness of the emerging components of online shopping (retail, grocery, and food) and in-person activities in both BCs and non-BCs. A unique household-level database is created by linking the 2021 Puget Sound Household Travel Survey and the US Department of Transportation’s burdened community databases, and a conditional mixed process model is estimated to account for unobserved endogeneity. The findings suggest households living in BCs are less likely to order online retail goods and groceries compared to non-BC households. Additionally, the probability of making more restaurant trips decreases for households living in BCs. The study highlights the digital divide that exists in BCs and the differences in online and in-person shopping activities across socioeconomic levels. Policymakers may address these disparities to promote better access to goods and services for all. Besides, planners may need to improve the travel demand models by accounting for the emerging components of online shopping and the trip frequencies by purpose in BCs.
An affordable and efficient public transport system is fundamental for the well-being of the inhabitants of a city. It improves their access to economic opportunities, enabling them to rise out of poverty and overcome social inequities. A good public transport system benefits the low-income and marginalised segments of society the most, resulting in greater social inclusion. Pakistan lacks an effective public transport system even in large cities, undermining its economic potential. Thus, developing efficient public transport systems is vital for urban mobility in Pakistan. However, the government lacks resources, and the private sector is reluctant to invest due to limited policy support and incentives. Against this background, this study proposes low-cost interventions to improve public transport in three main Pakistani cities. These interventions are backed by data from a stated choice survey, analysed using state-of-the-art choice modelling techniques. Findings reveal that travellers tend to value level-of-service improvements more than reduced fares. At the same time, they show a strong dislike for vehicle transfer, probably due to high uncertainty in waiting times during transfers. Small improvements such as providing free WI-FI and reserved seats for women in buses may increase its market share up to 8
School choice offers families an opportunity for children to attend schools other than those they are zoned to by residential location. Most families live beyond walking distance to choice schools, so vehicular transportation is needed to attend. Yet, most choice programs have not been designed with such access in mind, and most school bus programs are designed only for neighborhood schools. Such diminished access for children in households without cars presents an equity issue. This study examines the relationship between household transportation resources—automobiles, transit, and walkability—and the likelihood that children attend a choice school. We match data from the 2017 National Household Travel Survey California Add-On with school characteristics to identify the school each respondent ages 5–17 attends and the school to which school districts zoned them. We then fit a model that predicts choice of school type: neighborhood zone public, choice public, or private. Controlling for student, household, transportation, neighborhood, and assigned zone school characteristics, a student in a household with at least one vehicle had more than double the predicted probability of attending a choice public school compared with a student in a zero-vehicle household. The difference in predicted probabilities grew for students who were low-income, non-white, or zoned to low-performing schools. Living in neighborhoods with high-quality public transit access or high walkability did not affect the probabilities. These findings underscore the importance of transportation resources in enabling families to send their children to schools that best fit their needs.
