Research Output

@article{doi:10.1177/03611981251362154,

title = {Optimizing Sustainable Relay Freight Transport Networks: A Case Study in Japan},

author = {Zeying Wen and Tao Feng and Junyi Zhang and Xin Dou},

url = {https://doi.org/10.1177/03611981251362154},

doi = {10.1177/03611981251362154},

year  = {2026},

date = {2026-01-01},

urldate = {2026-01-01},

journal = {Transportation Research Record},

volume = {2680},

number = {1},

pages = {798-815},

abstract = {Relay transport networks design plays an important role in improving logistics efficiency, involving the determination of the number of relay points, costs of constructing these points, and constraints on the distance between relay points. The configuration of these unique characteristics must consider multiple aspects of sustainability, including economic, environmental, and social factors. In this study, we aim to determine the optimal relay network configurations and transportation routes through a multiobjective optimization model, considering three key objectives: maximizing freight demand for relay transport, minimizing total costs, and maximizing CO2 emission reduction. The proposed model was solved using the Gurobi optimization solver. A case study conducted in Japan revealed that the optimal relay network configuration consists of 23 relay points, including 21 small, 1 medium, and 1 large point. This optimized configuration yielded three significant contributions for environmental, economic, and social aspects, namely 34.72% reduction in CO2 emissions, 3.65% reduction in transportation costs, 27.47% reduction in overtime for short-haul trips, and 14.92% reduction in overnight stays for long-haul trips. Our findings recommend the following policies: (i) trucking companies should set the distance constraints between two adjacent relay points as 150 km for short-haul trips and 450 km for long-haul trips to balance environmental, social, and economic priorities; (ii) the Japanese government’s proposed budget (2.5 billion yen) for relay point construction is relatively sufficient, exceeding our estimated construction cost (2.095 billion yen); and (iii) relay transport implemented as a standalone measure has the potential to help achieve the Japanese trucking industry’s CO2 emission reduction target.},

keywords = {},

pubstate = {published},

tppubtype = {article}

}

@article{QI2026101159,

title = {Modeling the effects of additional crowd-sourced delivery service on transportation mode choice in mobility-as-a-service},

author = {Qiang Qi and Soora Rasouli and Tao Feng},

url = {https://www.sciencedirect.com/science/article/pii/S2214367X25001772},

doi = {https://doi.org/10.1016/j.tbs.2025.101159},

issn = {2214-367X},

year  = {2026},

date = {2026-01-01},

journal = {Travel Behaviour and Society},

volume = {42},

pages = {101159},

abstract = {This paper investigates the impact of people’s decision to be a crowd-shipper within Mobility-as-a-Service (MaaS) on their transportation mode choice. Leveraging the unique characteristics of MaaS as a comprehensive mobility service platform, we propose embedding Crowd-Sourced Delivery (CSD) as an additional service. This integrated service not only allows travelers to choose a transportation mode but also provides the option to accept (or decline) parcel delivery tasks during their journeys. Given the aim of MaaS to promote the use of sustainable transportation modes and replacement of car-based transport, we designed a stated choice experiment to capture respondents’ choices of transportation modes with and without the additional CSD service. This approach enables us to explore whether being a crowd-shipper has an impact on transportation mode choice. A random parameter mixed logit model was estimated to identify the effects of various factors. The results reveal that car-based transport is generally preferred for delivering parcels. Despite this preference, which contradicts the aim of MaaS, several factors can encourage the use of sustainable transportation modes, including public and active transport, within integrated CSD and MaaS. The parcels not being fragile, less strict delivery time window, and higher monetary incentives for using sustainable transport in CSD are among those factors. Additionally, travel contexts such as travel distance and weather, as well as the age of respondents, also influence the transportation mode choice decision once a parcel delivery task is accepted. These insights offer valuable information for policymakers and service operators to promote the use of sustainable transportation modes within CSD.},

keywords = {},

pubstate = {published},

tppubtype = {article}

}

@article{JI2026108251,

title = {How urban sprawl patterns shape the thermal environment during hot summers: An empirical analysis of 338 Chinese cities},

author = {Yifeng Ji and Zhitao Li and Ying Liu and Hongyu Tang and Hao Sun and Tao Feng},

url = {https://www.sciencedirect.com/science/article/pii/S0195925525004482},

doi = {https://doi.org/10.1016/j.eiar.2025.108251},

issn = {0195-9255},

year  = {2026},

date = {2026-01-01},

journal = {Environmental Impact Assessment Review},

volume = {117},

pages = {108251},

abstract = {With accelerating global warming trends, urban sprawl has emerged as a key driver of thermal environment changes in urban areas. However, limited attention has been given to how sprawl patterns shape the thermal environment at the city level. This study sets out to reveal the heterogeneous relationship between the spatial patterns of urban sprawl and thermal environment by analyzing 338 Chinese cities during hot summers from 1990 to 2020. Multi-source data were employed to assess the spatial arrangement and temporal dynamics of four urban sprawl patterns (centering, clustering, fragmentation, and complexity) alongside thermal conditions. A geographically weighted regression model revealed spatially varying associations between sprawl patterns and thermal environment. Results indicate a northward gradient of thermal deterioration, with the most severe warming in Northeast and Northwest China, while some southern tropical cities exhibit localized cooling. Sprawl patterns vary across regions and exhibit dual heterogeneity in their thermal impacts. Fragmentation demonstrates cooling effects in the southeastern coast, eastern Northwest, and northeastern Southwest China. Clustering improves thermal conditions in the southeastern coast, Bohai Rim, and northeastern and northwestern China, but intensifies heat stress in the southwest. Complexity worsens the thermal environment in the Yangtze River Delta and the junction of eastern Northwest and northeastern Southwest China, while slightly mitigating the thermal environment in northern cities. Centering shows limited impact, with minor improvements observed in southeastern coastal cities. Based on these findings, seven optimization zones have been identified with tailored thermal-mitigation strategies, providing evidence to guide regionally adaptive planning for climate-resilient urban development.},

keywords = {},

pubstate = {published},

tppubtype = {article}

}

@article{HU2026101141,

title = {Crowdshipping with private autonomous vehicles: Exploring the propensity of car owners},

author = {Yan Hu and Ying Zhao and Xiaodong Li and Tao Feng},

url = {https://www.sciencedirect.com/science/article/pii/S2214367X25001590},

doi = {https://doi.org/10.1016/j.tbs.2025.101141},

issn = {2214-367X},

year  = {2026},

date = {2026-01-01},

journal = {Travel Behaviour and Society},

volume = {42},

pages = {101141},

abstract = {As e-commerce and on-demand mobility services continue to experience rapid growth, the demand for parcel delivery has increased, prompting the exploration of innovative solutions like crowdshipping to address transportation challenges. Despite its potential, the current crowdshipping delivery mode grapples with efficiency issues and contributes to heightened traffic congestion. The emergence of autonomous vehicles presents a promising prospect for a new and potentially more efficient delivery mode. Upon completing passenger drop-off at their destinations, autonomous vehicles face the challenge of finding parking spaces or cruising in proximity while waiting for the next call, escalating the pressure on parking and traffic. Integrating idle autonomous vehicles into crowdshipping, however, provides an effective means to optimize resources, reduce the dependence on parking infrastructure, and generate revenue for car owners. In this study, we aim to analyze and understand the behavior and decision-making mechanism of autonomous vehicle owners regarding the adoption of crowdshipping. A stated choice experiment was designed focusing on people’s adoption behavior when traveling in various contexts related to travel purpose, duration, revenue, platform fee, and waiting time. Findings derived from a mixed logit model reveal notable variations in preferences among car owners. Revenue, platform fee, waiting time, and social influence all significantly influence the crowdshipping decision. Moreover, individuals with a college education, employed, and higher income groups showed a greater propensity for adopting crowdshipping. These findings provide valuable insights for crowdshipping operators, urban planners, and other stakeholders involved.},

