The allocation of resources in transport network investments is central to shaping accessibility, sustainability, and equity within infrastructure systems. Yet, balancing diverse public preferences with equity considerations remains a persistent challenge, often resulting in decisions that fail to adequately reflect the needs of different societal groups. This study addresses this gap by investigating public preferences for transport network investments in Australia through a discrete choice experiment with 2,050 respondents. Using a Latent Class Choice Model (LCCM), we capture heterogeneity in preferences and identify two distinct groups. The first, labelled Environmental, Societal, and Governance (ESG)-centric, prioritises environmentally friendly, safe, and inclusive transport options. The second, labelled Self-centric, favours affordability, reduced travel times, and maximisation of personal benefits. By distinguishing these classes, this study contributes new evidence on the trade-offs that shape investment preferences and offers policy recommendations that emphasise accessibility, safety, sustainability, and innovative funding models. The findings provide actionable insights for designing transport policies that are both equitable and responsive to diverse societal needs.

The conventional approach to modelling last-mile distribution operations involves use of the discrete formulation method which renders a representative mathematical model necessitating use of sophisticated solution techniques. While such an intensive approach is justified when decision-makers require a precise plan to support operational planning, this level of precision is redundant for strategic planning, wherein available information is representative but not necessarily exact. To this end, continuous approximation (CA) method offers a practical alternative with use of continuous density functions that estimate parameters approximately, thus striking a balance between estimation accuracy and computational effort. However, typical CA-based routing frameworks assume a simplified distribution environment that significantly limits the capability of such frameworks to accurately model last-mile operations in real-world distribution environments, such as postal service, e-commerce, q-commerce, waste collection, emergency service etc. Thus, the objective of this work is to develop a holistic CA framework for the Vehicle Routing Problem (VRP) to facilitate strategic last-mile distribution planning within diverse logistic settings. Specifically, this study synthesizes a total of 1080 distribution environments, each with peculiar distribution structure and distinctive customer characteristics. Subsequently, it optimizes last-mile distribution for distribution environment instance using an Adaptive Large Neighborhood Search (ALNS) metaheuristic. Finally, this work develops a robust CA-based functional form through symbolic regression to precisely estimate the total distribution tour length across the various distribution environments. In doing so, this work supports strategic decision-making for a wide scope of logistic operators that may have representative but not necessarily exact information.

Operationalizing social justice is challenged by the absence of a universal equity definition and inherent policy trade-offs. This study bridges theoretical social justice principles with operational decision-making by integrating four commonly discussed distributive justice theories—Utilitarianism, Rawls’ Egalitarianism, Prioritarianism, and Capabilitarian Sufficiency—into a bilevel bus frequency optimization model. It introduces linearized equity-oriented formulations to reflect the distinct distributive principles of these theories and proposes justice theory-driven equity metrics to evaluate distributive impacts using different ethical criteria. Using cumulative opportunity as a unit of distribution, the proposed frameworks and metrics are applied to Canberra’s southern suburbs. The results demonstrate that the choice of justice framework is not a neutral modelling decision but a normative one, with each framework producing distinct redistributive patterns and community-level trade-offs. This highlights the importance of aligning equity frameworks with policy objectives and governance contexts, as these choices ultimately shape public perception, political feasibility, and the long-term societal goals of transport investment.

Abstract War conflicts constantly generate spatiotemporal disruption waves, particularly affecting transport systems. Limited transport observability during conflicts challenges assessing real-time disruptions and elucidating underlying causes and consequences. However, predicting event-specific system response and capacity loss is paramount for rehabilitation and resiliency. This study proposes a comprehensive framework that leverages pervasive data from conflict regions to automate time-series traffic models development, derive multiple time-series mobility metrics, and discern the spatiotemporal patterns through statistical learning to isolate war-induced disruptions. The disruptions are categorized based on plausible war events to quantify cause-specific impacts for three Ukrainian cities. Kyiv is most severely disrupted by migration, and Mariupol by occupation and battles with impact area, accessibility loss, and capacity reduction up to 39 km2, 300%, and 85%, respectively. This approach provides a rapid, exploratory alternative for understanding the immediate and dynamic consequences of conflicts on transport systems in large-scale, data-scarce conflict environments, informing resilient transport planning.

