ABSTRACT. One major challenge faced by individuals with reduced mobility when travelling with their own vehicles to new areas is the lack of information about where to find disabled parking spots places and their usage constraints. While some large cities offer this information through official platforms such as the city hall webpage or mobile application, this is not a universal standard and often this information is not available for small towns. This study proposes a solution to alleviate these issues through the detection of accessible parking spots using artificial intelligence on satellite imagery. The aim is to develop an up-to-date database that can be used by interactive maps and mobile applications to enable users to locate accessible parking spaces and access detailed information about their characteristics and usage regulations.

ABSTRACT. India's increasing road accident statistics highlight the significant exposure faced by motorized two-wheelers (MTWs), which constitute most urban automotive traffic. This study examines MTW behaviour on a roadway in Surat, where MTWs make up over 90% of the traffic, effectively creating a naturally segregated lane for these vehicles. Data was collected for one hour using a UAV along a 180-m access-controlled stretch. The trajectory data was extracted and smoothened to remove outliers from the dataset. Key micro-level parameters, such as lateral placement and following behaviour, were analysed across midblock, merging, and intersection sections. Additionally, motion kinematics was explored across sections. The results indicated that MTWs predominantly occupied the middle portion of the road, regardless of section type. The mean speed was highest at midblock (130%), followed by merging sections (120%), compared to the mean speed at intersections. Due to complex weaving patterns, longitudinal and lateral accelerations varied significantly at the merging and intersection sections. Leader-follower interactions were analysed using hysteresis plots, and parameters related to the following behaviour were estimated based on Wiedemann 74 conditions. Further analysis revealed that the following time decreased from the midblock (2.29 s) to intersections (1.71 s), indicating denser interactions in complex areas. The Time to Collision (TTC) analysis showed relatively higher safety at midblock (mean TTC 2.863 s) and slightly elevated collision risks at intersections (mean TTC 2.668 s). The results obtained from this study can be useful for calibrating simulation models and testing policies related to exclusive MTW lanes.

ABSTRACT. Origin-Destination (OD) matrices are essential inputs for various traffic planning and management situations. A time-sliced OD matrix (usually hourly), which reflects traffic patterns over different times of the day, is essential for understanding peak-hour traffic and for managing congestion effectively. In this paper, we propose a method that uses mobile network data, link counts, and vehicle probe data, within a data-driven network assignment (DDNA) and OD estimation process, to derive a time-sliced OD matrix from an average day OD matrix. The method is evaluated on empirical data for Norrköping, Sweden, demonstrating that the method provides link flow estimations consistent with link flow observations for both training and test sets while maintaining structure from the demand model as well as mobile network data.

ABSTRACT. There is no one-size-fits-all answer to anomaly detection problems in time series data. The crude methods are often insufficient and sophisticated methods are often too particular for most purposes. It thus happens that a combination of simpler methods and tools from other areas does the job best. In this regard, we have not encountered any work that uses modularity-based community detection approach for anomaly detection problems. This note aims to present how this can be done with little effort and surprisingly good results, demonstrated on our data from studying lateral distances between cyclists and overtaking vehicles on rural roads.

ABSTRACT. In urban context, the use of motorized vehicles tends to be prioritized over other transport categories. This generates an injustice in the distribution of rights, possibilities and mobility freedom. Therefore, it is important to analyse and study disparities among transport users to be able to promote a fair distribution of road spaces with the aim of developing transport systems that offers equal access to public space and opportunities. In this study, we proposed an equity indicator for urban trips with the aim of evaluating spatial injustice in the area under consideration. It is a revisited Gini index which aims to show the injustice of the distribution of road spaces among road users. This methodology was applied to a case study.

ABSTRACT. This study presents a methodology for assessing transit stop accessibility and optimizing stop placement using a synthetic population and network-based catchment areas. Traditional approaches often rely on static buffer zones, which fail to capture real-world variations in pedestrian access and transit connectivity. Our method dynamically assigns transit stop watersheds based on optimal routes to key destinations, accounting for diverse travel behaviors and multimodal connections. A synthetic population is integrated to refine demographic profiling within each catchment area, enabling a more accurate estimation of transit usage potential. By combining demographic data and network-based walking accessibility, we compute a continuous probability function for transit stop use, improving upon rigid distance thresholds. Additionally, our approach allows for iterative stop placement evaluation, recalculating catchment areas and usage likelihood scores to optimize transit network design. This framework provides a data-driven, user-friendly tool for planners to enhance transit accessibility and equitable service provision.

ABSTRACT. In a smart city, it is essential to plan and design transport systems to reduce the environmental, social and economic impacts. Rethinking urban mobility, starting with the predominant use of Public Transport (PT) is essential, especially in the light of current policies aimed at discouraging the use of private cars, thus reducing vehicle congestion, pollution and land consumption. As most daily trips are related to commuting to workplaces, providing efficient connections to PT systems is paramount in promoting a modal shift. Improving the access and connectivity between residential and industrial areas also helps to strengthen the local productive fabric. The aim of this work is to develop a data-driven methodology, where simulation results serve as ground truth, to assess the accessibility to the workplaces. The methodology will be applied to the case study of the metropolitan city of Catania (Italy), which has one of the highest motorisation rates in Europe. Accessibility indicators are analysed in relation to the current PT supply and the assessment of inequalities will be carried out between different parts of the study area. Finally, by identifying areas of low accessibility, intervention strategies will be proposed and simulated, and their impact on the indicators will be directly assessed. This study is proposed as a key tool for administration to help them identify areas for intervention and evaluate the impact of new infrastructure on improving accessibility.

