<?xml version="1.0" encoding="utf-8"?><rss version="2.0" xmlns:atom="http://www.w3.org/2005/Atom"><channel><title>TRB Publications Index</title><link>http://pubsindex.trb.org/</link><atom:link href="http://pubsindex.trb.org/common/TRIS Suite/feeds/rss.aspx?tc=NN%3ACfc" rel="self" type="application/rss+xml" /><description></description><language>en-us</language><copyright>Copyright © 2015. National Academy of Sciences. All rights reserved.</copyright><docs>http://blogs.law.harvard.edu/tech/rss</docs><managingEditor>tris-trb@nas.edu (Bill McLeod)</managingEditor><webMaster>tris-trb@nas.edu (Bill McLeod)</webMaster><image><title>TRB Publications Index</title><url>http://pubsindex.trb.org/Images/PageHeader-wTitle.png</url><link>http://pubsindex.trb.org/</link></image><item><title>Multiobjective Vehicle Routing Optimization With Time Windows: A Hybrid Approach Using Deep Reinforcement Learning and NSGA-II</title><link>http://pubsindex.trb.org/view/2553276</link><description><![CDATA[This paper proposes a weight-aware deep reinforcement learning (WADRL) approach designed to address the multiobjective vehicle routing problem with time windows (MOVRPTW), aiming to use a single deep reinforcement learning (DRL) model to solve the entire multiobjective optimization problem. The Non-dominated sorting genetic algorithm-II (NSGA-II) method is then employed to optimize the outcomes produced by the WADRL, thereby mitigating the limitations of both approaches. Firstly, we design an MOVRPTW model to balance the minimization of travel cost and the maximization of customer satisfaction. Subsequently, we present a novel DRL framework that incorporates a transformer-based policy network. This network is composed of an encoder module, a weight embedding module where the weights of the objective functions are incorporated, and a decoder module. NSGA-II is then utilized to optimize the solutions generated by WADRL. Finally, extensive experimental results demonstrate that our method outperforms the existing and traditional methods. Due to the numerous constraints in VRPTW, generating initial solutions of the NSGA-II algorithm can be time-consuming. However, using solutions generated by the WADRL as initial solutions for NSGA-II significantly reduces the time required for generating initial solutions. Meanwhile, the NSGA-II algorithm can enhance the quality of solutions generated by WADRL, resulting in solutions with better scalability. Notably, the weight-aware strategy significantly reduces the training time of DRL while achieving better results, enabling a single DRL model to solve the entire multiobjective optimization problem.]]></description><pubDate>Wed, 31 Dec 2025 15:48:35 GMT</pubDate><guid>http://pubsindex.trb.org/view/2553276</guid></item><item><title>Exploring the Correlation Between Winter Severity Indices and Winter Maintenance Costs: Insights and Comparative Analysis from Nationwide Practices</title><link>http://pubsindex.trb.org/view/2633003</link><description><![CDATA[Winter maintenance operations are essential for safe and efficient transportation, yet unpredictable weather introduces major challenges in cost and material planning. A single severe snowfall can deplete resources midway through the season, while milder-than-expected winters may result in unused surpluses. The Winter Severity Index (WSI) provides a standardized measure of winter intensity, offering a promising tool to address these challenges. This study explores the relationship between WSIs and winter maintenance costs through a nationwide survey of state DOTs and Clear Roads survey data. The Clear Roads survey analysis involves regression models, assessing the relationship between the Accumulated Winter Season Severity Index (AWSSI) and winter maintenance expenditures across different U.S. climate divisions. AWSSI was used not as a recommended tool for operations, but as a nationally available and consistent index to investigate whether winter severity is statistically associated with cost trends on a broad scale. The study also assessed nonlinear regression models, which yielded similar results to linear models; the linear model was, therefore, retained for its simplicity and interpretability. Results show significant correlation within various climate clusters, indicating that weather severity can explain cost variability. The study reveals that WSIs can improve winter maintenance budget management, but their effectiveness is regionally nuanced. Customized WSIs are essential to improve resource planning, as the findings indicate that AWSSI does not work optimally in all geographic contexts. The study also identified best practices by state departments of transportation (DOTs), such as integrating WSIs with real-time tools such as the Road Weather Information Service and the Maintenance Decision Support System, to enhance planning accuracy and responsiveness.]]