keywords = {},

pubstate = {published},

tppubtype = {article}

}

@article{SHAO2026105166,

title = {Exploring preference heterogeneity of electric ride-hailing drivers’ charging decisions under time uncertainty},

author = {Mengru SHAO and Chao CHEN and Tao FENG},

url = {https://www.sciencedirect.com/science/article/pii/S1361920925005760},

doi = {https://doi.org/10.1016/j.trd.2025.105166},

issn = {1361-9209},

year  = {2026},

date = {2026-01-01},

journal = {Transportation Research Part D: Transport and Environment},

volume = {151},

pages = {105166},

abstract = {The rapid electrification of ride-hailing fleets intensifies demands on public charging infrastructure, making drivers’ charging behavior pivotal for effective system planning. Unlike studies focusing on private electric vehicle (EV) users or taxi drivers, this study examines electric ride-hailing drivers’ charging preferences under waiting time uncertainty using a conjoint stated choice experiment. A latent class mixed logit model is employed to capture behavioral heterogeneity, identifying two driver classes: flexible and proactive drivers and time- and price-constrained drivers. Results show that while both classes value shorter waits and lower prices, the time- and price-constrained drivers are considerably more sensitive to uncertainty and costs, diverging from patterns observed among private EV users. Preferences for ancillary facilities also differ across classes. These findings extend existing knowledge by providing new behavioral evidence on how electric ride-hailing drivers manage trade-offs under operational uncertainty and offering actionable implications for differentiated infrastructure and policy design.},

keywords = {},

pubstate = {published},

tppubtype = {article}

}

@article{ZHANG2026104827,

title = {Charging mode choice of electric micromobility users under uncertainty},

author = {Fan Zhang and Huitao Lv and Zijie Xu and Tao Feng and Yanjie Ji and Phil Blythe},

url = {https://www.sciencedirect.com/science/article/pii/S0965856425004604},

doi = {https://doi.org/10.1016/j.tra.2025.104827},

issn = {0965-8564},

year  = {2026},

date = {2026-01-01},

journal = {Transportation Research Part A: Policy and Practice},

volume = {204},

pages = {104827},

abstract = {The need to develop new charging practices for electric micromobility vehicles (EMVs) is inevitable due to the challenges associated with charging and managing their batteries. Intelligent battery swapping services for EMVs have expanded from being business-to-business to business-to-customer. This battery swapping model offers the advantages of convenience and battery controllability, effectively addressing the current charging issues faced by EMVs. To assess the feasibility and potential success of promoting intelligent battery swapping services, this study conducts a stated choice experiment to investigate the charging behavior of EMV users across different modes, such as self-operated charging, charging at charging stations, and intelligent battery swapping. We propose a charging and swapping choice modeling approach that combines cumulative prospect theory and multi-attribute decision making methods to model charging decisions under uncertainty. The results demonstrate that our proposed method surpasses conventional models in terms of model goodness-of-fit and behavioral interpretation. Furthermore, heterogeneity was observed among EMV users in their charging and swapping behaviors. Based on these findings, we discuss policy implications for creating sustainable cities, with a particular focus on establishing intelligent battery swapping facilities.},

keywords = {},

pubstate = {published},

tppubtype = {article}

}

@article{LAN2026101176,

title = {Modeling interdependent choices of remote working centers and transportation with attitudes through latent variables},

author = {Jieyuan Lan and Tao Feng},

url = {https://www.sciencedirect.com/science/article/pii/S2214367X25001942},

doi = {https://doi.org/10.1016/j.tbs.2025.101176},

issn = {2214-367X},

year  = {2026},

date = {2026-01-01},

journal = {Travel Behaviour and Society},

volume = {43},

pages = {101176},

abstract = {The paper aims to investigate the interdependent choice behavior of individuals regarding Remote Working Centers (RWCs) and transportation modes, focusing particularly on latent factors like attitudes and personality traits. Using stated preference data from Tokyo, Japan, an integrated choice and latent variable (ICLV) model was applied. The results reveal that RWCs with six square meters of workspaces, minimal distractions, and no nearby amenities increase the probability of using RWCs. Work commitment, opportunity loss, workplace attire, timesaving attitude, workplace attachment, and workplace aversion significantly influence RWC usage. Individuals having higher work commitment, greater concerns about opportunity loss, higher workplace attire, and stronger workplace attachment and aversion are more likely to use RWCs, while timesaving people prefer working from home. Transportation choices also vary, with workplace attire investments favoring driving, and timesaving attitudes leaning toward walking or rail. These findings provide valuable insights for policymakers and stakeholders to promote teleworking, alleviating traffic and reducing environmental impacts.},

keywords = {},

pubstate = {published},

tppubtype = {article}

}

@article{ZHANG2026101177,

title = {Willingness to become a ride-sourcing driver: results of an error components model},

author = {Hexin Zhang and Tao Feng and Harry J. P. Timmermans},

url = {https://www.sciencedirect.com/science/article/pii/S2214367X25001954},

doi = {https://doi.org/10.1016/j.tbs.2025.101177},

issn = {2214-367X},

year  = {2026},

date = {2026-01-01},

journal = {Travel Behaviour and Society},

volume = {43},

pages = {101177},

abstract = {Research on individuals’ participation in being a driver for ride-sourcing companies is rare and this decision has not been fully understood. This study uses stated choice experiment data collected in China through face-to-face interviews to explore the effects of job conditions, car ownership and socio-demographics on individuals’ decisions to become a driver for ride-sourcing companies. The results show that car owners and those who work in the tertiary industry have a higher interest in driving for a ride-sourcing company beyond regular working hours, but less interest in becoming a full-time driver. Social insurance is an important trigger of participation for individuals without a stable job. Middle-aged people have an interest in driving and prefer full-time driving over part-time driving or driving after work. These findings provide useful insights for improving driver recruitment and ride-sourcing sustainability.},

keywords = {},

pubstate = {published},

tppubtype = {article}

}

@article{TANG2026104758,

title = {Modeling heterogeneous individual choices for multi-person household MaaS bundles},

author = {Hongyu Tang and Tao Feng},

url = {https://www.sciencedirect.com/science/article/pii/S096585642500391X},

doi = {https://doi.org/10.1016/j.tra.2025.104758},

issn = {0965-8564},

year  = {2026},

date = {2026-01-01},

journal = {Transportation Research Part A: Policy and Practice},

volume = {203},

pages = {104758},

abstract = {Mobility as a Service (MaaS) has been investigated broadly regarding its potential market share through individual-oriented bundles or pay-as-you-go offerings. However, its potential for individual-specific offerings within multi-person households remains underexplored. This study attempts to investigate individual heterogeneous choice behavior in household-level MaaS, where bundles are specified both for the decision-maker and for their partner. Using data from a dedicated stated choice experiment involving 402 couples, we applied a latent class and latent variable model incorporating attitudinal factors such as altruism, egoism, empowerment, and household task responsibility, and socio-demographic characteristics. Results show that prioritizing mobility services for husbands may be more effective in promoting household-level MaaS adoption. Furthermore, three distinct decision-maker profiles were identified: male-dominant individualists, female-dominant individualists, and non-individualists. The significant preference heterogeneity highlights how individuals differently weigh their own versus their partner’s utility when evaluating household-level MaaS bundles. These findings provide valuable insights for designing more inclusive and targeted MaaS products that account for behavioral diversity at the individual level.},

keywords = {},

pubstate = {published},

tppubtype = {article}

}

@article{CHEN2026100176,

title = {Assessing the importance of travel habit and the built environment in shaping elderly’s mobility patterns: A case study in the Netherlands},