ABSTRACT. The implementation of 30 km/h (20 mph) speed limits has gained global attention as a strategy to enhance road safety and urban livability. While such interventions have been widely adopted across Europe and North America, Greece is at early stages of exploring their potential benefits. This study examines public perceptions of citywide 30 km/h speed limits in Greece, assessing levels of support, anticipated safety improvements, and broader implications for urban quality of life. Data was collected through an online survey targeting urban and rural residents, yielding 718 valid responses. Statistical analysis, including ordered probit modeling, identified key factors influencing public acceptance. Findings reveal that over 50% of respondents support the intervention, with many expecting reductions in air and noise pollution, as well as improved walkability. Younger individuals, public transport users, and those in single-person households were more likely to support the speed limit, whereas daily car users exhibited opposition and skepticism towards a 30 km/h speed limit. The results suggest that while road safety remains a key motivator, respondents also acknowledge the possibility of benefits in urban livability. These insights can potentially inform policymakers in designing effective speed limit policies that balance mobility needs with sustainability goals.

ABSTRACT. Last-mile delivery is significantly affected by and contributes to the issue of urban congestion. In times when e-commerce activities are on the rise, traditional delivery cars and vans are increasingly problematic. While the delivery staff has difficulties finding places to stop, they often have to occupy dedicated walking and cycling paths, often with the engine idling and contributing to emissions. Cargo bikes are a viable, more sustainable alternative in many urban regions. However, detailed information on the impact of this new form of delivery service is lacking. So far, no detailed analyses of route choice with delivery cargo bikes have been conducted. This study analyzed the route choice of commercially used cargo bikes in Munich. While chosen routes were found to be significantly longer than the shortest paths between delivery stops, the availability of bike-friendly infrastructure was impacting the delivery process. This study provides insights into real-life delivery processes that are often difficult to analyze, serves to start evaluating sustainable delivery scenarios, and helps to plan appropriate infrastructure.

ABSTRACT. Parking and walking are important but frequently disregarded features of last-mile delivery. This study quantifies the hidden costs associated with these activities by utilizing inverse optimization (IO) to estimate cost parameters based on observed delivery routes. By modeling drivers' parking decisions as part of a recurring decision-making process, we can identify trade-offs between parking availability and walking distance. Using data from two different types of logistics suppliers operating in Seattle (U.S.), we provide insights into the operational trade-offs in a real-world setting.

ABSTRACT. This paper presents a bi-objective mixed-integer linear programming (MILP) optimization model to enhance last-mile delivery efficiency within an integrated freight-passenger network. The model aims to minimize both delivery costs and delivery time by incorporating pedestrian participation in the delivery process. By using capacity of pedestrian, this approach optimizes urban logistics while reducing congestion and environmental impact. A small-scale case study is conducted to validate the feasibility of the proposed model. The findings highlight the importance of policy interventions, infrastructure improvements, and digital coordination tools in ensuring effective implementation.

ABSTRACT. INTRODUCTION “Last mile delivery” is a short geographical segment of the delivery of products to individual customers, usually located in dense, downtown areas of the cities. Due to the increase of internet sales, which requires supplies to individual customers, “las time deliveries” constitute the critical component of e-commerce transactions.

Despite the fact that urban freight traffic has a relatively low share in the total traffic within the metropolitan areas (3%-5%), it contributes dis-proportionally to air pollution, noise, congestion, traffic accidents and reduced accessibility of urban areas. "Last mile" deliveries result in a growing flow of loads in the centers of the cities, very high costs (up to 40% of logistics costs) and deterioration of natural environment and the quality of life of residents. Therefore, technologically-oriented logistics solutions are needed to eliminate or reduce these problems. The considered problem falls within the concept of city logistics/ urban freight logistics.

PROBLEM DEFINITION AND FORMULATION The paper is focused on the analysis and development of certain modern, innovative transportation/ logistics solutions for a last mile delivery system of goods in Poznan City in Poland. The project is a real life case study analysis of an e-commerce business servicing roughly 400 customers a day in a highly populated, historical segment of the city. The delivered products are cosmetics (perfumes, soaps, deodorants, air refreshers, shower gels) and most shipments are small in size and light in weight, up to 3 kg.

The analysis refers to the supply chain segment that covers deliveries between the logistics hub and final consumers. The area being served has a radius of roughly 3 km and the customers are randomly located in this area.

In our paper we propose 9 alternative, technologically advanced “last mile” delivery concepts and we investigate their feasibility. These concepts are based on: drones, autonomous vehicles, cargo bikes, crowd shipping and freight car pooling. Our solutions belong to innovative contributions classified as: innovative vehicles, collaborative and cooperative urban logistics, optimization of transport management and routing. The proposed logistics solutions are simulated and quantitatively tested. They are evaluated from different perspectives and their strengths and weaknesses are analyzed.