This study addresses the problem of managing traffic in urban transportation networks via path-based congestion pricing. Path-based congestion pricing policies consist of rewarding or tolling users for their path selections with the objective of mitigating system-wide congestion effects. The design of optimal path-based congestion pricing policies is notoriously difficult due to the large number of paths existing in transportation networks. Hence, both the problems of identifying candidate paths for pricing and that of computing optimal path-based congestion pricing policies are challenging. This study addresses these challenges by presenting novel statistical regression-powered optimization methods based on machine learning. We consider a bilevel optimization problem where a network planner aims to minimize congestion by providing path-based reward credits to commuters who are modeled as selfish agents minimizing a generalized cost function. We develop a statistical regression-powered heuristic approach to solve this path-based congestion pricing problem at scale. Our methodology integrates machine learning and optimization techniques to generate representative path sets and to model the relationship between congestion effects and path-based credit allocation. Sampling procedures and feature selection are used to synthesize a training data for a statistical regression model of congestion. Multiple supervised learning approaches are explored. The trained models are embedded in a surrogate optimization problem to determine path credits. An algorithm is designed to find feasible and efficient solutions to the bilevel optimization problem. Numerical experiments are conducted on small to large size networks. A comprehensive comparison is conducted between statistical regression-powered optimization methods, an exact branch-and-bound algorithm, and a model-based heuristic. The results demonstrate the efficiency and the computational scalability of the proposed statistical regression-powered optimization methods for solving problem instances based on large-scale networks.

Ethical considerations in evacuation network planning are crucial for ensuring the strategy's effectiveness and equity, as well as for increasing public trust in the government and the emergency management system. Whilst existing research often focuses on equity, prioritisation, humanitarianism, etc., there is a gap in directly analysing the ethical dimensions. This study aims to address these gaps by leveraging insights from existing literature to resolve ethical dilemmas and develop a comprehensive framework for ethical research. The scope of this review is limited to ethical considerations in the planning and design stages, rather than operational or post-disaster management phases. Our main contributions in this study are threefold: (1) examine ethical concepts and dilemmas in evacuation network planning, detailing key relationships of concepts and literature analysis; (2) summarise and evaluate qualitative and quantitative research approaches on disaster evacuation ethics, highlighting their importance; and (3) identify research gaps in disaster evacuation ethics research and propose a framework to address these issues and guide future work. We are concerned with conflict dilemmas in ethical concepts such as equity and prioritisation, internal and external contradictions of vulnerable groups, and so on. To this end, we propose a new concept, Ethical Cascading Failure (ECF), based on the literature review. Finally, we present a research pathway for evacuation network planning that addresses future ECF, along with a framework integrating multiple philosophical perspectives. The intended and potential beneficiaries of this study are government officials, planners, and individuals interested in ethical considerations in evacuation planning.

Transport resilience work is often split across separate data preparation scripts, network models, simulation tools, image inspection systems and reports. This fragmentation makes it difficult to move from an observation to a tested and reviewable decision. We introduce ResiliFlow, an open transport world model concept and an implemented platform for infrastructure resilience, response and recovery. The platform connects two workspaces. Disaster Transport Resilience Analysis provides six map-centred functions for critical-road and critical-area identification, recovery prioritisation, disruption routing, resilience testing and scenario simulation. AI-based Transport Infrastructure Perception and Decision Support organises street-level and satellite evidence, detects visible road, footpath and kerb conditions, and prepares these observations for human-reviewed intervention planning. Both workspaces share an eight-step cycle of perception, prediction, model development, verification, execution, decision, feedback and memory. Research Validation records assumptions and checks, while a local Assistant and an optional multi-provider large language model Copilot translate user questions into bounded calls to executable tools. We document the platform architecture, representative mathematical models, interface evidence and computer-vision learning results. Examples show accurate recognition across eight visible-condition classes, while compact error analysis demonstrates how difficult cases guide continued learning. ResiliFlow shows how transport models can become an inspectable, reusable and question-led system rather than a collection of disconnected analyses. The accompanying release is intended to support research collaboration, public scrutiny and extension under institutional review.

This study presents an innovative multistage methodology for decomposing urban traffic flows into light vehicle (LV) and heavy vehicle (HV) categories, addressing a critical gap in transportation network analysis. Utilizing data from Greater Sydney's road network, we develop a comprehensive approach comprising three main stages: origin-destination (OD) matrix estimation using RapidEx, quadratic programming optimization for HV/LV proportion estimation, and XGBoost regression for generalization. Our analysis examines associations between HV proportions and urban characteristics, including points of interest (POI), nightlight intensity, and zonal attributes. The XGBoost model achieves a test R2 of 0.637, demonstrating strong predictive power for real-world applications. Through SHAP (SHapley Additive exPlanations) analysis, we uncover complex nonlinear relationships between nightlight intensity and HV proportions, with significant interaction effects between urban features. The model performs particularly well in predicting common urban HV proportion ranges (0.2-0.6), reflecting typical urban traffic compositions. These findings provide valuable insights for urban planning and policy development, especially in contexts where detailed vehicle classification data are limited.