author = {Chao Chen and Mengru Shao and Xiaoning Gu and Tao Feng},

url = {https://www.sciencedirect.com/science/article/pii/S2667091725000780},

doi = {https://doi.org/10.1016/j.urbmob.2025.100176},

issn = {2667-0917},

year  = {2026},

date = {2026-01-01},

journal = {Journal of Urban Mobility},

volume = {9},

pages = {100176},

abstract = {In the context of social resource inequality among the elderly, understanding elderly’s travel behavior is of importance for enhancing mobility services. Recent attention has been investigated the built environment impacts on travel decisions, however, the influencing mechanisms and its extent on elderly’s mobility patterns are not yet well understood. In particular, how the mode choice behavior of older population is shaped by their habitual behavior remains underexplored. Using the Dutch national travel survey data, this study explored the non-linear effects of historical travel patterns and the built environment on mode choice behaviors, as well as the heterogeneity across different aged groups. We found that historical travel patterns exert a larger impact on mode choice comparing to the built environment factors. Significant threshold effects of built environment factors on mode choice (indicating non-linear impacts) are observed. Older adults tend to walk within a travel distance of 2.5 km and use bike for longer trips beyond 3.5 km. The findings are valuable for policymakers in promoting active mobility patterns for older adults through infrastructure planning.},

keywords = {},

pubstate = {published},

tppubtype = {article}

}

@article{SHAO2026103983,

title = {Scenic toll roads or cost-effective free roads? A stated preference analysis of private car users},

author = {Mengru Shao and Chao Chen and Tao Feng},

url = {https://www.sciencedirect.com/science/article/pii/S0967070X25005268},

doi = {https://doi.org/10.1016/j.tranpol.2025.103983},

issn = {0967-070X},

year  = {2026},

date = {2026-01-01},

journal = {Transport Policy},

volume = {179},

pages = {103983},

abstract = {The operation of scenic toll roads serves a dual purpose of enhancing drivers’ travel experiences while generating revenue for road operators, however, its effects on drivers’ route choice decisions remained underexplored. In this paper, we attempt to investigate the determinants influencing private car drivers’ route choice behavior over scenic toll roads and cost-effective free roads in the context of travel time variability, considering various combinations of travel contexts, route-specific attributes, marketing policies, and drivers’ latent attitudes. Using the data collected via a stated choice experiment, we also develop a hybrid prospect theoretic choice model (HPTCM) to further understand drivers’ preferences in scenarios under risky conditions. The results show that increased travel time uncertainty significantly reduces the attractiveness of cost-effective free routes, prompting a shift toward scenic toll roads. Service areas designed by including water and cultural landscapes increases the likelihood of using scenic toll roads. Furthermore, drivers’ attitudes toward safety, regulatory adherents and food health emerge as critical influencing factors. These insights offer practical guidance for toll-road operators in developing pricing mechanisms and enhancing service area infrastructure, thereby contributing to the broader advancement of the transport–tourism nexus.},

keywords = {},

pubstate = {published},

tppubtype = {article}

}

@article{ZHAO20261,

title = {How airport service quality and expectation confirmation shape airline passengers’ behavioral intentions: Evidence from Papua New Guinea},

author = {Ying Zhao and Yan Hu and Podi Richard Vai and Tao Feng},

url = {https://www.sciencedirect.com/science/article/pii/S2949899625000358},

doi = {https://doi.org/10.1016/j.team.2025.12.002},

issn = {2949-8996},

year  = {2026},

date = {2026-01-01},

journal = {Transport Economics and Management},

volume = {4},

pages = {1-13},

abstract = {This study aims to investigate the distinct effects of airline passengers’ initial expectations and perceived airport service quality on behavioral intentions. Drawing upon the Expectation Confirmation Theory, we examine the interrelationships among airport service quality, expectation confirmation, passenger satisfaction, and behavioral intentions, namely airport reuse and word-of-mouth recommendation. Data were collected from 484 departing passengers across three major airports in Papua New Guinea. Using structural equation modeling and multi-group analysis, the study analyzes passengers’ perceptions and attitudes toward their airport experiences. The results reveal that airport service quality significantly influence airport reuse and word-of-mouth intentions, with expectation confirmation and satisfaction serve as key mediators. These findings provide actionable insights for airport managers aiming to enhance service quality and the overall passenger experience.},

keywords = {},

pubstate = {published},

tppubtype = {article}

}

@article{KYI2026100059,

title = {Understanding pre-adoption intentions for urban air logistics: A hybrid LC-SEM-Fuzzy Kano model},

author = {Ei Phyu Kyi and Tao Feng},

url = {https://www.sciencedirect.com/science/article/pii/S3051064326000117},

doi = {https://doi.org/10.1016/j.sbr.2026.100059},

issn = {3051-0643},

year  = {2026},

date = {2026-01-01},

journal = {Strategic Business Research},

volume = {2},

number = {1},

pages = {100059},

abstract = {Customer acceptance is fundamental to both product sales and service relationships, as it depends on offerings that align with customer needs. However, since air logistics delivery is still in a pre-adoption context, satisfaction cannot be assessed through actual service experience. Instead, adoption decisions are shaped by expected satisfaction (ES), formed through psychological perceptions and expectations regarding key service attributes. This study investigates how psychological factors and attribute-level expectations jointly influence behavioral intention (BI) toward drone delivery. We propose a hybrid customer behavioral intention framework that integrates a latent class structural equation model (LC-SEM) and a threshold-based fuzzy Kano model (FKM). Using survey data from 528 respondents in Japan, LC-SEM identifies unobserved consumer segments and estimates class-specific psychological pathways from perceived usefulness (PU), perceived ease of use (EU), perceived innovativeness (PI), and social influence (SI) to BI. FKM is employed as a complementary descriptive tool to classify the relative importance structure of primary service attributes across segments. The LC-SEM results reveal three distinct consumer classes: innovation-oriented adopters influenced primarily by PI and EU; socially driven adopters whose intentions depend on peer influence and novelty; and low-certainty users for whom psychological drivers do not significantly predict BI. The FKM analysis indicates that safety, reliability, and privacy consistently emerge as critical requirements, while delivery speed and environmental considerations function as baseline expectations. Cost efficiency and convenience operate as value-adding features that enhance service appeal when expectations are exceeded. By integrating latent psychological pathways with attribute-level expectation structures, this study provides a nuanced understanding of consumer heterogeneity in drone delivery adoption. The findings offer actionable insights for the design and communication of trustworthy urban air logistics services in pre-adoption markets.},

keywords = {},

pubstate = {published},

tppubtype = {article}

}

@article{SHAO2026101729,

title = {From road waiting to road leading: can toll-road marketing strategies make a difference for truck drivers under uncertain travel times?},

author = {Mengru Shao and Chao Chen and Tao Feng},

url = {https://www.sciencedirect.com/science/article/pii/S2213624X26000258},

doi = {https://doi.org/10.1016/j.cstp.2026.101729},

issn = {2213-624X},

year  = {2026},

date = {2026-01-01},

journal = {Case Studies on Transport Policy},

volume = {23},

pages = {101729},

abstract = {Understanding truck drivers’ route choice behavior amid the rapid expansion of freight transport is crucial for road operators to improve their services and secure stable revenue. In addition to the strict delivery schedules, truck drivers’ route choice decisions are highly sensitive to uncertain traffic conditions and policy constraints. However, to what extent operator-led marketing strategies shape their route choice decisions remains insufficiently addressed in the literature. Therefore, this study aims to investigate truck drivers’ route decisions under travel time uncertainty with a particular focus on the effects of promotional marketing strategies. A stated choice experiment was designed to collect truck drivers’ responses under different cargo-delivery contexts to toll and parallel toll-free route alternatives. A hybrid prospect-theoretic Probit model (HPTPM), incorporating interaction effects, is further developed to account for risk attitudes in decision-making under uncertainty. The findings reveal that cargo-delivery context variables, marketing strategies, and risk preferences all significantly shape truck drivers’ route choices, with notable heterogeneity observed across driver groups. Based on these insights, this study provides practical recommendations for road operators, supporting the development of tailored marketing strategies to improve the effectiveness of toll-road management.},