The analysis is performed for the 3-rd party logistics (3 PL) being a freight logistics operator servicing different customers at the cosmetics market, including the e-commerce seller (denominated Company N). The decision maker is the Top Management of the 3PL.

The decision problem is formulated as a multiple criteria ranking problem in which 9 variants are evaluated by a consistent family of 7 criteria. The variants constitute the developed last mile delivery solutions such as: freight car pooling, crowd shipping, cargo-bikes, autonomous vehicles, drones, combined crowd shipping and autonomous vehicles, combined drones and carpooling, combined cargo-bikes and autonomous vehicles and combined crowd shipping and cargo-bikes are considered. As indicated above they are based on either a single transportation mode (mean) or a multimodal solution (e.g. cargo-bikes and autonomous vehicles) or a single transportation mode combined with a certain transportation policy (e.g carpooling or crowdshipping). The following criteria are used to evaluate the variants: Cost, Delivery Time, Safety, Flexibility, Reliability, Environmental Friendliness, and Timeliness. Based on their evaluation the considered variants – last mile delivery solutions are ranked from the best to the worst. Their ranking is generated in the computational phase carried out with the application of selected Multiple Criteria Decision Making/ Aiding (MCDM/A) methods, including ELECTRE III/IV and AHP. The results of computational experiments are demonstrated. SOLUTION PROCEDURE - COMPUTATIONAL EXPERIMENTS In the computational phase the following steps have been performed:

• Modeling of the DM’s preferences (different for the ELECTRE III/IV and AHP methods), taking into account two components of the DM’s preference model: weights/ importance of criteria and sensitivity of the DM towards the differentiation between variants on specific criteria. • Generation of the final ranking based on the computation of the utility of each variant (in the AHP method) and the outranking relation (in ELECTRE III/IV method). • Analysis of generated results. Comparison of the rankings produced by AHP and ELECTRE III/IV method. Final recommendations. Sensitivity analysis – conclusions about the stability of the rankings. CONCLUSIONS The problem of the design and evaluation of “Last Mile Delivery” System has a multiple criteria character and can be formulated as a multiple criteria ranking problem. It can be solved with the application of selected MCDM/A ranking methods.

In the analyzed case, the most promising variants for the “Last Mile Delivery” system are: Drones, Crowd shipping + Cargo Bikes and Drones + Freight Carpooling. Drones are ranked first in the generated ranking due to low overall cost of delivery, excellent delivery time and timeliness as well as environmental friendliness. The results of the simulation show that drones are not affected by congestion, however, they have not been legalized in many countries, yet. Additionally, the authors of the paper want to indicate that “Crowd Shipping + Cargo bikes” is also a highly ranked and thus strongly recommended solution.

In the final version of the paper further computational experiments will be carried out and their results will be demonstrated. The authors will present the comparison between the rankings generated by ELECTRE III/IV, AHP and possibly other methods.

ABSTRACT. Digitalization has great potential for making air transport more efficient and sustainable, e.g. by holistic optimization or improved usage of resources. This applies particularly to processes within an airline. In this context, an automated decision-making tool with high degree of integration and compatibility to different business strategies could enable airlines, politicians and research institutions to assess, for example, how new fleet types can be integrated into an airline's fleet or how flight demand can be met with limited resources. For this purpose, a digital airline twin (DAT) is currently being created by the German Aerospace Center (DLR). Given airline-specific input data the DAT will automatically execute several integrated and/or sequential optimization procedures. Those procedures represent all relevant airline planning processes starting with the question of which airports to include in the network and ending with the actual flight execution. During execution, the results of the long-term planning processes are used as inputs for subsequent planning steps. Additional sources of input are public databases. Unfortunately, publicly accessible data is often neither complete nor flawless. Therefore, there is a need for automatic procedures for dealing with these imperfect data sets. This study presents, compares and evaluates two algorithms to repair schedules in order to find the most promising input data supply procedure for the fleet assignment of the DAT.

ABSTRACT. Forecasting travel demand in aviation is a critical phase in airline management since it affects other management decisions, such as network design and fleet assignment. The aim of this study is to propose and develop an optimisation framework for estimating the Origin-Destination (O-D) demand matrix of a given air transport network and to demonstrate the benefits of this novel framework through applying it to Australia’s domestic aviation network/system. Our proposed O-D estimation problem is novel because it combines different data sources, including population data, operational data from airline companies, and count data within a single optimisation framework. We take the O-D estimation framework that is typically used in road transport as starting point and customise/tailor it to an aviation context. In our optimisation framework we apply a gravity model to determine a prior O-D matrix using population data. We then embed a traffic assignment based on a network consisting of observed operational flight routes and distance data obtained from Australian airlines, and calibrate predicted O-D flows to inbound and outbound passenger counts at airports.

