The car-following behaviour of autonomous vehicles (AVs) directly influences road capacity and infrastructure design. In this study, a potential game-based car-following model (PG-CFM) was developed, and its parameters were calibrated using a reinforcement learning-guided genetic algorithm. A potential game cost function was employed to represent multi-objective decision making under spacing, speed-difference and acceleration constraints. The model was calibrated and validated using real-world trajectory data from the Waymo data set and compared with the intelligent driver model and the Gipps model. The results showed that the PG-CFM outperformed both benchmark models in predictive accuracy and stability. Under open-loop validation, the model achieved a mean absolute error of 1.00 m/s2 and a root mean square error of 1.31 m/s2, representing a 4.2% reduction relative to the Gipps model. The error distribution was also more concentrated, indicating improved robustness. The calibrated parameters suggested that AVs operate with an expected time headway of approximately 1.18 s while maintaining smoother acceleration profiles. This behaviour reflects a stable balance between safety and traffic efficiency. These findings provide quantitative support for headway setting, road-capacity assessment and infrastructure planning for autonomous mobility systems.
Traditional traffic conflict analysis typically relies on indicators such as time to collision () and post-encroachment time. However, while comparing these one-dimensional indicators and their derivatives can to some extent evaluate vehicle-pedestrian collision risk, they struggle to intuitively reflect the macroscopic interaction states between vehicles and pedestrians in a planar environment. Considering the polar-coordinate perception characteristics of light detection and ranging sensors (utilising angle and distance data), a traffic conflict relationship representation method based on measured target bearing angle () is proposed, which can directly represent traffic conflict relationships using polar-coordinate data without further calculating traditional conflict indicators such as . By analysing variation, it can directly assess the traffic conflict state. By employing graph envelope and regression methods, the state boundary of traffic interactions at different distances from the conflict point can be defined and expressed as a boundary function. On this basis, the approach can further support autonomous vehicles in evaluating the yielding necessity. Moreover, the study employs the random forest method to train an illustrative yielding speed model as an example, demonstrating the potential of the proposed methods to support yielding speed planning. The proposed approach may also provide insights for studies on restricting human-driven vehicles from forcing priority passage or delaying deceleration.
Pedestrian safety at unsignalized crossings remains a critical challenge, as the absence of traffic signals increases the risk of accidents, particularly among vulnerable groups such as children and young adults. The aim of this systematic review is to identify and evaluate effective infrastructure measures to improve pedestrian safety at unsignalized intersections. A structured search was conducted across major engineering databases, and eligible studies were assessed using a standardized weight of evidence framework. The identified infrastructure measures were categorized into four key groups: engineering modifications, additional lighting, signage enhancements and smart technologies. This review provides a comprehensive synthesis of current best practices and highlights the importance of integrating quantitative assessment frameworks, such as the iRAP star rating system, to guide future infrastructure planning and policy decisions aimed at enhancing pedestrian safety at unsignalized crossings.
This study quantifies the full costs of automobility across five Australian states: New South Wales, Victoria, Queensland, South Australia and Western Australia. The analysis estimates the total annual cost of the vehicle economy at A$316.12 billion (A$1 = 0.53) pound. Of this, A$158.64 billion (50.18%) is borne by the public, including direct budgetary expenditures and indirect externalities such as congestion, pollution, public parking construction, vehicle crashes and land use. The remaining A$157.48 billion (49.82%) is directly incurred by private consumers through vehicle purchase, operation and parking. These findings challenge the belief that drivers fully bear automobility costs.
This study is an investigation of the effect of shared autonomous vehicle (SAV) deployment on private car ownership among households in Tehran. Data from 318 households were collected through a field survey and analysed using a binary logistic regression model. The receiver operating characteristic curve, Hosmer-Lemeshow test and model coefficients confirmed the model's accuracy and goodness of fit. The findings show that familiarisation with SAVs significantly increases the willingness to use these vehicles and reduces the intention to own private cars. Specifically, 59% of respondents indicated a willingness to reduce their use of private vehicles, and 54% reported that they would reconsider purchasing a new car. Although socioeconomic variables, such as household income and the number of employed household members, remain influential, their effects diminish after the introduction of SAVs. Lower costs, improved accessibility and such issues as parking shortages were identified as the main drivers encouraging the substitution of SAVs for private vehicles. Furthermore, a random forest model revealed that household size and the number of employed members are most significant in predicting usage. Overall, the results indicate that adopting SAVs can reduce dependence on private car ownership and contribute to notable improvements in urban traffic conditions.
