Shan Liu , Zhenhao Meng , Rui Jiang , Zhengli Wang , Ya Zhang , Ying-En Ge
{"title":"Dynamic recovery sequence optimization of subway-bus network with constrained reinforcement learning","authors":"Shan Liu , Zhenhao Meng , Rui Jiang , Zhengli Wang , Ya Zhang , Ying-En Ge","doi":"10.1016/j.trd.2026.105238","DOIUrl":"10.1016/j.trd.2026.105238","url":null,"abstract":"<div><div>Subway and bus networks constitute the backbone of urban public transportation and cater to the daily commuting needs of city residents. For damaged subway-bus networks caused by extreme events, aligning the repair sequence with dynamic passenger flow is crucial to reducing passenger delays. Most existing studies on station recovery strategies ignore the dynamic changes of passenger flow, and this study aims to design the station recovery sequence for the subway-bus network considering dynamic passenger flow. To enhance practicality, a constrained reinforcement learning framework is formulated to incorporate passenger demands, mobility distance, and repair duration into a Markov decision process. Moreover, dynamic graph convolutional networks are adopted to capture time-varying characteristics of passenger flow. The reward function proposed in this work integrates node importance, recovery progress, transferred passengers, and excessive distance penalty to enable efficient post-disaster recovery. Empirical analysis reveals that this methodology increases the number of transferred passengers by 5.53%-11.30%.</div></div>","PeriodicalId":23277,"journal":{"name":"Transportation Research Part D-transport and Environment","volume":"154 ","pages":"Article 105238"},"PeriodicalIF":7.7,"publicationDate":"2026-05-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"146153276","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":1,"RegionCategory":"工程技术","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
Huazhi Yuan , Tao Dong , Tengwang Ge , Qian Chen , Chaohui Wang
{"title":"Long-lasting emission reduction modified asphalt: preparation, pavement performances and emission reduction efficacy","authors":"Huazhi Yuan , Tao Dong , Tengwang Ge , Qian Chen , Chaohui Wang","doi":"10.1016/j.trd.2026.105254","DOIUrl":"10.1016/j.trd.2026.105254","url":null,"abstract":"<div><div>To reduce the emission of asphalt fumes from asphalt pavements, an asphalt emission reduction modifier (TAP) and its modified asphalt were prepared. Its high and low-temperature rheological properties and medium-temperature fatigue resistance were analyzed and evaluated. The influence of different TAP dosages on the emission reduction efficacy of asphalt pollutants was explored. The long-term emission reduction efficacy of long-lasting emission reduction modified asphalt was verified. The emission reduction efficacy of TAP on asphalt VOCs was verified by HS-GC–MS test. This lays a solid foundation for further popularizing low-carbon asphalt in the transportation infrastructure field. The results show that TAP can effectively improve the high-temperature deformation resistance, low-temperature crack resistance, and fatigue resistance of asphalt. Long-lasting emission reduction modified asphalt with different TAP dosages all show significant emission reduction efficacy. Among them, when the TAP content is 20%, the emission reduction rates of gaseous and solid pollutants during the construction period both exceed 55%, and those during the operation period both exceed 45%. The long-term emission reduction efficacy of long-lasting emission reduction modified asphalt is remarkable. The coefficients of variation of the emission reduction rates of gaseous and solid pollutants in asphalt fume before and after long-term aging are both less than 4%.</div></div>","PeriodicalId":23277,"journal":{"name":"Transportation Research Part D-transport and Environment","volume":"154 ","pages":"Article 105254"},"PeriodicalIF":7.7,"publicationDate":"2026-05-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"146110262","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":1,"RegionCategory":"工程技术","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
Amir Shafiee , Hanieh Rastegar Moghaddam , Sudheer Ballare , Jane Lin