ABSTRACT. Aligning with the objectives of the Paris Agreement, airlines are committed to achieving net-zero carbon emissions by 2050. In order to help achieve this goal, this paper addresses the creation of flight schedules across airlines under the premise that airlines share their aircraft fleets and that passenger flows are bundled across airlines. We propose a mixed-integer linear programming model to determine the departure times of all flights within a given network, considering desired frequencies by airport pairs and aircraft types in the network. The objective is to minimize the number of aircraft needed to perform the schedule. As the model is hard to solve for flight networks of realistic size, we discuss heuristic approaches and investigate their performance using synthetically generated instances. We solve the model for scenarios with and without aircraft sharing, in order to estimate the maximum potential of aircraft sharing and passenger bundling.

ABSTRACT. Extended aircraft ground times between two flights are widely recognised to have the potential to increase costs, to generate network inefficiencies, and to diminish the attractiveness of both airports and airlines. As a result, minimizing ground times, and thereby maximising aircraft utilisation, has become an objective within airline operation planning. Furthermore, incorporating operational data such as ground times into scenario-based analyses has become relevant for identifying suitable design parameters in future aircraft concepts, such as the design of liquid hydrogen or electric aircraft.This study aims to explore the dynamics of ground times in relation to future aircraft design and airline networks. A dual approach is employed, combining both descriptive and analytical methods to analyse ground times across factors as aircraft types, regions, airports, and airlines.An openly available ADS-B data set to extract ground time information from flight operations between 2022 and 2024 is utilised. Additionally, flight schedules and passenger demand data are incorporated to conduct a network analysis from airport and network level, evaluating how ground times affect overall airline performance and connectivity.

The analysis of ground times across aircraft categories highlights the need for range- and operation-specific insulation strategies in future hydrogen aircraft to optimise hydrogen storage. Furthermore, the regression-based network analysis reveals that reduced ground times are associated with greater network robustness, higher passenger numbers, increased connectivity, enhanced centrality, and improved aircraft utilisation at the network level, with variations observed across regions and business models. Regional differences are more evident at the airport level, while business model variations became more apparent at the network level.

ABSTRACT. Intelligent Intersection Management (IIM) approaches aim to optimize urban traffic flow but face inherent limitations depending on their operational mode. Sequential approaches suffer from inefficiencies under high traffic demand, rigid cyclic operations, and poor scalability. Parallel approaches struggle with turning conflicts, unbalanced traffic flows, and strict lane discipline requirements. Synchronous approaches depend heavily on precise timing, reducing their flexibility. To address these challenges, we propose the Software-Defined Intelligent Intersections (SDI^2) framework, an SDN-based IIM solution that intelligently orchestrates intersection management through centralized control. The SDI^2 framework dynamically switches between sequential, parallel, and synchronous modes based on real-time traffic patterns, effectively accommodating both individual vehicles and groups. This adaptive strategy optimizes traffic flow by reducing reaction times to green signals, minimizing unnecessary stopped delays by up to 89.5%, and significantly reducing energy waste by 63.3% and associated emissions by 76.9% compared to conventional approaches.

ABSTRACT. This research addresses the challenge of OD matrix calibration in transport modeling, where an initial OD matrix is available alongside traffic data collected from sensors on specific road segments. We propose a refined Simulation-Optimization algorithm that improves the accuracy of transport simulations while preserving the consistency of historical demand patterns. A key aspect of the calibration process is determining the calibration parameter α, which governs the trade-off between the historical OD matrix and the observed traffic count data. The approach identifies the value of α that better maintains both the integrity of historical OD matrices and the accuracy of simulated traffic flows. This approach, employing the PTV Visum simulation tool and heuristic techniques, is tested on a real-world urban network and is designed to be robust, scalable, and applicable to a variety of transport networks.

ABSTRACT. Fare evasion is a pressing issue in public transport networks, impacting the financial sustainability of Transit Agencies (TAs) and Public Transport Companies (PTCs). While prior studies have largely focused on fare evasion frequency, research on severity—the financial and operational impact of detected fare evasion cases—remains limited according to authors knowledge. This study addresses this gap by specifying, calibrating, and validating two predictive models for fare evasion severity using real-world inspection data from a mid-sized Italian PTC. Two modelling approaches are employed: an Econometric Approach (EA) using a Generalised Linear Regression Model (GLRM) based on logistic regression and a Machine Learning Approach (MLA) leveraging an Artificial Neural Network Model (ANNM). The models are quantitatively assessed and compared using Confusion Matrices (CMs) and Performance Metrics (PMs), including True Positive Rate (TPR, Sensitivity), True Negative Rate (TNR, Specificity), and Accuracy (ACC). Results indicate that the ANNM outperforms the GLRM, demonstrating higher predictive accuracy and a stronger ability to detect high-severity fare evasion cases while reducing false negatives. The ANNM improves sensitivity, making it more effective in identifying severe offenders. However, the GLRM remains valuable for policy analysis, offering greater interpretability to help transit agencies understand the significance of specific factors influencing fare evasion severity. These findings provide critical insights for optimising fare inspection policies and enforcement resource allocation. Future research should explore other econometric or machine learning models, as well as hybrid modelling approaches that could balance predictive accuracy and interpretability.