This study introduces a multi-objective slot-allocation model designed for coordinated airport clusters, with a focus on adaptive fairness. The model optimises the trade-off between minimising schedule deviations, promoting fairness among airlines and airports, and limiting individual flight adjustments. Utilising an epsilon-constraint method, the model transforms the multi-objective problem into a single-objective framework, with dynamically adjustable fairness constraints based on real-time flight volume. A novel piecewise function links the trade-off parameter (epsilon) to flight volume, allowing the model to adaptively balance fairness and efficiency. Key operational constraints - airport and waypoint capacity limits, turnaround times and maximum deviation thresholds - are incorporated. A case study of the Beijing-Tianjin-Hebei airport cluster validates the model, showing significant improvements in fairness and operational efficiency. Slot-allocation adjustments of 75 min (Beijing Capital airport), 70 min (Beijing Daxing airport; ) and 35 min (Tainjin Binhai airport; ) were achieved, with fairness indices improving at (from 1.43 to 1.11) and (from 2.42 to 0.93). The adaptive model ensures equitable resource distribution among airlines, reducing maximum fairness deviation from 1.35 to 0.51. Sensitivity analysis confirms the model's robustness, providing a scalable and practical solution for multi-airport systems facing dynamic traffic demands.
To enhance the design rationality and material utilisation of heavy-haul railway sleepers, this study takes the widely used type IIIa prestressed concrete sleeper from the Shuohuang railway as a case study. Based on the bi-directional evolutionary structural optimisation () method, topological solutions are first generated and then explicitly interpreted as strut-and-tie models, leading to two optimised reinforcement schemes (model and model ). Using Abaqus software, a comparative non-linear finite-element analysis of the mechanical performance between the optimised schemes and the conventional design (model I) was conducted. The results demonstrate that by bending up prestressed tendons, the optimised schemes achieve a more uniform stress distribution in the reinforcement, significantly improve the ultimate bearing capacity (with increases of 13.39% for model and 24.89% for model ) and exhibit superior crack resistance compared to the conventional design. Moreover, the proposed designs enhance structural efficiency without additional material use, achieving an implicit steel saving of 12-18% and a corresponding carbon dioxide emission reduction of 0.044-0.066 kg 2-eq per sleeper. This method provides an innovative and intelligent design approach for sleepers, which aligns well with the principles of low carbon dioxide and energy-saving construction, offering practical value for sustainable railway infrastructure development.
Research on predicting urban air quality using machine learning techniques typically employs global models that assume spatial homogeneity. This study aims to comprehensively evaluate the predictive limitations of global and station-based XGBoost methods and to examine the assumption of spatial homogeneity in complex and data-constrained environments such as Istanbul. To examine the link between traffic and air pollution, data from a single month were analysed, minimising seasonal variability. Global models showed satisfactory performance for coarse particulate matter (10; R-2 = 0.92), but their effectiveness for nitrogen dioxide (2) was limited due to their inability to distinguish factors affecting local distribution (R-2 = 0.35). However, the best accuracy was achieved with the proposed station-based methodology (R-2 = 0.93 for nitrogen dioxide), resulting in an 85% improvement in root mean squared error. Furthermore, feature significance analysis and pure prediction trials show that traffic data, when considered as a single variable, can explain approximately 7-10% of the prediction variance in the models. This finding demonstrates that, contrary to popular belief, the impact of road traffic on urban air quality is quite limited, while region-specific spatial and meteorological factors have a more significant effect. Consequently, it is concluded that standard and traffic-focused interventions are insufficient for effective environmental management, and the adoption of region-specific methods is necessary.