keywords = {},

pubstate = {published},

tppubtype = {article}

}

@article{ZHOU2026104606,

title = {Does telecommuting promote active travel? New evidence from the Netherlands},

author = {Ting Zhou and Astrid Kemperman and Tao Feng},

url = {https://www.sciencedirect.com/science/article/pii/S0966692326000608},

doi = {https://doi.org/10.1016/j.jtrangeo.2026.104606},

issn = {0966-6923},

year  = {2026},

date = {2026-01-01},

journal = {Journal of Transport Geography},

volume = {132},

pages = {104606},

abstract = {Telecommuting has been proposed as a strategy to promote sustainable mobility, yet its impact on active travel remains unclear, particularly when considering socio-demographics and telecommuting-related behavioral changes. To address this gap, this study investigates the relationships between telecommuting frequency, telecommuting-related behavioral changes, and the frequency of walking, conventional bike use and e-bike use in the Netherlands. Using data from the 2022 Netherlands Mobility Panel, we estimated a Bayesian Network model. Results show that moderate telecommuting (2–4 days/week) is associated with higher walking frequency, while full-time telecommuting (5+ days) generally is linked to lower likelihoods of frequent conventional bike and e-bike use. Moreover, respondents who reported ‘started to make more trips’ showed higher probabilities of using conventional bikes, whereas those who “started travelling on other days” exhibited distinct patterns of e-bike use, including greater probabilities of occasional use. Additionally, e-bike ownership corresponds to more e-bike use while coinciding with lower possibilities of frequent walking and conventional bike use, indicating a substitution effect. These findings provide empirical insights for policymakers seeking to design telecommuting and mobility policies that encourage active travel.},

keywords = {},

pubstate = {published},

tppubtype = {article}

}

@article{YANG2026108481,

title = {Decoding the influencing factors of pedestrian-automated vehicle interaction in shared spaces: A multi-level path analysis approach},

author = {Liu Yang and Gamze Dane and Tao Feng and Astrid Kemperman},

url = {https://www.sciencedirect.com/science/article/pii/S0001457526000904},

doi = {https://doi.org/10.1016/j.aap.2026.108481},

issn = {0001-4575},

year  = {2026},

date = {2026-01-01},

journal = {Accident Analysis & Prevention},

volume = {231},

pages = {108481},

abstract = {In the near future, pedestrians will increasingly face automated vehicles (AVs) in urban environments, particularly in shared spaces. Ensuring safe and effective pedestrian-AV interactions requires a deeper understanding of the factors that affect pedestrian behaviour. This study examines how environmental and individual factors influence pedestrians’ perception and crossing behaviour in front of AVs in shared spaces. A virtual reality (VR) experiment with 60 participants was conducted to simulate diverse traffic scenarios, and both subjective and behavioural data were collected after each trial. Using multi-level path analysis, we modelled the direct and indirect effects of environmental factors (e.g., lane width, visual load, surface condition, time of day, traffic markings, traffic conditions) and individual factors (e.g., age, gender, educational level, personality) on perceived safety, comfort, legibility, trust and behavioural outcomes including crossing initial time, crossing duration and safety margin. The findings highlight that traffic markings and traffic conditions are the most influential factors, while educational level, transport modes, and personality traits also play a significant role. For example, the presence of zebra and yielding AV behaviours were associated with more positive perception and safer crossing behaviour. Participants with higher education levels and greater openness tended to show more supportive attitudes during interactions with AVs. In addition, perception served not only as an outcome but also as a mediator associating context and behaviour. The results provide valuable insights for enhancing the design of AV systems and shared spaces to improve pedestrian safety and trust.},

keywords = {},

pubstate = {published},

tppubtype = {article}

}

@article{WEN2026104033,

title = {Relay transport network design for sustainable trucking industry considering carbon emissions and driver working conditions},

author = {Zeying Wen and Xin Dou and Xiaowen Fu and Tao Feng},

url = {https://www.sciencedirect.com/science/article/pii/S0967070X26000430},

doi = {https://doi.org/10.1016/j.tranpol.2026.104033},

issn = {0967-070X},

year  = {2026},

date = {2026-01-01},

journal = {Transport Policy},

volume = {180},

pages = {104033},

abstract = {Relay transport is an emerging collaborative approach that divides traditional long-haul shipments into multiple shorter segments, each handled by a different truck driver. By jointly completing shipments and enabling the shared use of transportation resources, this approach has the potential to alleviate operational inefficiencies, environmental pollution, and poor driver working conditions in the trucking industry. However, existing studies on relay transport network design primarily concentrate on the operational perspective of private companies. To address environmental and social issues in the trucking sector, this study proposes a bi-level programming model to develop a sustainable relay transport network that accounts for carbon reduction targets, carbon tax policies, driver work regulations, relay point capacity limits, and investment budgets. The upper-level government agency aims to minimize carbon emissions, driver overtime hours, driver overnight stays, and infrastructure construction costs by determining the number, location, and capacity of relay points. The lower-level trucking companies select freight transport routes based on the given network configuration, influenced by transport costs, cargo transit times, and carbon taxes. The model is solved using a genetic algorithm integrated with the method of successive averages. A case study in Japan identifies 22 relay points (14 small, 7 medium, 1 large). The optimized relay transport network offers substantial improvements compared to the traditional direct transport network, achieving reductions of 40.48 % in carbon emissions, 53.21 % in driver overtime hours, 70.38 % in driver overnight stays, 20.07 % in operational costs, and 5.13 % in total cargo transit times. Sensitivity analyses highlight the positive impact of appropriate relay point capacity and higher carbon tax on the network's sustainability performance. These findings offer valuable insights for government agencies to configure relay transport networks and support policymaking for sustainable freight transport planning.},

keywords = {},

pubstate = {published},

tppubtype = {article}

}

@article{Shao26032026,

title = {To be patient or to take a risk? E-hailing drivers’ charging decisions under waiting time uncertainty},

author = {Mengru Shao and Chao Chen and Tao Feng},

url = {https://doi.org/10.1080/19427867.2026.2649367},

doi = {10.1080/19427867.2026.2649367},

year  = {2026},

date = {2026-01-01},

journal = {Transportation Letters},

volume = {0},

number = {0},

pages = {1–22},

publisher = {Taylor & Francis},

abstract = {Advancements in charging speed and battery capacity are accelerating electric vehicles (EVs) adoption, particularly in ride-hailing services. However, limited battery capacity and charging infrastructure require e-hailing drivers to interrupt operations for recharging, where time spent waiting or charging directly translates into lost revenue. This study examines charging decisions of e-hailing drivers under waiting time uncertainty, considering travel contexts, station attributes, and latent attitudes. Using data from a conjoint choice experiment, a hybrid choice model (HCM) is developed. Results indicate that travel context, charging price, waiting time, and latent attitudes significantly influence charging behavior. The estimated willingness-to-pay (WTP) for reducing waiting time is 69.37 CNY/h, exceeding that reported for private EV users. Elasticity analysis shows that charging price has a stronger effect on charging decisions than waiting time and charger availability. These findings provide insights for optimizing charging infrastructure planning and management for e-hailing services.},

keywords = {},

pubstate = {published},

tppubtype = {article}

}

@article{Dou13032026,

title = {Competition of ridesharing autonomous vehicles: a stated preference analysis of air and ground mobility alternatives},