ABSTRACT. Bike-sharing systems have emerged as a crucial component of urban mobility infrastructure across major metropolitan areas. While these systems provide significant benefits for sustainable transportation, the disparate nature of monitoring platforms poses challenges for both users and transportation planners. In response to this, this paper presents a unified real-time dashboard that integrates multiple bike-sharing systems across major North American cities. Notably, selected bike-share systems generate 86 percent of total trips in 2023. These cities include Boston (Bluebikes), Washington, D.C. (Capital Bikeshare), Chicago (Divvy), New York City (Citi Bike), Montreal (BIXI), San Francisco (Bay Wheels), and Toronto (Bike Share).

Furthermore, our dashboard implements a scalable architecture that uses the General Bikeshare Feed Specification (GBFS) to aggregate and visualize real-time data from multiple bike-sharing networks through a single interface. Moreover, the system employs a configuration-based approach that facilitates the easy integration of additional cities and bike-sharing systems. Built on a Python-based stack utilizing Panel for the user interface and Folium for interactive mapping, the dashboard provides comprehensive visualization of system status, including conventional and electric bicycle availability, docking station capacity, and operational status across all integrated networks.

ABSTRACT. Railway punctuality remains a critical measure of service quality and operational efficiency. Traditional performance metrics such as on-time performance and delay increments inform about punctuality goals and guide maintenance planning but often neglect the passenger experience due to the lack/limited availability of data. This study integrates forecasted ridership data with delay evaluation models to assess passenger delays and their socio-economic impacts. By combining passenger-centric delay contributions with the Swedish framework for socio-economic evaluations, we enable a more informed prioritization of maintenance interventions. A case study on the Southern Main Line in Sweden illustrates the methodology’s potential to optimize maintenance planning, highlighting its relevance for achieving data-driven improvements in train service reliability.

ABSTRACT. Public transport networks in European cities integrate both schedule-based and frequency-based services. However, existing optimization methodologies typically address these services separately, often leading to suboptimal results. To enhance service efficiency and passenger experience, this study advocates for a joint optimization approach. Building on the work of Jiang et al. (2022), which introduced an integrated optimization framework that balances passenger costs and operational cost by adjusting timetables for mixed-service (schedule-based and frequency-based) public transport networks, this paper extends their framework by incorporating (i) additional capacity constraints to mitigate overcrowding and (ii) heterogeneity considerations to improve accessibility for elderly individuals and people with disabilities. These enhancements contribute to advancing research in equitable and efficient public transport network optimization.

ABSTRACT. The rapid evolution of urban mobility presents significant challenges for transport infrastructure development as micromobility options, like e-scooters and cargo bikes, become increasingly prevalent. A comprehensive analysis of technical guidelines, regulatory frameworks, and practices reveals how current bicycle-centric infrastructure approaches may face limitations in supporting emerging mobility patterns and technologies. The findings contribute to understanding infrastructure requirements for sustainable urban mobility development by examining vehicle diversity, operational patterns, and integration needs.

ABSTRACT. Meeting public transportation quality standards is essential for creating sustainable and livable urban environments, yet this goal is often challenged by data access and integration needs, as well as the system's complexity and heterogeneity [1, 2, 3]. Mobility patterns in public transport are continuously evolving due to factors such as demographic shifts, the integration of new transport modes (e.g., micro-mobility), and changes in commuting behavior driven by remote work [4,5]. In addition to these evolving and unpredictable factors, the system is shaped by the constant interaction of diverse actors, including varying user profiles, multimodal services, and operational challenges. In the age of intelligent transport systems, information systems play a pivotal role in monitoring the dynamics of the system and, in turn, supporting the measurement of key indicators that reflect the quality standards of public transport [1, 2]. Maintaining real-time tracking of these quality indicators is not only vital for assessing service levels but also for enabling the dynamic planning of efficient, equitable, and multimodal public transport systems.

The real-time tracking of public transport quality faces several challenges that have yet to be fully addressed in the research field. One major obstacle is the integration of heterogeneous data sources, including public transport service data, real-time transit data, and spatial indicators such as land use and demographic information, among others [6]. Another challenge lies in defining comprehensive key indicators that account for both the spatial and temporal variability of the service quality while further ensuring their quantitative, objetive and traceable nature. These indicators must align with state-of-the-art practices and European directives, which establish criteria for evaluating public transport based on factors such as availability, accessibility, time compliance, mode dependability, multimodality, among others [7,8]. However, there is still a need to develop tailored metrics that focus on the entire (end-to-end) journey, considering its variability along time (e.g., vulnerable periods along the day and week) and space (e.g., transfer connectivity). Another challenge is the need for an analytical system that evaluates the suitability of services for specific user groups (e.g., elderly passengers) while giving information about fairness in service delivery across the entire network and urban area. The literature has provided advances in this last concern, e.g., by tackling the problem through the spatial distribution of accessibility indicators, enabling cross-sectional analysis of city zones [10, 11]. However, temporal factors (e.g. availability of services throughout the day) may have a significant social impact, especially if user needs are not met during certain periods. Finally, it is of utmost importance that the information system offers statistical guarantees and captures seasonal trends and emerging patterns for each available service, accompanied by clear and informative visualizations.