Integration of innovative solutions into the transportation sector have led to the emergence of intelligent transportation systems () to reduce traffic congestion and accidents, provide environmentally friendly, safe and cost-effective infrastructure and increase efficiency. A key prerequisite for public institutions, local authorities and industry stakeholders to benefit from is the availability of qualified technical staff. This study examined the presence of courses within civil engineering departments across universities in T & uuml;rkiye. The analysis revealed that only four out of 143 undergraduate programmes, 14 out of 111 master's programmes and seven out of 69 doctoral programmes offer dedicated courses. This demonstrates that only a limited proportion of civil engineering students receive education. Expert interviews emphasised the urgent need to incorporate courses into the civil engineering curriculum. Experts also emphasised the need to supplement theoretical content with field training and simulation studies. Survey results further indicated strong agreement among participants regarding the importance of hands-on applications. Overall, the findings suggest that education and awareness in T & uuml;rkiye are still at an early developmental stage; however, increasing demand for qualified staff and rising interest in the field indicate strong potential for rapid advancement in the near future.
To fill a research gap, this paper aims to assess service quality for ferry transportation using the Servqual measurement scale. In doing so, the integration of fuzzy analytical hierarchy process (AHP) and importance and performance analysis (IPA) was developed to identify priority areas for improvement for ferry operations. Additionally, ferry operators (FOs) in the south of Vietnam were investigated to validate the proposed research model. The paper’s originality is as follows: first, this paper expands the Servqual framework to make it predictive of customer loyalty and satisfaction. Second, the integration of fuzzy set theory into AHP enables the capture of respondents’ subjective judgments in terms of service requirement attributes, thereby limiting biased and uncertain evaluation. Third, the utilisation of the IPA model allows FOs to identify priority areas for scarce resource allocation, thus contributing to the enhancement of ferry service quality and economic growth, which is relevant to UN SDG 3.
The safety of high-speed train operations is a critical issue in railway transportation. The wheel-rail relationship is the connection between vehicles and tracks, and its condition determines the safety of high-speed railways. This article conducts an analysis of the characteristics of vehicle response data and the response characteristics caused by short, medium and long wavelength irregularities in the track. Utilising the comprehensive inspection train's detection of the entire section's dynamic response, a multi-wavelength irregularity evaluation method is proposed. This involves the use of axle box acceleration (ABA), bogie frame acceleration (BFA) and car body acceleration (CBA). The response characteristics of ABA, BFA and CBA at different running speeds, along with the sensitive frequency components at different speeds, are analysed. It is proposed to use the integrated ABA to obtain the vertical track displacement and combine this with the root mean square of acceleration after bandpass filtering to diagnose track irregularities. Envelope analysis is performed on the vertical displacement after bandpass filtering. Subsequently, peak extraction is conducted to identify the positions of track slabs, base plates and simply supported girder bridges. For changes occurring in a short period, the proposed method can serve as an auxiliary tool in evaluating the track status.
Overburnt brick aggregate (OBBA) has lower strength and durability relative to natural stone aggregate (NSA). This study evaluates the impact of nano-chemicals on the characteristics of cement-stabilised OBBA base course layers. Mixtures SACN.25, SACN.5, SACN.75 and SACN1 containing 0.25 to 1 kg/m3 of nanocompounds, respectively, were used for surface treatment. Based on California bearing ratio and unconfined compressive strength testing results, SACN1 was ascertained as the optimal mixture. The X-ray diffraction analysis indicates that calcium-silicate-hydrate and kaolinite minerals enhance the strength and cohesiveness of the SACN1 mixture, respectively. Scanning electron microscopy analysis confirms that organosilane creates water-resistant alkyl siloxane layers, enhancing the durability of road surfaces. Flexural and durability testing confirm that SACN1 provides sufficient strength and resilience against environmental variations. Moreover, the SACN1 mixture complies with the standard values prescribed by India's Ministry of Rural Development and the Indian Roads Congress requirements. Consequently, OBBA can serve as a substitute stone aggregate in road building when modified with nanochemicals. A Random Forest-based machine learning (ML) framework was employed to model the non-linear relationship between mix constituents and performance parameters and to perform multi-objective optimisation. The ML results successfully identified the optimal nano-modified mix that maximised strength and stiffness while minimising permeability, showing strong agreement with experimental observations.