{"title":"An alternative last-mile delivery system: Leveraging microhubs and crowdshipping","authors":"Amir Shafiee , Hanieh Rastegar Moghaddam , Sudheer Ballare , Jane Lin","doi":"10.1016/j.trd.2026.105239","DOIUrl":"10.1016/j.trd.2026.105239","url":null,"abstract":"<div><div>Urban freight demand has increased with e-commerce, intensifying traffic congestion, air pollution, and greenhouse gas emissions. To address these challenges, we propose a delivery model combining microhubs and crowdshipping (M+C). Trucks visit only microhubs, while crowdshippers—individuals making personal trips—handle first and last-mile deliveries. This reduces truck related emissions, noise pollution, and congestion in urban areas. We formulate the truck routing problem as a Many-to-Many Pickup-and-Delivery Problem with Split Loads (M2MPDPSL), which allows heterogeneous commodities and multi-visits to microhubs. We synchronize trucks and crowdshippers by linking parcel availability and time windows across both layers. An enhanced Adaptive Large Neighborhood Search (ALNS) with tailored operators is developed to solve the problem and validated using both standard benchmark instances and customized scenarios. We then compare the M+C system to a traditional Depot-Based (DB) approach and results show that M+C reduces operational cost and truck miles traveled, especially in dense customer settings.</div></div>","PeriodicalId":23277,"journal":{"name":"Transportation Research Part D-transport and Environment","volume":"154 ","pages":"Article 105239"},"PeriodicalIF":7.7,"publicationDate":"2026-05-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"146135120","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":1,"RegionCategory":"工程技术","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
{"title":"Impact of electric vehicle subsidies on a mixed urban delivery fleet","authors":"Yongling Gao , Yuan Qiao , Meng Xu , Yupeng Jiang , Taesu Cheong , Chi Xie","doi":"10.1016/j.trd.2026.105237","DOIUrl":"10.1016/j.trd.2026.105237","url":null,"abstract":"<div><div>Electric vehicles (EVs) play an important role in reducing carbon emissions for logistics companies. This study examines three types of government subsidy policies for EVs: (1) vehicle purchase subsidy (<em>PS</em>); (2) fleet electrification subsidy (<em>FS</em>); and (3) mileage subsidy (<em>MS</em>). For each subsidy policy, a bi-objective optimization model is proposed to minimize the fleet cost and carbon emissions. The model considers the en-route charging decisions of EVs and carbon emissions of internal combustion engine vehicles (ICEVs). This study reveals that among the three policies (i.e., <em>PS</em>, <em>FS</em>, and <em>MS</em>), the policy resulting in the largest number of EVs in the fleet can maximize social welfare only when the per-vehicle environmental impact of EV production is low; When the impact is high, the greatest social welfare could be achieved under the policy yielding the fewest EVs. Varying individual policies has divergent effects: increasing the <em>PS</em> or <em>FS</em> can enhance fleet electrification and reduce emissions, whereas increasing the <em>MS</em> may primarily boost the average EV mileage. Comparing the same total subsidy level and using the <em>PS</em> as a benchmark, both the <em>FS</em> and <em>MS</em> policies can enhance fleet electrification and reduce emissions, but their limitations differ: the <em>MS</em> policy’s environmental benefits may diminish at high subsidy levels, whereas the <em>FS</em> policy may raise the total fleet cost. Any of the three policies can achieve the lowest total fleet cost, highlighting the need to tailor subsidy choice to a specific objective.</div></div>","PeriodicalId":23277,"journal":{"name":"Transportation Research Part D-transport and Environment","volume":"154 ","pages":"Article 105237"},"PeriodicalIF":7.7,"publicationDate":"2026-05-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"146135142","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":1,"RegionCategory":"工程技术","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
{"title":"Mitigating container port congestion and carbon emissions through AI and capacity sharing","authors":"Shan Lyu, Ruisi Yang","doi":"10.1016/j.trd.2026.105251","DOIUrl":"10.1016/j.trd.2026.105251","url":null,"abstract":"<div><div>Container port congestion results in substantial economic losses and carbon emissions. To examine the effectiveness and interaction of two common congestion mitigation strategies, capacity sharing and artificial intelligence (AI) investment, this study develops a game-theoretical model of two competing ports (a dominant port and a non-dominant port), and explores their impacts on carbon emissions. The results show that capacity sharing can alleviate the non-dominant port’s congestion by transferring service volume to the dominant port. This process reduces the non-dominant port’s carbon emission but increases those of the dominant port, and total emissions may rise when the non-dominant port’s capacity is relatively low. In contrast, AI investment consistently alleviates congestion and reduces carbon emissions. However, when the two ports have similar capacities, they may fall into a prisoner’s dilemma in their AI investment decisions. Moreover, we identify the capacity and cost thresholds for the non-dominant port’s optimal congestion mitigation strategy.</div></div>","PeriodicalId":23277,"journal":{"name":"Transportation Research Part D-transport and Environment","volume":"154 ","pages":"Article 105251"},"PeriodicalIF":7.7,"publicationDate":"2026-05-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"146191343","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":1,"RegionCategory":"工程技术","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