At the light of these challenges, the research reported in this paper is centered on the proposal of an advanced analytical decision support system for the comprehensive assessment of public transport service quality using smartcard data and relevant sources of situational context. The system aims to provide actionable insights for each journey by offering: (i) a tailored multi-dimensional metric system that reflects the quality of service (accessibility, availability, cross-modal network/multimodality, among other metrics); (ii) context-sensitive analytics that captures temporal trends and seasonal patterns, of different spatial locations/ land-use indicators; (iii) capabilities to evaluate service suitability across key domains such as efficiency, equity, and multimodality; and (iv) a ranking system that classifies the spatiotemporal features of an end-to-end journey and compares its quality against comparable journeys.

The proposed analytical decision support system (ADS2) is based on the following structured approach: ● Data collection and integration: consolidation of real-time transit data, public transport information (including user profiles, smart card data, and timetable scheduling), and spatial indicators such as land use and demographic data. Data preprocessing methods, including imputation of missing data, spatial database queries, and constraints to ensure the completeness, spatial and temporal alignment of data sources. ● Definition of quality metrics: using the previous data, tailored metrics are formulated and computed to evaluate the quality of services along journeys, covering dimensions such as spatial and temporal accessibility (e.g., walking needs), service availability, reliability, and multimodality, among other relevant factors. ● Machine learning and statistical modelling: representation of end-to-end journeys, with a focus on their spatiotemporal features and time-varying multi-dimensional indicators of service and quality, are proposed using multivariate time series data structures. Embeddings of a journey learned through autoencoders, are given to descriptive and predictive models for detecting anomalies and identifying temporal patterns. Finally, clustering methods are further applied to group journeys with similar dynamic behavior (e.g., journeys with comparable spatiotemporal profile, demand, or user group characteristics, such as senior passengers). ● Statistical analysis: For each time series cluster, statistical models decompose the series to identify trend variations and seasonal patterns. Additionally, descriptive statistics are computed to capture other cluster characteristics, such as transfer variability, multimodality, user demographics, or land use indicators. ● Journey quality ranking system: Within each cluster, journeys are ranked according to the quality metrics to ensure fair and context-aware assessments. The underlying principle is that journeys should only be compared with those sharing similar dynamics. The ranking method evaluates the position of a journey based on the metrics inserted in the system, using both individual and integrative views of quality. ● Dashboard capabilities: To support decision-making, an intuitive user interface is provided, featuring: (i) the ability to select the origin and destination of a journey, (ii) visualizations of key statistics related to the journey (e.g., temporal patterns of the metrics, anomalous behaviour, user profile distribution), and (iii) the ranking position of the journey compared to others with similar dynamics. The case study in this research focuses on Lisbon's public transport system, a multimodal network that integrates subway and bus services, used for the testing and validation of the ADS2. The proposed tool aims to support effective transport planning – operational, tactical and strategic. However, it is designed for the easy adaptation to other urban contexts and transport networks, including those outside the scope of public transportation, such as private or shared multimodal systems. To ensure broad applicability and support continuous improvements, the code and dashboard for the underlying system will be made open access. This open-source approach enables operators to modify and extend the tool to suit different service models, incorporate new metrics, and adapt to changing operational requirements.

[1] Cats, O. (2024). Identifying human mobility patterns using smart card data. Transport Reviews, 44(1), 213-243. [2] Kuo, Y. H., Leung, J. M., & Yan, Y. (2023). Public transport for smart cities: Recent innovations and future challenges. European Journal of Operational Research, 306(3), 1001-1026. [3] de Oña, R., López, G., de los Rios, F. J. D., & de Oña, J. (2014). Cluster analysis for diminishing heterogeneous opinions of service quality public transport passengers. Procedia-Social and Behavioral Sciences, 162, 459-466. [4] Christidis, P., Christodoulou, A., Navajas-Cawood, E., & Ciuffo, B. (2021). The post-pandemic recovery of transport activity: Emerging mobility patterns and repercussions on future evolution. Sustainability, 13(11), 6359. [5] Reck, D. J., Martin, H., & Axhausen, K. W. (2022). Mode choice, substitution patterns and environmental impacts of shared and personal micro-mobility. Transportation Research Part D: Transport and Environment, 102, 103134. [6] Hussain, E., Bhaskar, A., & Chung, E. (2021). Transit OD matrix estimation using smartcard data: Recent developments and future research challenges. Transportation Research Part C: Emerging Technologies, 125, 103044. [7] EN 13816:2002. Transportation - Logistics and Services - Public Passenger Transport - Service Quality Definition, Targeting and Measurement. European Committee for Standardization. [8] Ngoc, A. M., Hung, K. V., & Tuan, V. A. (2017). Towards the development of quality standards for public transport service in developing countries: analysis of public transport users’ behavior. Transportation Research Procedia, 25, 4560-4579. [10] Arbex, R., & Cunha, C. B. (2020). Estimating the influence of crowding and travel time variability on accessibility to jobs in a large public transport network using smart card big data. Journal of Transport Geography, 85, 102671. [11] Cui, M., & Levinson, D. (2018). Full cost accessibility. Journal of Transport and Land Use, 11(1), 661-679.