author = {Xin Dou and Xiaofeng Pan and Tao Feng},

url = {https://doi.org/10.1080/19427867.2026.2639046},

doi = {10.1080/19427867.2026.2639046},

year  = {2026},

date = {2026-01-01},

journal = {Transportation Letters},

volume = {0},

number = {0},

pages = {1–18},

publisher = {Taylor & Francis},

abstract = {This study contributes empirical evidence on competition among Autonomous Air Taxis (AATs), Autonomous Taxis (ATs), and Autonomous Buses (ABs). An offline survey including a stated choice (SC) experiment and personal attitude statements was conducted. Observations from 561 residents of Shanghai, China were collected. A latent class hybrid choice model was developed and estimated. The estimation results show that personal attitudes and heterogeneous preferences significantly affect the acceptance of the AV modes. Approximately 80.13% of respondents with low incomes and unstable employment are highly sensitive to the costs and travel times of ATs and AATs. About 19.87% of respondents who have high incomes and stable work are easily influenced by choices of others in their social networks. Moreover, choice preferences for AVs are significantly influenced by personal attitudes toward AVs, such as perceived safety of autonomous taxis and a dislike of public transportation.},

keywords = {},

pubstate = {published},

tppubtype = {article}

}

@article{Nie2025,

title = {What drives public participation in urban regeneration? An integrated choice and latent variable analysis},

author = {Mingxi Nie and Tao Feng},

url = {https://doi.org/10.1038/s41598-025-32386-z},

doi = {10.1038/s41598-025-32386-z},

issn = {2045-2322},

year  = {2025},

date = {2025-12-18},

journal = {Scientific Reports},

abstract = {Resident participation is essential to the success of urban regeneration, ensuring plans align with local needs and foster community consensus. However, limited studies have examined residents' renewal preferences and the role of socio-psychological factors in shaping them. This study investigates how facility features, psychological factors, and individual characteristics influence the choice of renewable energy plans. Using data collected through a face-to-face survey, we estimated an integrated choice and latent-variable model. Results show that plans emphasizing green space, public activity areas, and electric vehicle charging stations are most favored. Among psychological factors, social influence, place attachment, and social trust emerge as key latent drivers. The relatively weaker sense of community among younger, short-term, and less educated residents highlights an opportunity to tailor communication and engagement efforts to increase their support for renewal initiatives. These findings highlight the importance of consensus-building in the renewal process and underscore the need for strategies that strengthen public acceptance, providing policymakers with a practical pathway to design effective urban regeneration programs.},

keywords = {},

pubstate = {published},

tppubtype = {article}

}

@article{ZHAO2025106997,

title = {Optimization of passenger flow control and customized bus bridging in urban rail transit network},

author = {Qingqing Zhao and Jinjin Tang and Chao Li and Qiuhan Dong and Tao Feng and Xingwei Yang},

url = {https://www.sciencedirect.com/science/article/pii/S0305054825000255},

doi = {https://doi.org/10.1016/j.cor.2025.106997},

issn = {0305-0548},

year  = {2025},

date = {2025-01-01},

journal = {Computers & Operations Research},

pages = {106997},

abstract = {When dealing with large passenger flows in urban rail transit, developing reasonable passenger flow control plans is a crucial aspect of operational management. When many passengers wait outside the stations, customized buses can serve as an effective way to alleviate passenger pressure. To obtain reasonable passenger flow control and customized bus connection plans, we propose an optimization method for passenger flow control and customized bus connections. Firstly, we construct a two-stage model. The first-stage model formulates the passenger flow control plan to minimize the number of waiting passengers inside and outside the station considering factors such as passenger arrivals and urban rail transit train departures and arrivals. The second-stage model develops the customized bus plan based on a time–space-state network. According to the characteristics of the model, the CPLEX solver is used to solve the first-stage model, and a Lagrange relaxation-based solution method is designed to solve the second-stage model. To verify the effectiveness of the model, a small-scale case study including two lines and five stations is designed, and the results show that the proposed method can achieve effective passenger flow control and bus connection plans. Additionally, actual data from the Chengdu Metro is used to further validate the feasibility of the method. The results indicate that the passenger flow control and bus bridging plans can effectively reduce passenger waiting time. The proposed method can effectively alleviate passenger pressure, ensure passenger travel safety, and further promote the coordinated development of public transportation},

keywords = {},

pubstate = {published},

tppubtype = {article}

}

@article{ZHANG2025104416,

title = {The battery-swapping revolution: Exploring user preferences in electric micro-mobility sector},

author = {Fan Zhang and Huitao Lv and Chenchen Kuai and Tao Feng},

url = {https://www.sciencedirect.com/science/article/pii/S0965856425000448},

doi = {https://doi.org/10.1016/j.tra.2025.104416},

issn = {0965-8564},

year  = {2025},

date = {2025-01-01},

journal = {Transportation Research Part A: Policy and Practice},

volume = {194},

pages = {104416},

abstract = {Battery swapping services (BSS) offer innovative solutions to address the challenges of charging, safety, and battery management for electric micro-mobility vehicles (EMVs), such as electric bicycles and mopeds. However, the preferences and willingness to pay (WTP) for BSS among EMV users remain uncertain. This study examines consumer preferences for battery-swapping versus rechargeable EMVs within the Chinese market, identifying key factors that influence user choices. We employ a Stated Preference (SP) approach combined with a hybrid choice model (HCM) to assess both the economic viability and psychological impacts, including risk perception and social influence. Key findings demonstrate a higher willingness to pay for battery-swapping EMVs due to reduced charging time and enhanced safety features. Specifically, consumers are willing to pay an additional $2.86 for each minute reduction in facility accessibility time and an additional $27.52 for the safety features of battery-swapping EMVs. Existing users of traditional EMVs show a strong preference for BSS, willing to pay an additional $97.86. Frequent riders, who are more cost-sensitive, prefer rechargeable EMVs and are willing to reduce their expenditure by $47.95. In contrast, long-distance riders value battery endurance and are willing to pay an additional $55.96 for battery-swapping EMVs. These findings deepen our understanding of consumer behavior and provide valuable insights for policymakers and manufacturers aiming to optimize EMV adoption strategies.},

keywords = {},

pubstate = {published},

tppubtype = {article}

}

@article{@DOUXIN202501,

title = {Investigating the Impact of the COVID-19 Pandemic on Travel Mode Choice Behaviour – A Stated Preference Case in Wuhan, China},

author = {Xin Dou and Xiaofeng Pan and Tao Feng},

url = {https://hrcak.srce.hr/en/327643},

doi = {doi.org/10.7307/ptt.v37i1.736},

year  = {2025},

date = {2025-01-01},

journal = {Promet - Traffic&Transportation},

volume = {37},

issue = {1},

keywords = {},

pubstate = {published},

tppubtype = {article}

}

@article{LAI2025103423,

title = {Grid-level assessment on spatial equity in access to urban public facilities by vulnerable groups based on the multi-source data},

author = {Wenbo Lai and Dahu Lin and Zhuowen Li and You Peng and Weiqiang Zhou and Tao Feng},

url = {https://www.sciencedirect.com/science/article/pii/S0197397525001390},

doi = {https://doi.org/10.1016/j.habitatint.2025.103423},

issn = {0197-3975},

year  = {2025},

date = {2025-01-01},

journal = {Habitat International},

volume = {161},

pages = {103423},

abstract = {In the context of promoting human-centered sustainable cities and communities, achieving spatial equity in the development of urban public facilities (UPF) plays an essential role and faces a significant challenge. Although the regional and social disparity of UPF's distribution is well documented, researchers seldom consider the spatial distribution disparities between multiple types of UPF and vulnerable groups at a fine scale. This study built a comprehensive evaluation framework with a fine-grained model, which is comprised of grid-based partitioning, multi-source data crawling, cumulative opportunity model, and spatial statistical analysis, to analyze spatial equity in Guangzhou, China. The accessibility to UPF and the social vulnerability index (SVI) are measured utilizing the cumulative opportunity measure and principal component analysis (PCA). This study, therefore, analyzed the imbalance between the distribution of UPF and vulnerable groups. Moreover, the study accurately identified spatial inequity areas using the Gini coefficient and bivariate Moran's I. The results show that numerous areas within the study area have limited opportunities to access UPF. Nevertheless, the formation of multiple urban centers during urbanization has increased the diversity of accessible types of UPF within the community life unit. In addition, due to the aging population in the old city center and the complex population structure of urban villages, the areas in the old town often become high-value clusters spatially with SVI. In particular, due to planning problems caused by urban villages, these areas are often prone to become spatial inequity areas characterized by "high SVI - low comprehensive accessibility." This evaluation framework is user-friendly and suitable for application in other cities, helping to identify spatial inequity areas that require planning intervention and priority.},