{"title":"The potential uptake and climate impacts of Hydrogen-Fuel-Cell vehicles in Beijing","authors":"Ying Zhang , Xingjun Huang , Junbei Liu , Chengxiang Zhuge","doi":"10.1016/j.trd.2026.105253","DOIUrl":"10.1016/j.trd.2026.105253","url":null,"abstract":"<div><div>Hydrogen-fuel-cell vehicles (HFCVs) can deliver near-zero life-cycle emissions with green hydrogen, yet urban uptake remains negligible. Most diffusion studies treat cities as static backdrops. To overcome this, we build a dynamic, spatially explicit agent-based model (SelfSim-HFCV), calibrated to Beijing (2018–2023) and simulating to 2035, which co-evolves demographics, land-use change, and vehicle markets. Under the baseline, almost no HFCVs emerge. Redirecting growth to the Tongzhou Subcenter barely alters HFCV uptake but reallocates charging-station density southeast rather than increasing totals. Introducing demographic heterogeneity boosts HFCV adoption and reveals profiles: owners are typically older, wealthier, and concentrated in child-free, retiree, multi-license households. Only synchronized purchase subsidies with hydrogen-refueling-station (HRS) rollout shift applications to HFCVs and deliver sustained emission reductions, while HRS alone has limited effect due to scaling delays. These findings highlight the importance of coordinating infrastructure timing, urban form, and social composition, suggesting a transferable framework for urban hydrogen transition assessment.</div></div>","PeriodicalId":23277,"journal":{"name":"Transportation Research Part D-transport and Environment","volume":"154 ","pages":"Article 105253"},"PeriodicalIF":7.7,"publicationDate":"2026-05-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"146110845","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":1,"RegionCategory":"工程技术","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
{"title":"Love thy neighbor – reducing driving through green slot labels in e-commerce","authors":"Nora Svarstad Ytreberg , Guri Natalie Jordbakke","doi":"10.1016/j.trd.2026.105252","DOIUrl":"10.1016/j.trd.2026.105252","url":null,"abstract":"<div><div>With the rapid growth of e-commerce and its negative externalities, an emerging literature examines the potential of nudging consumers towards more environmentally friendly delivery options. We contribute by analyzing a quasi-natural experiment of green labeling, investigating revealed preference data from 35,828 Norwegian customers over five weeks in 2024. A green label marks delivery slots that enable route optimization, thereby minimizing driving and related externalities. We exploit the quasi-random assignment of green labels, and a pricing randomized control trial, to estimate the label effect and the willingness to pay for green delivery using a multinomial logit model.</div><div>Results show that the probability of choosing a slot more than doubles when labeled as green. Moreover, we find a willingness to pay of €1 for green delivery, ultimately suggesting that consumers value green services and that green labeling is an effective tool for online retailers to minimize delivery costs.</div></div>","PeriodicalId":23277,"journal":{"name":"Transportation Research Part D-transport and Environment","volume":"154 ","pages":"Article 105252"},"PeriodicalIF":7.7,"publicationDate":"2026-05-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"146135151","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":1,"RegionCategory":"工程技术","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
Chenglin Liu , Yang Liu , Zhigang Xu , Liang Wang , Haotong Tang
{"title":"En-route charging and sequencing policies for heterogeneous electric truck platoons","authors":"Chenglin Liu , Yang Liu , Zhigang Xu , Liang Wang , Haotong Tang","doi":"10.1016/j.trd.2026.105264","DOIUrl":"10.1016/j.trd.2026.105264","url":null,"abstract":"<div><div>Electrifying long-haul freight with electric truck platoons can improve safety and efficiency, yet deployment remains constrained by limited range and further complicated by nonlinear charging characteristics and vehicle heterogeneity in battery capacity, energy consumption, and charging power. These challenges call for charging policies that are not only energy-feasible but also operationally implementable for corridor operations. We present a policy design and evaluation framework that produces implementable en-route charging and intra-platoon sequencing policies for heterogeneous electric truck platoons. The upper layer generates platoon charging policies that specify where to stop and how long to charge under nonlinear battery dynamics, using a schedule-integration algorithm. The lower layer derives sequencing policies via Monte Carlo Tree Search to maximize effective charging power by reallocating position-dependent aerodynamic benefits, thereby reducing cross-vehicle energy disparities. Evaluations on a real expressway corridor with heterogeneous platoons show consistent reductions in total charging times compared with two benchmarks. Sensitivity analysis quantifies how parameter dispersion affects policy performance and offers operational policies for long-haul corridors.