ABSTRACT. Barge transportation is a sustainable and cost-effective alternative to road transport, offering a viable solution to reduce congestion and lower emissions. However, the increasing frequency and severity of drought seasons, driven by climate change, threaten its reliability and efficiency. Since water levels play a critical role in barge operations, their varying conditions directly impact transportation capacity. The ability of a barge to carry cargo depends not only on the physical characteristics of the vessel but also on the available water levels along its route. When water levels decrease, draught restrictions limit the amount of cargo that can be safely loaded and transported, creating operational challenges that make it difficult for carriers to efficiently meet shipper demands while maintaining profitability. This study presents a modeling framework for tactical planning in consolidation-based barge transportation. By explicitly examining the relationship between water levels and barge load capacity, with a focus on barge dimensions, the framework assesses their impact on transportation efficiency and profitability. The proposed model establishes a tactical plan that optimizes resource utilization and maximizes expected carrier revenue while incorporating predicted water levels to ensure efficient demand fulfillment. Through extensive computational experiments using commercial software, we analyze how varying water levels affect carrier profitability, system efficiency, customer satisfaction, and network structure, providing key insights into the operational challenges of barge transportation.

ABSTRACT. Container terminals are of pivotal importance to global trade, as they act as a bridge between maritime and land transport. However, inefficiencies in operations, such as long waiting times and high emissions, continue to challenge the industry. Current practices, including first-come-first-served (FCFS) berth allocation, often result in ships arriving too early and idling at anchorage, leading to increased fuel consumption and negative environmental impacts. Just-in-Time (JIT) strategies have been identified as a potentially effective approach to address these issues, by aligning ship arrivals with berth availability, thus optimising speed and reducing emissions. In this work, we present a simulation-based decision-support tool to evaluate JIT strategies in container terminal operations. By analysing scenarios involving speed optimisation and resource investments, the tool provides insights into key performance metrics, including waiting times, emissions, and resource utilisation. A case study designed around a large Portuguese seaport was used to validate the approach and lead to significant reductions in emissions and operational inefficiencies. These findings highlight the potential of JIT operations to enhance sustainability and efficiency in the maritime sector.

ABSTRACT. The efficient scheduling of vehicles within container terminals is a critical factor in optimizing berth operations and enhancing overall terminal productivity. This paper addresses the Container Terminal Vehicle Scheduling Problem (CTVSP) with a focus on finally integrating quay crane operations, transportation means (trailers and reach stackers), and yard operations into a comprehensive decision-support system. As a fundamental step towards this end, we propose a novel Mixed Integer Linear Programming (MILP) model that minimizes vessel turnaround time, reduces operational costs, and synchronizes the interdependencies among terminal resources. The effectiveness of the model will be analyzed considering realistic data for an Italian terminal container port.

ABSTRACT. Seaborne oil trade is the lifeblood of the world economy. At the other end of the spectrum, economic forces act as major drivers of the demand for tanker ships. The unpredictability of macroeconomic conditions poses considerable challenges to tanker operators. In this context, it is essential that the economic environment be closely monitored when making tanker chartering and investment decisions. Given that not all economic factors are equally impactful, this study proposes a new composite indicator that is designed to capture the aggregate impact of the macroeconomic forces on tanker freight rates. The selection of the most critical sub-indicators from a comprehensive set of representative macroeconomic variables is accomplished using stepwise and lasso regression models. Next, a linear programming model is developed for the aggregation, weighting, and estimation of the composite indicator’s values. Finally, the indicator is validated and tested for its causal relationship with the BDTI values. Overall, the results indicate that there is a statistically significant causal relationship between the proposed indicator and the BDTI.

ABSTRACT. Through the proliferation of modern traffic technologies like autonomous driving and connected vehicles, the need for reliable and precise positioning of vehicles has arisen. The default solution for locating a vehicle is through the use of Global Navigation Satellite Systems (GNSS). GNSS can under certain circumstances provide imprecise positional data. In this paper we will present a system to refine GNSS positional data using Floating Car Observer (FCO) environment data. An FCO is a vehicle that can, through the use of its sensors, detect other vehicles in its environment. By matching all detected vehicles onto known street lanes, we improve the provided GNSS positional data.