Traditional disaster evacuation planning has long emphasised efficiency, typically by minimising evacuation time or cost. With increasing attention to equity and social sustainability, new conflicts have emerged, such as whether to prioritise vulnerable groups, protect minority interests, or respect individual autonomy while ensuring collective order. These conflicts represent critical ethical dilemmas. This article introduces the novel concept of Ethical Cascading Failure (ECF), which describes how the violation of a single ethical principle can propagate across multiple principles and stakeholders, eventually leading to systemic breakdowns of evacuation governance. To illustrate this, qualitative analyses of recent disaster cases including the 2025 Southern California wildfires and the 2022 New South Wales floods, demonstrate how budget cuts, delayed warnings and biased communication escalated into wider ethical failures. For the quantitative analysis, three methods are proposed: complex network simulation to capture cascading propagation, multi-objective optimisation to balance efficiency and fairness, and explainable deep reinforcement learning with agent-based models to reveal adaptive and interpretable strategies. Together, these approaches show how ECF can be systematically analysed and mitigated. The framework offers new theoretical foundations and practical guidance for designing evacuation models, evaluating decision-making, and strengthening policy formulation.

With the increasing urgency to mitigate greenhouse gas (GHG) emissions from urban transportation, this study presents a novel automated planning framework to estimate road traffic carbon emissions. The framework replaces traditional data-intensive and time-consuming planning model development, offering a scalable and transferable approach for coherent comparative analysis across regions worldwide. Utilizing pervasive and open data sources for network and traffic data, we infer origin–destination (O-D) travel demand and integrate it with a macroscopic emission modeling approach. A set of 45 global cities has been modeled to highlight disparities in carbon emissions, congestion, and vehicle kilometers traveled (VKT) levels against varying demand, socioeconomic, and geographic contexts. The findings suggest that while global metropolises such as New York and London exhibit pronounced increases in emissions during demand surges, cities such as Tel Aviv and Ankara demonstrate greater resilience. A comparative analysis with International Energy Agency (IEA) data confirms the validity of our estimates, revealing strong correlations in ranking trends. The findings emphasize the significant impact of congestion and VKT on urban emissions, with some cities experiencing congestion-induced emission increases of up to 300% for a mere 50% demand rise. Given these insights, the study underscores the necessity for tailored policy interventions, including enhanced public transport systems, congestion pricing, and urban mobility strategies to curb emissions effectively. By providing a rapid, scalable modeling approach, this study aids policymakers in developing targeted strategies for sustainable urban mobility and carbon reduction.

The negative social and environmental impacts of car-oriented road designs have led transport policymakers to concentrate on alternative modes of travel in their recent development plans. However, in this paradigm shift, they must deal with the challenge of allocating limited resources to non-car users of transport networks in the most efficient manner. In this study, we address the resource allocation problem to maximise the accessibility of active modes on a multimodal network, accounting for the comfort of car and bus users, from a cost-benefit analysis (CBA) perspective. A bilevel optimisation model is developed in which, at the upper level, a planner moderates the network configuration by making investments in bus frequency and bike lane development to maximise the accessibility of the cycling mode over a pre-determined investment horizon and project life cycle considering a budget constraint containing operating and long-term funds. A combined modal split and traffic assignment (CMSTA) problem is considered at the lower level. The bilevel problem is converted to a single-level non-linear optimisation model by implementing KKT conditions. In a nutshell, our analysis shows that when the project life cycle is less than the planner’s investment horizon, having the option to modify the design throughout the horizon (dynamic designs) leads to higher average accessibility levels compared to static designs where that option does not exist. Furthermore, charging users to use the network regardless of the link type they choose provides cyclists with higher accessibility levels. Finally, difference in value of time among users can negatively impact accessibility of cyclists.

Despite considerable public investment, transport systems commonly result in significant disparities in outcomes across different socioeconomic backgrounds. We propose the Agent-Based Model for Equity-Transport Optimisation and Policy (ABM-ETOP) framework, a novel computational approach that integrates a custom-built agent-based model and Bayesian optimisation specifically designed for the rigorous evaluation of transport policy interventions. In this study, the ABM-ETOP framework is used to investigate the optimal allocation of transport subsidies for the improvements of mode share equity, travel time equity, and overall system efficiency. Analysis using this framework reveals considerable pre-intervention disparities (e.g., low-income: 35.6% public transit, 18.4% car vs. high-income: 14.5% public transit, 58.9% car). Furthermore, applying optimal subsidy levels identified by the framework demonstrates substantial improvements: a 3.0x enhancement in modal equity distribution, an 8.0x improvement in temporal accessibility fairness, and a 1.3x reduction in aggregate system travel burden. Thus, ABM-ETOP offers a powerful tool for evidence-based policy design that addresses persistent equity concerns while supporting system efficiency.