></description><pubDate>Wed, 03 Dec 2025 09:19:38 GMT</pubDate><guid>http://pubsindex.trb.org/view/2633003</guid></item><item><title>Investigating Grocery Shopping Choice and Behavioral Changes during COVID-19 Using Machine Learning Methods</title><link>http://pubsindex.trb.org/view/2628345</link><description><![CDATA[This study leverages machine learning (ML) models to predict grocery shopping channel choices in Florida, comparing findings across two waves of stated preference (SP) survey data (collected February to April and November to December 2021). The SP survey design considered three alternatives (online, curbside, in store) and five cost and time attributes, (product price, delivery cost, travel time, delivery time, shopping time). The data were analyzed to examine changes in the importance of the explanatory variables including cost and time attributes, personal attitudes, and sociodemographic characteristics. In addition to ML models, a mixed logit model was applied to provide a reference to compare with the ML models. Among the six ML models evaluated, XGBoost was the best performing, and product price was the most influential variable predicting grocery shopping channel choices. The results also showed that recreational shoppers tended to prefer the online and curbside options over in-store shopping, during both waves. Furthermore, cost-consciousness was linked to a preference for the curbside pickup alternative during the second wave, and some attitudes changed between the two waves. The comparison revealed that both mixed logit and ML models consistently identified cost-related and demographic factors as key influences on shopping choices. ML models offered more nuanced insights into evolving factors like vehicle ownership, whereas mixed logit models provided better interpretability, especially in capturing the stable impact of socioeconomic factors. The findings of this study provide a better understanding of consumer behaviors, and their changing attitudes toward and perceptions of different shopping channels.]]></description><pubDate>Tue, 25 Nov 2025 08:51:44 GMT</pubDate><guid>http://pubsindex.trb.org/view/2628345</guid></item><item><title>Comparing Location-Based and Home Address–Based Approaches in Evaluating the Cost to Society of Crashes and Congestion</title><link>http://pubsindex.trb.org/view/2612392</link><description><![CDATA[Traffic congestion and safety pose two of the largest challenges for transportation planners and engineers worldwide, placing significant economic, environmental, and social burdens on the public. In the United States crashes were estimated to have cost the American public $340?billion in 2019, with congestion costs estimated to be over $166?billion in 2017. With funding from the Infrastructure Investment and Jobs Act (IIJA) becoming available to transportation agencies around the country, it is vital that agencies identify the most impactful infrastructure projects on which to prioritize funding to address these costs. At the same time, the United States is seeking to address the lack of infrastructure spending in disadvantaged communities in previous decades. To guide this infrastructure spending, data are required to properly identify the locations most in need of funding. This study assesses the trade-off between data accuracy and data availability in two methodologies to estimate the cost to society of crashes and congestion, focusing on the comparison between county-level location-based aggregation and home address–based approaches. The study assessed crash data in the state of Massachusetts and used the results of the Boston Region Metropolitan Planning Organization’s TDM23 travel demand model to evaluate these trade-offs. The study found that the county-level location-based aggregation acceptably mitigated for the unknown home addresses of those involved when creating a metric for the cost to society of crashes. By contrast, it did not mitigate well for the costs of congestion.]]></description><pubDate>Fri, 24 Oct 2025 16:43:21 GMT</pubDate><guid>http://pubsindex.trb.org/view/2612392</guid></item><item><title>Research on Lightweight Design and Fatigue Life Optimization of a Commercial Vehicle Frame</title><link>http://pubsindex.trb.org/view/2596622</link><description><![CDATA[To reduce vehicle manufacturing costs and improve overall vehicle reliability, a lightweight optimization design was implemented on the frame of an electric commercial vehicle based on HyperWorks and Lsight, and its various performance impacts were analyzed and verified. Based on the static, dynamic, and modal performance analysis of the original frame, 19 key components were identified through sensitivity analysis. An optimization model was established using optimal Latin hypercube test sampling method combined with the radial basis function surrogate model, and the multi-objective non-dominated sorting genetic algorithm-II was used to solve the optimal design scheme. The optimized design achieved the goals of reducing frame weight by 12.24% and increasing first-order natural frequency by 14.27%, effectively enhancing ride comfort and structural safety. The optimized frame meets the fatigue life requirements under various working conditions, demonstrating the effectiveness of this design in ensuring vehicle reliability and safety.]]></description><pubDate>Fri, 12 Sep 2025 16:14:17 GMT</pubDate><guid>http://pubsindex.trb.org/view/2596622</guid></item><item><title>The Attractiveness of Electric Vehicles for Heterogeneous Consumers: A Range-Cost-Emission Comparative Analysis</title><link>http://pubsindex.trb.org/view/2591973</link><description><![CDATA[This study offers a complete comparative method that examines the long- and short-term range–cost–emission attractiveness of different electric vehicle (EV) types for heterogeneous consumers. Despite the centrality of range satisfaction in EV adoption decisions, existing studies rely heavily on subjective user perceptions. This study bridges this gap by proposing the first data-driven framework to quantify range satisfaction, comparing EVs and conventional vehicles (CVs) through real-world mobility patterns and