ABSTRACT. The fight against air pollution is one of the main challenges faced by urban areas. The United Nations Sustainable Development Goal (SDG) number 11, “Sustainable Cities and Communities,” considers that by 2030, it is essential to reduce "the adverse per capita environmental impact of cities, including by paying special attention to air quality and municipal and other waste management" (United Nations, 2020). According to the United Nations (2024), an estimated 4.2 million people die each year due to environmental pollution. Despite a global reduction in pollutants such as PM2.5, air quality, especially in urban areas, continues to fall short of World Health Organization (WHO) recommendations. The WHO's global annual mean concentration should move from a population-weighted average of 35.7 µg/m3 to the recommended value of 5 µg/m3 (World Health Organization, 2021). Transportation is a major contributor to pollution, responsible for significant portions of emissions such as 55% of NOx, 21% of CO, 20% of PM2.5, and 13% of PM10 in Europe (European Environment Agency, 2019). These numbers tend to be higher in urban areas due to denser populations and higher concentrations of residential and industrial activities, leading to greater congestion and a physical environment that limits the dispersion of pollutants (Health Effects Institute, 2022). Sustainable mobility policies, therefore, play a crucial role in reducing urban pollution. Among these policies, Low Emission Zones (LEZs) have become one of the most widely implemented tools aimed at reducing the presence of highly polluting vehicles, particularly in the most congested urban areas. However, the effects of LEZs on urban areas can vary, especially in terms of mobility patterns and modal distribution as a result of their implementation. The literature surrounding LEZs has been focused on multiple dimensions, including health, mobility, air quality, and public acceptance. Several studies have shown that LEZs can effectively reduce pollution levels and promote cleaner vehicle technologies, but they also highlight the challenges associated with the behavioural changes required for their success. For example, studies by Morton et al. (2021) and Oltra et al. (2021) have explored how citizens' perceptions, attitudes, and socio-demographic characteristics impact their acceptance of such policies. These findings suggest that factors such as positive emotions toward the policy, the perceived fairness of the implementation process, and the policy's effectiveness in improving air quality are crucial in determining public support. On the mobility side, research has demonstrated that LEZs can influence modal shifts, particularly from private cars to public transport and active modes. For instance, Charleux (2014) found that the implementation of a LEZ in Grenoble could significantly promote the use of active transport modes and public transit. Similarly, studies in Madrid (Tarriño-Ortiz et al., 2022; Moral-Carcedo, 2022) have observed modal shifts towards public transport and active transportation modes, with some evidence of a reduction in traffic congestion within LEZ boundaries, although the latter can sometimes result in increased traffic in surrounding areas (Tarriño-Ortiz., 2022). These findings point to the importance of understanding how various factors, such as socio-economic status and prior mobility patterns, influence the likelihood of a shift in travel behaviour. Building on this state of the art, the goal of this study is to develop policies that enhance the effectiveness of LEZs by better understanding the factors influencing citizens' willingness to change their travel behaviour and transportation modes in response to these zones. Through the use of Structural Equation Modelling (SEM), this research aims to provide a deeper insight into the socio-economic, psychological, and attitudinal factors that shape individuals' decisions to adapt to LEZs, contributing to the design of more effective and targeted policies that can lead to improved air quality and more sustainable urban mobility.

ABSTRACT. Within a city, several networks coexist to support their activities and operations, each playing a crucial role in ensuring the distribution of goods and services throughout the territory. Their structure must provide a high level of connectivity, avoiding dependence on individual elements: in this way, each component contributes to the stability of the system, with no single element being essential. Among these, the transport network plays a crucial role, enabling the mobility of people and goods, thus promoting access to services, economic growth and sustainability. In modern cities, such networks are designed according to the principle of redundancy, but in established urban centres that have evolved over time, they do not always comply with this framework. In these contexts, the performance of the network can become critical under high traffic flows, e.g. during peak hours, highlighting the vulnerability of the transport system. In addition to ordinary traffic conditions, it is also important to consider anomalous events that may partially or totally reduce the capacity of certain road segments, which may affect the functionality of the whole network, or a significant part of it. The saturation of a network segment not only affects travel times, but can also have significant Knock-on effects, particularly if it affects strategic routes, leading to key attractors, such as hospitals or civil protection facilities. This study aims to analyze and assess the vulnerability of the transport network through the definition of a composite indicator, i.e. the Network Resilience Index (NRI), which integrates structural and functional characteristics of the network, followed by scenario-based analyses to assess its resilience. The NRI considers link connectivity, traffic flow intensity, accessibility factors to strategic points of interest, and public transport network characteristics. The methodological approach is developed by implementing a macroscopic simulation model to estimate the factors needed to calculate the NRI. The index is composed of four components: (i) the number of routes passing through each link, derived from the zoning of the study area and the implementation of a traffic assignment procedure; (ii) the level of congestion, expressed in terms of vehicle volumes relative to capacity; (iii) the utilization of the link by public transport, evaluated according to the number of transit services; and (iv) the role of the link in the shortest paths to key attractors. The NRI allows for the identification of the most critical road segments in terms of network vulnerability, both under normal operating conditions and in extraordinary scenarios, that limit connectivity, capacity or overall resilience. The results can guide strategies for preventing and managing critical issues, both planned (e.g. maintenance) and unplanned (e.g. natural events), supporting transport planning, improving accessibility to strategic attractors, and helping to mitigate the externalities of network congestion, including on public transport.

ABSTRACT. Public transport (PT) systems are prone to inherent schedule deviations and disruptions, which require proper analytical frameworks to formulate effective interventions. This study develops an automated geofencing algorithm for PT route segmentation and classification based on large-scale AVL data. It reduces the dimensionality of PT operational performance analytics and allows to identify the critical route sections for travel time improvements. As such, it can serve as a useful evidence-based decision support for PT agencies and operators.

(extended abstract is provided in the attached PDF file)