keywords = {},

pubstate = {published},

tppubtype = {article}

}

@article{Zhang14052025,

title = {Transportation mode choice prediction using a new multi-class association rule model},

author = {Jiajia Zhang and Tao Feng and Harry J. P. Timmermans and Qing Dai and Zhengkui Lin and},

url = {https://doi.org/10.1080/23249935.2025.2497897},

doi = {10.1080/23249935.2025.2497897},

year  = {2025},

date = {2025-01-01},

journal = {Transportmetrica A: Transport Science},

volume = {0},

number = {0},

pages = {1–54},

publisher = {Taylor & Francis},

keywords = {},

pubstate = {published},

tppubtype = {article}

}

@article{JI2025106445,

title = {Stratified strategies for enhancing thermal comfort through multidimensional compactness optimization in urban built-up areas during heatwaves},

author = {Yifeng Ji and Ying Liu and Hongyu Tang and Zhitao Li and Yihang Bai and Tao Feng},

url = {https://www.sciencedirect.com/science/article/pii/S221067072500321X},

doi = {https://doi.org/10.1016/j.scs.2025.106445},

issn = {2210-6707},

year  = {2025},

date = {2025-01-01},

journal = {Sustainable Cities and Society},

volume = {127},

pages = {106445},

abstract = {Thermal comfort (TC) in built-up areas with varying levels of compactness is unevenly affected during heatwaves (HW). However, identifying zones that should prioritize compactness optimization to effectively enhance TC is often overlooked. This study constructed a research framework for enhancing TC through stratified planning strategies by identifying key compactness-optimized areas and patterns during HW. Taking the built-up area of Shenyang, China, as an example, the compactness index containing spatial, functional and socio-economic dimensions and the TC index were first constructed based on multi-source data. Afterwards, different types of compactness-optimized areas, dominant compactness in different regions, and trade-off and synergy patterns among various dimensions of compactness were revealed using a geographically weighted regression model (GWRM) and local bivariate spatial autocorrelation analysis. The results show that compactness decreases from the center to the periphery of the built-up area, while TC follows the opposite trend. A total of 10 types of compactness-optimized areas are identified, including 8 key types covering 40.472 % of the built-up area. Based on the trade-offs and synergies between different compactness dimensions, 8 optimization patterns are revealed, with synergistic optimization across all three dimensions representing the largest share (71.256 %). Furthermore, 4 optimization categories with different priorities are identified, each exhibiting distinct spatial patterns and targeted optimization strategies. These findings support hierarchical resource allocation and strategic intervention to enhance thermal comfort and promote climate-resilient cities.},

keywords = {},

pubstate = {published},

tppubtype = {article}

}

@article{HU2025104520,

title = {Examining the temporary use behavior of autonomous vehicles under uncertainty: A stated preference analysis},

author = {Yan Hu and Tao Feng and Mengxia Li},

url = {https://www.sciencedirect.com/science/article/pii/S096585642500148X},

doi = {https://doi.org/10.1016/j.tra.2025.104520},

issn = {0965-8564},

year  = {2025},

date = {2025-01-01},

journal = {Transportation Research Part A: Policy and Practice},

volume = {198},

pages = {104520},

abstract = {The rapid advancement of autonomous vehicles (AVs) is reshaping traditional parking, posing challenges for urban and transportation planning. This study investigates AV owners’ parking choice behavior to inform policy decisions. A stated choice experiment was designed to evaluate three parking modes including regular parking, cruising, and car-sharing, integrating dual uncertainties in parking fees and sharing revenues, an aspect largely overlooked in prior research. Unlike previous studies focusing solely on parking versus cruising decisions, we incorporate car-sharing as an additional behavioral option for AVs users upon reaching their destinations. Using a random parameter mixed logit-cumulative prospect theory model, we analyze AV owners’ risk attitudes and parking decisions under uncertainty. Results indicate different risk attitudes toward parking mode choice and varied preferences concerning the uncertain parking fee and revenue. Additionally, socio-demographic characteristics, travel purposes, waiting times, social influences, and congestion charges significantly impact parking choice behaviors. The findings provide valuable insights for urban planning, transportation policy, and car-sharing platform development.},

keywords = {},

pubstate = {published},

tppubtype = {article}

}

@article{LIU2025100192,

title = {SAFER-predictor: Sparse adversarial training framework for robust traffic prediction under missing and noisy data},

author = {Yutian Liu and Chengfeng Jia and Soora Rasouli and Jian Gong and Tao Feng and Melvin Wong and Tianjin Huang},

url = {https://www.sciencedirect.com/science/article/pii/S2772424725000320},

doi = {https://doi.org/10.1016/j.commtr.2025.100192},

issn = {2772-4247},

year  = {2025},

date = {2025-01-01},

journal = {Communications in Transportation Research},

volume = {5},

pages = {100192},

abstract = {Accurate traffic flow forecasting is essential for developing intelligent transportation systems (ITSs) to reduce congestion, optimize road management, and improve safety. While data-driven traffic prediction approaches have shown high accuracy, they rely heavily on precise measurements, making them vulnerable to perturbed environmental factors, like sensor malfunctions, data storage issues, and adverse weather conditions. To overcome the limitation, we propose SAFER-Predictor, a novel sparse adversarial training (Sparse AT) framework for enhancing the reliability of deep learning based spatiotemporal traffic prediction models. Sparse AT extends traditional adversarial training (AT) through a two-phase process: pre-training and fine-tuning. In the pre-training phase, the model is optimized to capture normal traffic patterns, enhancing predictive performance by understanding standard dynamics without external disruptions. In the fine-tuning phase, the focus shifts to strengthening robustness against corrupted inputs by employing an iterative min-max strategy during AT, optimizing performance for worst-case scenarios. Furthermore, we derive theoretical formulations that establish an upper bound on the model's prediction error following Sparse AT under certain noise levels. Experimental results indicate that incorporating Sparse AT into the representative traffic flow prediction models improves stability and ensures high accuracy under various perturbation scenarios.},

keywords = {},

pubstate = {published},

tppubtype = {article}

}

@article{ZHOU2025106146,

title = {Non-linear associations between the built environment and outdoor activity duration: An application of gradient boosting decision trees},

author = {Ting Zhou and Tao Feng and Astrid Kemperman},

url = {https://www.sciencedirect.com/science/article/pii/S0264275125004470},

doi = {https://doi.org/10.1016/j.cities.2025.106146},

issn = {0264-2751},

year  = {2025},

date = {2025-01-01},

journal = {Cities},

volume = {165},

pages = {106146},

abstract = {Outdoor activities benefit human physical and mental health. However, how the built environment influences the duration of location-based outdoor activities is unclear. To address this gap, we measured the built environment using different methods including Geographic Information Systems and a semantic segmentation technique with Google Street View images and associated these independent variables with outdoor activity duration by employing a gradient boosting decision tree model. We found neighborhood characteristics play a much more important role in influencing outdoor activity duration than location characteristics (features about a specific area where outdoor activities take place). Although all seventeen variables had a non-linear and threshold impact, distance to the nearest public transport station is the most significant predictor, followed by residential density, distance to the nearest school, and street connectivity. These findings allow urban planners and policymakers to give detailed advice on planning and designing built environments that better promote outdoor activities.},