Transportation networks are crucial for social and economic activities but are susceptible to disruptions. Rapid quantification of the impacts of network disruptions can assist in planning recovery efforts. However, gathering timely and comprehensive information for assessing transportation network state is often challenging and not always possible. This study introduces a network assessment strategy to estimate total link capacity reduction and origin-destination (OD) demand matrix (CRDM) for disrupted transportation networks subject to limited information, i.e., link travel time accessible from smartphone-based trajectory data. The CRDM problem can be formulated as a bi-level model, optimizing estimates of externally caused capacity reduction and OD demand matrix in the upper level while solving the user-equilibrium-based traffic assignment in the lower level. The proposed bi-level model with a generalized least squares (GLS) objective (to minimize the discrepancy between observed and estimated travel times) does not yield a unique solution. Therefore, we further employ the maximum entropy principle to develop a maximum entropy-least squares (MELS) model, which has a unique solution. To solve the MELS model, we develop a tailored augmented Lagrangian algorithm and conduct numerical studies on different transportation networks (i.e., a two-link single- OD network, the Sioux-Falls network and a real-world regional transportation network). The proposed approach is able to provide a rapid post-disruption evaluation of the overall link capacity loss in transportation network under limited information, i.e., without OD demand information and with limited information on link travel time.

In this study, we develop an innovative data-driven optimization approach to solve the drone delivery service planning problem with online demand. Drone-based logistics are expected to improve operations by enhancing flexibility and reducing congestion effects induced by last-mile deliveries. With rising digitalization and urbanization, however, logistics service providers are constantly grappling with the challenge of uncertain real-time demand. This study investigates the problem of planning drone delivery service through an urban air traffic network to fulfil online and stochastic demand. Customer requests, if accepted, generate profit and are serviced by individual drone flights as per request origins, destinations and time windows. We cast this stochastic optimization problem as a Markov decision process. We present a novel data-driven optimization approach which generates predictive prescriptions of parameters of a surrogate optimization formulation. Our solution method consists of synthesizing training data via lookahead simulations to train a supervised machine learning model for predicting relative link priority based on the state of the network. This knowledge is then leveraged to selectively create weighted reserve capacity in the network and via a surrogate objective function that controls the trade-off between reserve capacity and profit maximization to maximize the cumulative profit earned. Using numerical experiments based on benchmarking transportation networks, the resulting data-driven optimization policy is shown to outperform a myopic policy. Sensitivity analyses on learning parameters reveal insights into the design of efficient policies for drone delivery service planning with online demand.

As disasters become more frequent and severe, disaster management has increasingly become a global concern, with technological innovation playing a crucial role. The introduction of artificial intelligence (AI) has greatly changed traditional methods of disaster management, opening new avenues for enhancing the efficiency and effectiveness of disaster response. However, despite the transformative impacts of AI in disaster management, there are still some potential issues and challenges that have not been fully explored and addressed. In this review article, we comprehensively review existing research related to AI and disaster management. We selected 5,893 articles from Web of Science and Google Scholar that fit the theme and conducted detailed discussions on the application of different AI technologies in the four stages of disaster management: prevention and mitigation, preparedness, response, and recovery, as well as on disaster technology and ethical issues, through bibliometric analysis and knowledge graph analysis. Based on this, we analysed the core challenges and issues in disaster management where AI has not yet been fully addressed, including pre-disaster and post-disaster resilience enhancement, reinforcement of critical infrastructure, evacuation network planning and design, long-term sustainable recovery after disasters, and related ethical issues in disaster management. We also proposed a future technological research framework. This study provides valuable insights for researchers and practitioners interested in AI and disaster management and clarifies the enormous potential for further research and application of AI in disaster management, potentially driving more progress in this professional community.

The role of transportation system in ensuring equitable access to essential services and promptly recovering it post‐disaster is critical to community resilience. This research introduces a framework aiming to strengthen transportation systems against external shocks, with an emphasis on geographical equity. To evaluate equity and address multiple network design objectives, we develop a two‐level consolidated resilience index that measures network performance and community equity, employing a data‐driven analytic hierarchy process for objective metric weighting, surpassing traditional expert scoring methods. Furthermore, we have implemented an equity‐weighted Shapley value method to prioritize candidate links prior to investment. Finally, we have established a multi‐objective bi‐level program that integrates traffic distribution and travel behavior analysis. Our findings reveal that integrating equity considerations into candidate links selection phase significantly enhances fairness outcomes. The results also underscore the inseparable relationship between pursuing fairness and efficiency. This framework could potentially extend to other transportation systems’ investment strategies during the preparation phase, contributing to broader applications in resilience planning.