subsidy scenarios. The distribution of daily vehicle miles traveled was used to cluster mobility patterns and driving demands. In addition, EV subsidy rollback scenarios were modeled to quantify their impact on EV adoption in both long- and short-term ownership situations. The findings show that the attractiveness of EVs for different user types for long- and short-term adoption varies. EVs may not provide the same level of range satisfaction that CVs do, especially for users with high travel demands. Although driving costs can be reduced with EVs, not all users will benefit from them. Before acquiring an EV, users must examine their usage patterns and driving demands, as well as prepare for long- or short-term ownership. Changes in subsidies can affect not only the attractiveness of EVs but also the adoption duration. Finally, acknowledging that the environmental attractiveness of EVs is tied to the electricity production profile, it also varies based on vehicle utilization pattern.]]></description><pubDate>Fri, 22 Aug 2025 13:42:51 GMT</pubDate><guid>http://pubsindex.trb.org/view/2591973</guid></item><item><title>Experiences in the Use of Mini and Modular Roundabouts by Highway Agencies</title><link>http://pubsindex.trb.org/view/2582435</link><description><![CDATA[While roundabouts continue to be a proven solution for addressing safety and efficiency at intersections in the U.S., the increasing costs of construction and right-of-way needs for conventional roundabouts have increased momentum of the mini roundabout as a viable option in favor of the traditional roundabout. By definition, a mini roundabout is a type of roundabout in which the central island is fully traversable and intended to be utilized by trucks or other large vehicles. This reduces the footprint of the intersection, and mini roundabouts can often be retrofitted within existing intersection footprints. Approximately 300 mini roundabouts with fully traversable central islands have successfully been constructed in the U.S., and states with eight or more mini roundabouts include Washington, Minnesota, North Carolina, Texas, Maryland, Michigan, Kentucky, Georgia, Ohio, Colorado, Arkansas, and Oregon. These states’ general experiences with mini roundabouts have been overwhelmingly positive. A modular roundabout is a specialized roundabout that incorporates prefabricated materials to reduce excavation, paving and drainage work, environmental impacts, utility and right-of-way impacts, construction duration, and ultimately, cost. The modular material is typically used for the central island and splitter islands but may also be used for outside curbing. The material is glued or anchored on top of existing pavement. In addition to these custom-made materials, modular roundabouts employ striping and may include quick-build curbs and flex-posts to delineate vehicle paths. Modular roundabouts are less common in the U.S., but several have been constructed in California, Georgia, North Carolina, Virginia, and Wisconsin. Most modular roundabouts have been received positively by agency staff and the public.  The objective of the domestic scan was to identify leading states and describe the experiences and lessons learned that may be valuable to others who may be considering using mini or modular roundabouts. The scope of the scan included a range of topics for each form of roundabout, including the following: Capacity and traffic efficiency data, Crash history (before and after), Design and performance checks, Construction costs, Installation and construction timeline, Maintenance, and Public/community acceptance.]]></description><pubDate>Sat, 16 Aug 2025 11:31:23 GMT</pubDate><guid>http://pubsindex.trb.org/view/2582435</guid></item><item><title>Sustainable and Resilient Schedule Coordination for Connecting Flights at Airport Hubs</title><link>http://pubsindex.trb.org/view/2567040</link><description><![CDATA[The arrival delay of connecting flights in a hub-and-spoke network (HSN) can increase operational costs and degrade the level of service, particularly for transfer passengers. Developing a sustainable and resilient schedule by incorporating suitable buffer times can significantly improve the likelihood of successful connections. This paper presents a mathematical model that considers probabilistic flight arrivals at the hub airport to minimize total costs, including those associated with aircraft, gate utilization, and passenger delays. A dual-step search algorithm is developed to determine the optimal buffer times. The model is applied in a case study to optimize the buffer times of connecting flights at a hub airport in Xi’an, China. Additionally, the impacts of model parameters on decision variables and total costs are analyzed. Results indicate that implementing the proposed model can significantly enhance HSN resilience while reducing operational and user costs.]]