keywords = {},

pubstate = {published},

tppubtype = {article}

}

@article{LI2025105252,

title = {Activity-aware urban area embedding with contrastive learning for intelligent transportation systems applications},

author = {Gen Li and Tao Feng and Dan He and Li Yan and Jiwon Kim},

url = {https://www.sciencedirect.com/science/article/pii/S0968090X25002566},

doi = {https://doi.org/10.1016/j.trc.2025.105252},

issn = {0968-090X},

year  = {2025},

date = {2025-01-01},

journal = {Transportation Research Part C: Emerging Technologies},

volume = {178},

pages = {105252},

abstract = {Embedding is a machine learning technique that represents data entities as continuous vector representations, capturing the underlying semantic relationships between them. Urban area embedding applies this concept to urban regions, representing each area as a vector that encapsulates its key characteristics. These embeddings enable models to better understand the relationships between different urban areas, facilitating applications such as traffic management, urban planning, and resource allocation. In this paper, we propose a comprehensive framework called AUAEC (Activity-aware Urban Area Embedding with Contrastive Learning) that integrates diverse open datasets including Location-Based Social Network (LBSN) check-ins, taxi flow data, and Points of Interest (POI) to produce enriched and context-aware region embeddings. To capture both mobility patterns and activity-aware semantics of LBSN users, we apply spatial interpolation based on road network, coupled with activity vector construction to represent user daily activity and movement patterns. To refine these embeddings into comprehensive urban regional representations, the AUAEC incorporates two complementary contrastive learning strategies: View-wise Contrastive Learning, which aligns representations across multiple data views, and Activity-aware Contrastive Learning, which captures inter-region relationships based on activity-aware semantics. The resulting embeddings are evaluated across four critical ITS tasks including land use distribution classification, traffic incident prediction, public transport delay prediction and traffic volume prediction using real-world data. Our approach demonstrates promising results, outperforming state-of-the-art solutions and highlighting the superiority of AUAEC in providing robust, contextual representations of urban areas for ITS and urban planning applications.},

keywords = {},

pubstate = {published},

tppubtype = {article}

}

@article{ZHAO2025960,

title = {Passengers’ vertiport choices for integrated urban air mobility and airline services},

author = {Ying Zhao and Yan Hu and Tao Feng and Anming Zhang},

url = {https://www.sciencedirect.com/science/article/pii/S0967070X25002811},

doi = {https://doi.org/10.1016/j.tranpol.2025.07.028},

issn = {0967-070X},

year  = {2025},

date = {2025-01-01},

journal = {Transport Policy},

volume = {171},

pages = {960-969},

abstract = {Vertiports are essential hubs for Urban Air Mobility (UAM) airport shuttles, however, passenger choice behavior in vertiports remains underexplored. In this paper, we conducted a stated choice experiment in Beijing and applied a mixed logit model with interaction effects to examine passengers’ heterogeneous choice preferences in the context of synchronized UAM and airline services. Results show passengers generally prefer small, nearby vertiports, integrated with public transport, and equipped with multiple facilities and services. Preferences vary notably among middle-income groups (100,000–200,000 CNY/year), and passengers with helicopter experience prefer less proximate vertiports due to noise and privacy concerns. Elasticity analyses underscore the critical influence of access distance, providing valuable insights for policymakers and UAM service providers in infrastructure planning and development.},

keywords = {},

pubstate = {published},

tppubtype = {article}

}

@article{WANG2025104623,

title = {The impacts of carbon emission regulatory policies on demand-responsive transit operation: a strategy to avoid service equity loss},

author = {Jiangbo Wang and Zhongyu Sun and Kai Liu and Jianjun Wu and Tao Feng},

url = {https://www.sciencedirect.com/science/article/pii/S0965856425002514},

doi = {https://doi.org/10.1016/j.tra.2025.104623},

issn = {0965-8564},

year  = {2025},

date = {2025-01-01},

journal = {Transportation Research Part A: Policy and Practice},

volume = {200},

pages = {104623},

abstract = {As one of the primary contributors to carbon emissions, road transportation systems are expected to be included in carbon emission management frameworks in the near future. Demand-responsive transit (DRT) will face a new constraint: carbon emission costs. Currently, no widely established framework exists for carbon emission management in transportation systems. Carbon tax and carbon trading are two commonly implemented carbon regulatory policies worldwide, and both are expected to be extended to the transportation sector. However, the effectiveness of these policies in incentivizing carbon reductions in DRT systems and the potential challenges they may pose remain uncertain. To fill this gap, this paper developed route planning models for the Open Capacitated Low-carbon Vehicle Routing Problem (OCLCVRP) and designed a series of experiments using standard benchmark instances to explore the impacts of carbon tax and carbon trading policies on the operation of DRT systems. The results indicate that although both carbon tax and carbon trading policies are effective in reducing carbon emissions, they lead to the loss of a fundamental attribute of DRT systems—service equity. As carbon prices rise, this loss of equity becomes more pronounced, ultimately increasing a higher failure rates of DRT systems and undermining the effectiveness of carbon emission regulatory policies. To mitigate this issue, this paper proposed a dynamic carbon quota-based carbon trading scheme. Through a case study, the proposed dynamic carbon quota proves to be effective for the public transportation sector in reducing carbon emissions while ensuring service equity under the carbon trading scheme.},

keywords = {},

pubstate = {published},

tppubtype = {article}

}

@article{XIA2025126242,

title = {Integration management of vessel transportation in complex waterway network of port clusters: A column generation-based solution approach},

author = {Zicheng Xia and Qi Tian and Tao Feng and Zijian Guo and Yun Peng and Ying Jiang and Wenyuan Wang},

url = {https://www.sciencedirect.com/science/article/pii/S0957417424031099},

doi = {https://doi.org/10.1016/j.eswa.2024.126242},

issn = {0957-4174},

year  = {2025},

date = {2025-01-01},

journal = {Expert Systems with Applications},

volume = {268},

pages = {126242},

abstract = {The integration of regional ports will heighten the complexity of maritime traffic in port waters, thus necessitating a unified management approach for this intricate traffic flow. Maritime Bureau implements the integration of multi-port vessel traffic organization based on the advance information provided. The staff manually arranged and organized plans, thereby eliminating planning conflicts during the actual implementation process. Inefficiency and limitation of manual scheduling result in frequent delays and cannot guarantee vessel safety. In this paper, the multi-port waterway network vessel traffic management problem (MPWN-VTM) under multi-rule constraints, e.g., navigation rules, speed control, priority weights of vessels, and tidal time windows requirements, in an actual waterway network is investigated. A mixed-integer linear program (MILP) is proposed to improve scheduling efficiency. The problem is formulated into a more compact set-partitioning formulation and a column generation (CG) algorithm is developed to solve this problem. Several acceleration strategies are used to improve the performance of CG and obtain the optimal or near-optimal solutions. Numerical experiments with the real port network demonstrate the efficiency and applicability of the algorithm by comparing it with practical port scheduling strategy, general heuristic algorithm, and branch-and-cut. The empirical result indicates that the proposed MILP can successfully handle the actual daily scheduling plan for up to 300 vessels within 15 min, with an optimality gap of no more than 5%. This work can be extended to provide intelligent support for potential regional port integration works by promoting the development of traffic management systems under multi-port vessel traffic organization.},

keywords = {},

pubstate = {published},

tppubtype = {article}

}

@article{Lv28122025,

title = {Integrating crowding and safety in the analysis of cyclists’ route preferences: insights from a hybrid choice model},

author = {Huitao Lv and Haojie Li and Fan Zhang and Xue Bai and Tao Feng},

url = {https://doi.org/10.1080/19427867.2025.2606311},

doi = {10.1080/19427867.2025.2606311},

year  = {2025},

date = {2025-01-01},

journal = {Transportation Letters},

volume = {0},

number = {0},

pages = {1–20},

publisher = {Taylor & Francis},

keywords = {},

pubstate = {published},

tppubtype = {article}

}

@article{math13244003,

title = {TFGCRN: Temporal–Frequency Graph Convolutional Recurrent Network for Incomplete Traffic Forecasting},