</div></div>","PeriodicalId":23277,"journal":{"name":"Transportation Research Part D-transport and Environment","volume":"154 ","pages":"Article 105264"},"PeriodicalIF":7.7,"publicationDate":"2026-05-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"146190265","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":1,"RegionCategory":"工程技术","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
Ann-Kristin Dugstad , Noureddine Bénichou , Maxine Berthiaume , Paul Geoerg , Steve Gwynne , Amanda Kimball , Kamryn Kubose-Peutz , Hui Xie , Enrico Ronchi , Max Kinateder
{"title":"Drilling for disaster preparedness: Insights from a community wildfire evacuation exercise","authors":"Ann-Kristin Dugstad , Noureddine Bénichou , Maxine Berthiaume , Paul Geoerg , Steve Gwynne , Amanda Kimball , Kamryn Kubose-Peutz , Hui Xie , Enrico Ronchi , Max Kinateder","doi":"10.1016/j.trd.2026.105258","DOIUrl":"10.1016/j.trd.2026.105258","url":null,"abstract":"<div><div>Community evacuation exercises (or drills) are one approach residents and authorities use to train wildfire emergency procedures in the Wildland Urban Interface (WUI). This paper presents results from a drill performed in Roxborough Park, a WUI community in Colorado, USA, in 2024. Observer and self-report data were collected to derive resident preparatory actions, pre-travel, travel, and total evacuation times, as well as route choice and the drill’s impact on their perceived preparedness. It took more than 28 minutes until 90 % of residents began traveling and more than 48 minutes until 90 % had completed their evacuation. Most of the participants reported following the instructions for the evacuation route, with a minority taking a shorter or more familiar route. The work underlines the value of drills for improving community disaster preparedness, providing data for developing/testing computational models, and deepening our understanding of human behavior in wildfire scenarios.</div></div>","PeriodicalId":23277,"journal":{"name":"Transportation Research Part D-transport and Environment","volume":"154 ","pages":"Article 105258"},"PeriodicalIF":7.7,"publicationDate":"2026-05-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"146152661","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":1,"RegionCategory":"工程技术","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
{"title":"From images to insights: ChatGPT and Google Street View for walkability assessments","authors":"Donghwan Ki , Zhenhua Chen","doi":"10.1016/j.trd.2026.105245","DOIUrl":"10.1016/j.trd.2026.105245","url":null,"abstract":"<div><div>A better understanding of walkability in minority neighborhoods requires close attention to <em>route quality</em>, which includes diverse streetscape features, but comprehensive large-scale assessment remains challenging. This study introduces a novel approach that integrates ChatGPT with Google Street View (GSV) to conduct scalable, fine-grained walkability evaluations in Los Angeles, California. For each GSV image, ChatGPT generates both numeric walkability scores and narrative descriptions of negative aspects. LDA topic modeling is used to uncover latent walkability topics in these narratives. Spatial analyses of both outputs reveal that minority neighborhoods consistently exhibit lower walkability scores and face disproportionate challenges, such as unkempt streetscapes. Validation against human ratings and computer vision models is conducted to assess the reliability of ChatGPT-based evaluations. The findings demonstrate that ChatGPT can capture nuanced microscale features and social cues beyond the capabilities of existing off-the-shelf computer vision methods. This approach provides a context-rich, scalable tool for targeted and equity-focused interventions in minority neighborhoods.</div></div>","PeriodicalId":23277,"journal":{"name":"Transportation Research Part D-transport and Environment","volume":"154 ","pages":"Article 105245"},"PeriodicalIF":7.7,"publicationDate":"2026-05-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"146110846","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":1,"RegionCategory":"工程技术","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}