This paper addresses the challenge of locating electric vehicle (EV) charging stations in large cities by proposing an improved sparrow search algorithm (ISSA), an enhancement of the original bio-inspired sparrow search algorithm (SSA). While SSA has shown effectiveness, it suffers from premature convergence and limited population diversity in complex, high-dimensional problems. ISSA mitigates these drawbacks through three modifications: dynamic inertia weights to balance exploration and exploitation; chaotic initialisation with a logistic map to increase population variety; and a population diversity index to prevent stagnation. A mathematical model for site selection is developed to minimise installation costs and meet demand while respecting urban constraints such as capacity limits and green zones. The research aims to improve SSA with adaptive components, construct a site-selection model and validate ISSA in a simulated urban environment. Simulation results in a hypothetical city demonstrate ISSA's superior performance, achieving 41.9% faster convergence, 21.1% lower total installation costs and 32.6% higher computational efficiency compared to SSA. In terms of charging infrastructure planning, the optimised station layout ensures effective coverage of EV charging demand while respecting station capacity constraints and urban land-use limitations, leading to a more balanced and cost-efficient deployment of charging stations.
Attitudes towards transport modes influence delivery of the twin imperatives of moving towards lower carbon dioxide transport while improving people's health and well-being. This study provides a macro-level assessment of social value across eight transport modes in the UK, forming the social component of a broader research agenda examining transport's economic and environmental dimensions. A nationally distributed questionnaire (n = 300) collected pre-pandemic data on ten key attributes, including travel time, cost, comfort, safety, environmental impact and health benefits. Utilising statistical analysis, the relationship between overall and individual evaluations was investigated, revealing significant positive correlations. The results confirmed the relevance of the aforementioned factors in decision making. Younger participants provided more positive evaluations than older participants, suggesting that older individuals may feel underserved by transport options. Transport evaluations were independent of gender or ethnicity. Car ownership correlated with higher appreciation of car travel and lower appreciation of taxis, but did not affect evaluations of public transport or active modes. Qualitative responses clarified the reasons behind transport mode choices, identifying time and comfort as the primary influences. By offering a consistent, pre-Covid-19 baseline, the study enables comparison with macro-level economic and environmental analyses and supports more integrated transport policy design.
The promotion of sustainable travel methods, such as public transportation, walking and bike-sharing, is being carried out in many countries around the world to raise awareness of the harmful effects of motorised traffic on the environment and form sustainable travel habits. Bike-sharing is considered a valuable option as it contributes to emission-reduction goals. This study investigates the transferability of a unified light gradient boosting machine (LightGBM) framework for bike-sharing demand prediction across three distinct socio-economic and climatic urban archetypes, namely Seoul, Washington D.C. and London, using variables including temperature, humidity, wind speed, season, hour, working day or holiday and location. While previous research focuses on localised models, this study tests the hypothesis that a single, high-fidelity model can transcend geographical heterogeneity. The results, validated through ten-fold cross-validation to ensure robustness, show that the predictive LightGBM model has a coefficient of determination, R2, of 0.947, root mean square error of 195.532 and mean absolute error of 107.548. Shapley additive explanations interpretability reveals that while temporal cycles and thermal comfort are universal predictors, the location feature captures latent socio-technical maturity, where London exhibits significantly higher peak-hour demand intensity compared to Washington D.C. and Seoul.