author = {Jiazhan Hu and Tao Feng},

url = {https://www.mdpi.com/2227-7390/13/24/4003},

doi = {10.3390/math13244003},

issn = {2227-7390},

year  = {2025},

date = {2025-01-01},

journal = {Mathematics},

volume = {13},

number = {24},

abstract = {Traffic forecasting is a crucial component that underpins an intelligent transportation system. Among the current mainstream forecasting algorithms, spatial–temporal graph neural networks (STGNNs), as the mainstream solution, have been used in traffic forecasting due to their ability to model spatial–temporal dependencies effectively. However, sensor failures caused by factors such as bad weather often lead to incomplete traffic data, which severely prevents STGNNs from modeling spatial–temporal dependencies and consequently degrades forecasting performance. To achieve accurate forecasting under incomplete traffic conditions, this paper proposes a Temporal–Frequency Graph Convolutional Recurrent Network (TFGCRN) model that embeds a Temporal–Frequency Graph Convolutional Network into gated recurrent units. During the recursive modeling process, TFGCRN fully leverages both global and local information to resist the adverse effects of missing values while generating more accurate spatial relationships, thereby achieving precise incomplete traffic forecasting. The experiments on four real-world datasets show that TFGCRN can achieve satisfactory results superior to multiple baselines and effectively adapt to different missing rates. Compared with the state-of-the-art baseline, TFGCRN can reduce forecasting error by 2–6%.},

keywords = {},

pubstate = {published},

tppubtype = {article}

}

@article{LAN2025105754,

title = {Role of remote working center in a hybrid work culture},

author = {Jieyuan Lan and Tao Feng},

url = {https://www.sciencedirect.com/science/article/pii/S026427512500054X},

doi = {https://doi.org/10.1016/j.cities.2025.105754},

issn = {0264-2751},

year  = {2025},

date = {2025-01-01},

journal = {Cities},

volume = {159},

pages = {105754},

abstract = {Remote Working Centers (RWCs) have gained popularity as alternative workspaces. However, to what extent are teleworkers willing to use RWCs in comparison to home and office considering transportation effects remains underexplored. This study investigates teleworkers' joint decisions on work location and transportation mode under different contexts. Using stated preference data collected in Tokyo Metropolis, Japan, a random parameter error component model was estimated, capturing the unobserved heterogeneity among teleworkers. The results indicate a preference for working in RWCs when tasks have a higher work intensity, involve teamwork, last between 4 and 6 h or longer. In general, the intention to travel to RWCs decreases with longer travel times. Females and high-income individuals are more likely to use RWCs. In accessing to RWCs, low-income individuals prefer active travel modes, while high-income individuals opt for public transportation or private cars. These findings provide insights for urban planners in shaping policies to promote the use of RWCs in a hybrid work culture.},

keywords = {},

pubstate = {published},

tppubtype = {article}

}

@article{ZHAO2025100990,

title = {Exploring the integration of urban air mobility into Mobility-as-a-Service: A stated preference analysis of commuters},

author = {Ying Zhao and Yan Hu and Tao Feng},

url = {https://www.sciencedirect.com/science/article/pii/S2214367X25000080},

doi = {https://doi.org/10.1016/j.tbs.2025.100990},

issn = {2214-367X},

year  = {2025},

date = {2025-01-01},

journal = {Travel Behaviour and Society},

volume = {39},

pages = {100990},

abstract = {This study introduces the concept of air mobility as a Service (AMaaS), integrating urban air mobility (UAM) into the framework of Mobility-as-a-Service (MaaS), with the goal of establishing highly synchronized mobility services. To facilitate the successful deployment of Urban Air Taxi (UAT) services, we aim to investigate people’s preferences for multimodal air taxi services. A stated choice experiment was designed incorporating pay-as-you-go options for various multimodal UAT services and subscriptions to ride-based discounts. Using data collected in Beijing, China, a random parameter error component model was estimated to identify preference heterogeneity among different individuals and potential correlations between alternatives. Our findings indicate a general preference for subscription schemes over pay-as-you-go options across all UAT services. The choice within AMaaS is significantly influenced by the various attributes of UAT alternatives and incentive measures, e.g., stronger government support and/or price discounts increase the probability of using AMaaS. Workers aged 44 or above, high-income groups, car owners, regular car commuters, individuals in managerial positions, and those having helicopter experience are more inclined to commute via multimodal UATs. These findings provide valuable insights for policy decision-making in the planning of UAM, especially when integrated into MaaS.},

keywords = {},

pubstate = {published},

tppubtype = {article}

}

@article{ZHAO2025103784,

title = {Assessment of passengers’ safety and risk attitudes on integrated urban air mobility and airline services},

author = {Ying Zhao and Yan Hu and Tao Feng and Anming Zhang},

url = {https://www.sciencedirect.com/science/article/pii/S0967070X25003270},

doi = {https://doi.org/10.1016/j.tranpol.2025.103784},

issn = {0967-070X},

year  = {2025},

date = {2025-01-01},

journal = {Transport Policy},

volume = {172},

pages = {103784},

abstract = {This study introduces the concept of Air Mobility as a Service (AMaaS) by integrating Urban Air Taxi (UAT) services into the Mobility as a Service (MaaS) framework, aiming to enabling seamless multimodal transportation. The objective is to investigate commuter preferences for adopting multimodal UAT services. A stated choice experiment was designed to capture joint choice of UAT-based alternatives and subscription schemes, alongside attitudinal measures assessing the influence of safety and risk perceptions on adoption behavior. Using data collected in Beijing, a hybrid choice model with latent variables was estimated. Results show that subscription-based schemes, particularly sustainable options like Bike + UAT and PT + UAT, are generally preferred over pay-as-you-go alternatives. Government support and discounts significantly increase adoption likelihood. Safety perceptions also play a critical role. Specifically, perceived UAT safety encourages adoption, while safety consciousness, and perceived UAT risks hinder the use of these services. Individuals with higher safety consciousness are less likely to use pay-as-you-go options, and those perceiving UAT as risky are less inclined to use subscription schemes, particularly Taxi + UAT. These findings provide valuable insights for policymakers and UAT service providers in designing effective policies and marketing strategies to promote UAT adoption.},

keywords = {},

pubstate = {published},

tppubtype = {article}

}

@article{ZHOU2025104426,

title = {Active travel and the built environment: A life-course perspective},

author = {Ting Zhou and Tao Feng and Astrid Kemperman},

url = {https://www.sciencedirect.com/science/article/pii/S0966692325003175},

doi = {https://doi.org/10.1016/j.jtrangeo.2025.104426},

issn = {0966-6923},

year  = {2025},

date = {2025-01-01},

journal = {Journal of Transport Geography},

volume = {129},

pages = {104426},

abstract = {Active travel has many health, environmental, economic, and social benefits, yet it is not clear how life events and factors from the built environment influence active travel over time. To address this gap and provide causal and dynamic evidence, this study used a 5-year longitudinal dataset from the Netherlands Mobility Panel and applied a Dynamic Bayesian Network approach to explore the temporal relationships between life events, built environment factors and active travel frequency, including frequency of conventional bike, e-bike, and walking. Results show that life events play a much more important role in influencing active travel frequency than the built environment factors. In addition, owning an e-bike has the greatest impact on cycling frequency, impeding conventional bike use and increasing e-bike use. Furthermore, childbirth has different temporal effects on the frequency of all three types of active travel. Moreover, workplace change negatively impacts e-bike use both concurrently and over time. These findings provide empirical evidence for urban planning and policymaking to promote active travel.},

keywords = {},

pubstate = {published},

tppubtype = {article}

}