></description><pubDate>Mon, 23 Jun 2025 08:44:37 GMT</pubDate><guid>http://pubsindex.trb.org/view/2567040</guid></item><item><title>A Novel Choquet Integral-Based Regression Approach for Analyzing Freight-Train Derailment Severity</title><link>http://pubsindex.trb.org/view/2537003</link><description><![CDATA[Freight-train derailments represents a significant hazard to human safety and the economy. However, few studies have explored the contributory and interactive characteristics of risk factors influencing the economic costs of derailments across various track classes. We proposed a novel Choquet integral-based multivariate polynomial regression (CI-MPR) model to predict the economic cost via the states of four risk factors—number of derailed cars (DC), derailment speed (DS), loading factor (LO), and residual train length (RL). The results indicated that the CI-MPR model attained better predictions than the Choquet integral-based multiple linear regression (CI-MLR) model. The Shapley value and interaction indices were used to elucidate the contributory and interactive characteristics of these risk factors. Across Track Classes 1 to 5, DC emerged as the most influential risk factor, contributing 34.22%, 45.81%, 51.82%, 63.94%, and 64.74%, respectively. As for interactive characteristics, DC prevalently interacted with DS and RL in a superadditive way across all track classes. Notably, the pair (DC, LO) demonstrated critical superadditive interactions in Track Classes 1, 2, and 3, while appearing as additive and subadditive in Track Classes 4 and 5. The findings suggest that managing operational speeds and enhancing the stability of leading cars could significantly reduce economic losses. Moreover, controlling operational speed may prevent interactions with LO, potentially mitigating the adverse effects of the superadditive interaction between DS and LO in higher track classes. This study provides insights into identifying critical risk factors for freight-train derailment and devising track-class-specific mitigation strategies to reduce derailment severity and enhance operational safety.]]></description><pubDate>Wed, 16 Apr 2025 11:23:08 GMT</pubDate><guid>http://pubsindex.trb.org/view/2537003</guid></item><item><title>State of the Art Transportation Noise Barriers</title><link>http://pubsindex.trb.org/view/2431656</link><description><![CDATA[This article contains a state of the art study of transportation noise barriers by the assistant program director of the National Cooperative Highway Research Program. He reviews these areas of interest: recent reports on noise and noise barrier research; planning for noise control; noise control measures; when and where barriers are needed; kinds of barriers; criteria for planning and design; paying for noise barriers; objections to barriers; noise barrier theory; design procedures; barrier costs; and noise barrier construction examples.]]></description><pubDate>Mon, 30 Sep 2024 11:43:44 GMT</pubDate><guid>http://pubsindex.trb.org/view/2431656</guid></item><item><title>Post-Purchase Trip Heterogeneity: Exploring the Impact of Free and Paid Return Deliveries on Shopping and Transport Mode Choices in the USA</title><link>http://pubsindex.trb.org/view/2417320</link><description><![CDATA[This paper explores aspects of the post-purchase trips generated by the necessity to return purchased clothing to the shops. Particular focus is given to online shopping that has become common for clothing purchases. This study’s novelty lies in delving into the underlying reasons behind post-purchase trips, particularly those initiated when customers seek to return clothing to retailers. It examines the impact of free and paid return delivery options as key factors driving consumers’ decisions in relation to clothing returns. The study consists of two branches and leverages the random utility maximization theory. The first branch focuses on the impact of free and paid return options on the preferred shopping channels by utilizing a stated preference dataset collected from 507 US online shoppers. The second branch of this study employs the revealed preference dataset and aims to explore the return trip behavior. The willingness-to-pay values estimated for the free return delivery option are higher among female online shoppers compared with males—$7.42 and $6.65?per delivery, respectively. It was found that among the identified “returners,” 84.62% of males and 79.91% of females showed a strong reliance on private cars for their return trips. The potential environmental consequences of return trips were evaluated, focusing on the case of the USA. Additionally, the estimated marginal probability effect revealed that factors such as an aging population, car ownership, and number of children in households positively influence car usage for post-purchase trips. The study’s implications for stakeholders are discussed.]]></description><pubDate>Thu, 15 Aug 2024 09:28:20 GMT</pubDate><guid>http://pubsindex.trb.org/view/2417320</guid></item><item><title>Research on Joint Distribution Path Planning of Electric Logistics Vehicles with Different Recharge Modes</title><link>http://pubsindex.trb.org/view/2414255</link><description><![CDATA[To address the route planning issues under the community group purchase model for joint delivery, this study thoroughly considers electric logistics vehicles with different recharging methods. The objective is to minimize the sum of operating costs, recharging costs, time window penalty costs, and carbon emission costs. Separate multi-objective optimization models for route planning are constructed for both charging and battery-swapping logistics vehicles. An improved seagull optimization algorithm, guided by the golden sine strategy of the Lévy flight guidance mechanism, is employed to avoid local optima and enhance the solution efficiency. The feasibility of the models and the algorithm is verified through simulation examples. Experimental results show that, at the current stage, battery-swapping logistics vehicles display significant advantages over charging electric logistics vehicles. Although battery-swapping logistics vehicles extend delivery time, they can reduce the total delivery costs to a certain extent. Therefore, the future development prospects of battery-swapping logistics vehicles will be even broader.]]