Accurate transport demand forecasting can benefit from multimodal data, yet practical challenges arise when different institutions hold separate datasets and cannot share them directly. While institutions may not share data directly, they may share models trained by their data, where such models cannot be used to identify exact information from their datasets. In this context, we propose a Knowledge Adaptation Demand Forecasting (KADF) framework that leverages pre-trained models from one transport mode (source) to forecast demand for another (target), without direct data sharing. The framework captures shared travel patterns across modes through a knowledge adaptation strategy, separating target-mode data into individual and shared components. A pre-trained source model transfers generalized knowledge to improve target-mode predictions. Experimental results on real-world datasets show that KADF outperforms baseline and state-of-the-art models, demonstrating the effectiveness of knowledge transfer without compromising data privacy. This approach supports collaborative forecasting in a decentralized data environment.

This study contributes to sustainable transportation modeling by proposing a user-centric approach to incentivize eco-routing travel behavior. We propose a novel reward credit scheme to provide path-based commuter incentives with the goal of reducing CO2 emissions and the total system travel time. The scheme takes into account multiple classes of commuters in the network that differ by their value of time and their vehicle energy type. Users subscribing to the scheme may earn monetary reward credits which act as incentives to promote sustainable mobility. Two types of reward credits are considered: subscription-and path-based credits. A discrete choice model is embedded within a traffic assignment model to capture the endogenous impact of commuters' scheme adoption onto network congestion effects. We introduce a bilevel optimization formulation to determine optimal non-additive, path-based reward credits and subscription-based reward credits within a predefined budget under traffic equilibrium conditions. In this formulation, the follower problem is a parameterized multi-class user equilibrium traffic assignment problem with non-additive path costs and incorporates a logit choice model for scheme adoption. The leader represent the network regulator whose goal is to maximize social welfare by minimizing the total system travel time and total CO2 emissions. We develop a single-level Karush-Kuhn-Tucker reformulation and propose a customized branch-and-bound algorithm to solve this bilevel optimization problem. Numerical experiments demonstrate the potential of eco-routing incentives to promote sustainable urban mobility and highlight the benefits of combining subscription-and path-based reward credits for traffic congestion management.

Macroscopic link-based flow models are efficient for simulating flow propagation in urban road networks. Existing link-based flow models described traffic states of a link with two state variables of link inflow and outflow and assumed homogeneous traffic states within a whole link. Consequently, the turn-level queue length change within the link cannot be captured accurately, resulting in underrepresented queue spillback. Moreover, a constant link free-flow speed was assumed to formulate models, restricting their applicability in modeling phenomena involving time-varying free-flow speed. This study proposed a new link-based flow model by introducing an additional state variable of link queue inflow and adapting the link outflow to be free-flow speed-dependent. In our model, the vehicle propagation within each link is described by the link inflow, queue inflow, and outflow, which depends on the link free-flow speed changes and signal control. A node model is further defined to capture the presence of potential queue spillback, which estimates the constrained flow propagation between adjacent road segments. Simulation experiments were conducted on a single intersection and on networks to verify the proposed model performance. Results demonstrate the predictive power of the proposed model in simulating traffic proppagations for networks with multiple turning movements and time-varying free-flow speed. Our model outperforms the baseline link-based flow model while preserving the computational tractability property of link-based flow models.

The equity-efficiency tradeoff is a perpetual challenge in public transport planning. There is a strong need to integrate equity considerations into transit planning, while respecting the long-term financial sustainability of the public transport system. We introduce an ‘Equity over Time (EoT)’ multi-period, biobjective, bilevel frequency optimization framework, to integrate fairness metrics into bus allocation, while incorporating practical considerations such as fleet rebalancing costs and transfers. By changing the recipient of benefits or penalties over time, the multi-period allocation perspective allows the model to improve the tradeoff between efficiency and equity. We use Pareto-front analysis to demonstrate the improved tradeoffs between efficiency and equity of the EoT approach, then show through numerical experiments that the EoT framework is able to achieve better or equal solutions than its single period counterpart in all instances. We propose a customized matheuristic combining pattern generation and scheduling to solve larger instances. Numerical experiments show that the heuristic is on average 99.03