This study tackles inefficient urban mobility in Indian cities by showing how bus trip rate (trip generation) modelling can strengthen public transit planning and mitigate congestion and pollution. Predictive models are developed across multiple Indian cities using multiple linear regression (MLR) and artificial neural networks (ANN), finding that ANN consistently achieves higher predictive accuracy - especially in high-density contexts - by capturing non-linear relationships in mobility patterns. Sensitivity analysis highlights trip purpose and age as dominant predictors: bus reliance varies by trip purpose, while ridership declines with age, particularly among older populations. Population density positively influences bus use, underscoring stronger dependence on public transport in dense areas. Gender also proves significant, reinforcing the need for safer, more accessible, gender-inclusive systems to enhance women's mobility. These results motivate targeted interventions - route optimisation tailored to diverse travel needs, improved access for vulnerable users and gender-sensitive policies - alongside strengthening networks in dense corridors, integrating inclusive infrastructure and aligning with national sustainability agendas (e.g., Smart Cities Mission, National Action Plan on Climate Change). To address prior limitations, a generalised model is proposed that integrates common socio-demographic drivers in developing-economy contexts and is applicable across cities of varying sizes.
This study presents a graph-based framework for managing urban transport infrastructure and identifying potential mobility hubs within the Seoul metropolitan rail network in the Republic of Korea. The approach integrates traditional network centrality analysis with graph neural networks to capture both structural influence and flow mediation. Using operational schedule data, Bonacich power and random-walk betweenness centrality were embedded into a graph learning model to evaluate node importance beyond the limitations of conventional shortest-path assumptions. The results reveal that top nodes exhibit high multimodal potential, acting as strategic connectors within the metropolitan transit structure. By linking network-derived hub scores with public bicycle usage data, the analysis identifies spatial overlaps between structural centrality and micromobility demand. These findings support road space reallocation and multimodal integration strategies that enhance sustainable and inclusive accessibility. The proposed framework expands the methodological scope of transport planning by combining network science and machine learning. It provides a data-driven basis for infrastructure management and policy development toward low carbon dioxide, human-centred and resilient mobility systems. The proposed framework expands transport planning by combining network science and machine learning, providing a data-driven basis for infrastructure management, mobility hub planning and policy development toward low carbon dioxide and resilient urban transport systems.
The strategic placement of advance guide signs (AGSs) is crucial in ensuring optimal driving efficiency and safety at expressway exits. However, the extant literature contains a paucity of research on the placement distance of AGS at expressway exits. The objective of this study is to establish the optimal design for AGS at exit ramps on eight-lane expressways across varying levels of service (LOSs). The present study establishes a bidirectional eight-lane vehicle lane-changing model by integrating a selection of vehicle lane-change points (pre-lane change), minimal safe distances between preceding and following vehicles (post-lane change) and the vehicle lane-changing process. Subsequently, a method for calculating the lane-change success rate on expressways is proposed. Finally, the success rate of vehicles exiting the eight-lane expressway under four LOS conditions was calculated using conditional probability while also investigating the relationship between the AGS deployment location and the expressway exit success rate. To ensure a high driving-out success rate across various LOSs, it is recommended that AGSs be positioned at distances of 3.5 km, 2.5 km, 1.5 km and 0.9 km from the exit ramp. The research findings can supplement the existing specifications for AGS deployment on expressways.
The optimal placement of railway maintenance depots is a critical, yet complex, component of modern transport system planning, with significant implications for operational efficiency, cost management and network reliability. This study introduces a decision-support model that integrates the picture fuzzy (PiF) analytic hierarchy process and the PiF evaluation based on distance from average solution to evaluate candidate depot locations under uncertainty. Applied to the metropolitan railway network in Istanbul, the proposed model incorporates quantitative and qualitative criteria, including cost, accessibility, operational capability, environmental resilience and social factors. The results indicate that cost-related aspects and travel time exert the most significant influence on depot site selection, whereas social factors play a secondary role. The proposed framework enables planners to systematically assess alternative sites and supports the adoption of advanced computational methods in transport infrastructure planning. Overall, this research provides actionable guidance for practitioners and highlights the value of uncertainty-aware, multi-criteria approaches in technology-driven transport system development. The findings align with UN Sustainable Development Goals 9 (Industry, Innovation and Infrastructure) and 11 (Sustainable Cities and Communities) by promoting resilient transport infrastructure.