></description><pubDate>Fri, 09 Aug 2024 08:40:49 GMT</pubDate><guid>http://pubsindex.trb.org/view/2414255</guid></item><item><title>Railway Signal Digitalization with the European Rail Traffic Management System and Positive Train Control: Industry 4.0 Expectations and Reality</title><link>http://pubsindex.trb.org/view/2414224</link><description><![CDATA[Positive train control (PTC) and European Rail Traffic Management System (ERTMS) are digital railway signal systems in North America and Europe, respectively. They are frequently described as interchangeable, but they are not. This paper explains the history and motivations for each continent, and the general technical and capability differences between the two signal systems. In general, North America revised their signal systems to respond to safety concerns, and Europe committed to replacing their signals to encourage cross-border train traffic. ERTMS is significantly more expensive than PTC, and the cost has been justified with expectations of greater capacity. Multiple studies find no basis for large capacity increases after implementation of ERTMS. In addition, the added cost of ERTMS threatens an already weak rail freight market in Europe.]]></description><pubDate>Fri, 09 Aug 2024 08:40:49 GMT</pubDate><guid>http://pubsindex.trb.org/view/2414224</guid></item><item><title>Path Optimization of Container Multimodal Transportation Considering Differences in Cargo Time Sensitivity</title><link>http://pubsindex.trb.org/view/2373767</link><description><![CDATA[To optimize the container multimodal transportation path selection problem with differences in cargo time sensitivity, this paper introduces the concept of mixed time windows and establishes an integer programming model from the perspective of carriers. The objective is to minimize the total transportation cost while meeting the customer’s on-time delivery requirements. The model imposes hard time window constraints on cargoes with on-time delivery requirements and soft time window constraints on general cargoes. A bi-level genetic algorithm is employed to solve the model. A case study is conducted to demonstrate the effectiveness of the proposed approach. Experimental results show that the bi-level genetic algorithm’s convergence ability, efficiency, stability, and global search capabilities are far superior to those of traditional single-level genetic algorithms. This indicates that the use of multi-stage solution approaches can mitigate the shortcomings of genetic algorithms in dealing with high-dimensional problems. In addition, when analyzing the effects of time penalty intensity and time window on decision-making results, it is found that the more time-sensitive cargoes are more likely to be affected by the time window setting, and will tend to sacrifice costs to ensure that the time demand is met. At the same time an increase in demand for time-sensitive services is accompanied by an increase in operating costs. Accurate assessment and classification of time sensitivity of cargoes can effectively prevent wasting resources in unnecessary areas, and ultimately maximize decision outcomes and ensure the most efficient use of resources.]]></description><pubDate>Mon, 29 Apr 2024 18:29:55 GMT</pubDate><guid>http://pubsindex.trb.org/view/2373767</guid></item><item><title>Choice-Based Service Region Assortment Problem with Statewide Synthetic Data: Toward Equitable Transportation Design</title><link>http://pubsindex.trb.org/view/2334490</link><description><![CDATA[Incorporating individual user preferences in statewide transportation planning is of great importance regarding revenue management and behavioral equity. However, an enduring challenge is that consistent population travel data remains scarce, particularly in underserved and rural areas. Moreover, large-scale optimization models are computationally demanding when considering stochastic travel demands in a discrete choice model (DCM) framework. These can be addressed with a combination of synthetic population data and deterministic taste coefficients. The authors formulate a choice-based optimization model, in which the mode share in each block group-level trip origin-destination (OD) is determined by a set of deterministic coefficients reflecting user preferences. In that case, statewide service region design becomes an assortment optimization problem with known parameters and linear constraints, which can be efficiently solved through linear or quadratic programming (depending on variant). The authors test the method using two hypothetical new mobility services considered for New York State: service A has shorter trip duration while higher trip cost; service B has longer trip duration while lower trip cost. The proposed model is applied to optimize their service region with one of the three objectives: (1) maximizing the total revenue; (2) maximizing the total change of consumer surplus; (3) minimizing the disparity between disadvantaged and non-disadvantaged communities. The results show that objectives 1 &amp; 2 lead to higher revenue while might increase inequity, while objective 3 decreases the welfare disparity by up to 8.97% at the cost of losing almost half of the revenue.]]></description><pubDate>Tue, 20 Feb 2024 09:16:43 GMT</pubDate><guid>http://pubsindex.trb.org/view/2334490</guid></item></channel></rss>