This article proposes an evidence-based policy recommendation framework integrating social media data and natural language processing methods, to support inclusive and efficient transport policy-making. Given that current research underscores the crucial role of both external and psychological variables in individual travel decisions, psychological features – such as beliefs, attitudes or values – are frequently used as latent variables for travel behaviour interpretation and travel choice modelling. However, user-centric policy recommendations based on dynamic psychological variables are still limited. Most studies rely on survey data, which neglects the urgent dynamic trend of user perception change and its underlying relationship with travel behaviour. Hence there is a lack of illustration on how these psychological variables can be further used at specific temporal and spatial levels for travel behaviour interpretation. This would be valuable to identify priorities for more targeted (sustainability and other) policies and interventions. In this article, we utilize sentiment analysis and dynamic topic modelling to represent the spatial–temporal variance of psychological features. Integrating with corresponding travel behaviour, we illustrate how these dynamic psychological features can distinguish travel dissonance, identify key motivations, and reflect urgent social demands at precise spatial–temporal levels. We demonstrate these advances in a case study in New York City from 2019 to 2022 using Twitter (X) data. A comparison with existing travel-related policies in the case study validates the feasibility of our framework to support evidence-based policy recommendations. We conclude by discussing the potential of this framework to support sustainable transport promotion.

Bicycle delay is an important variable to assess the performance of the cycling transportation system, especially as an indicator of intersection efficiency. This article estimates a machine learning (ML)-based model for estimating average bicycle delays at signalized intersections. This study evaluates various ML models with regressor features, including random forest, k-nearest neighbor, support vector regression, extreme gradient boosting, and neural networks. Sparse GPS cycling data (as reference data) from the Talking Bikes program in the Netherlands and the local control signal and flow detection information from the VLOG data provided by a Dutch city are adopted to train the ML models. The findings illustrate the viability of estimating bicycle delays by considering the interplay among weather conditions, temporal factors, junction topology, and local traffic conditions. The estimation model fit using the best-performing model - random forest - has doubled compared to the case without such additional traffic information, indicating its improved performance. Insights gained from the estimation model emphasize the potential of data-driven approaches to inform traffic management, bicycle policy, and infrastructure development.

Are you using tools like ChatGPT in your daily life to help write an email or even draft a construction plan? Just ten years ago, these kinds of capabilities would have seemed unimaginable. Today, they’re becoming part of everyday life for ordinary people. Behind these powerful tools are technologies known as Large Language Models (LLMs)—AI systems that can understand and generate human-like text and now even create images and videos. But what exactly are LLMs? Could they help transform fields like transportation and traffic management? Can they really do everything, or are there still limitations? In this article, we’ll walk you through a general introduction to LLMs: what they are, how they work, and what opportunities—and challenges—they bring to the transportation sector.

To escape a dangerous building emergency occupants may need to respond quickly, assess the environment, plan their actions and tackle possible problems during evacuation. In this study 147 participants were tested in an experimental evacuation design for the effects of three environmental factors (fire alarm, lighting and emergency exit signs illumination) on problem-solving abilities. The experimental evacuation scenarios consisted of: (1) fire alarm, normal lighting conditions and illuminated emergency exit signs, (2) fire alarm, dark environment and illuminated emergency exit signs and (3) fire alarm, dark environment and not illuminated emergency exit signs. The tested problem-solving abilities were the time to plan actions and number of excess moves on the Tower of London test. The main results indicate that the third experimental evacuation scenario led to a decrease of 25.9% in planning time, compared to the control scenario. Age also had a significant effect on planning time. The oldest participants took or needed on average 42 s more planning time than the youngest participants, an increase of 146.9%. Furthermore, the second and third experimental evacuation scenario led to significant more excess moves, compared to the control scenario. However, the older the participants the less excess moves they had. For gender no significant effects on problem-solving abilities were found. In addition, the relationships between problem-solving abilities and building evacuation time were investigated. Longer planning times were associated with longer evacuation times and more excess moves were associated with shorter evacuation times. Practical implications for building and safety managers are to add training in darkness or assume more evacuation time in darkness or for older aged populations in evacuation plans and drills. Future research should collect more quantitative data about effects of various environmental factors and personal characteristics, such as problem-solving styles, age and gender, on building evacuation behaviour.

Bicycle transportation, a low-carbon option, is essential for promoting sustainable urban mobility. However, predicting bicycle traffic is challenging due to limited investments in data collection, especially in smaller cities. This paper proposes a multi-source transfer learning spatial-temporal graph neural network (Multi-TLSTGCN) for accurate bicycle traffic prediction in target cities with limited available data. This study first examines how to transfer knowledge from single source domain to the target domain while mitigating the risk of negative transfer. Following this, a multi-source adaptive transfer learning approach is developed to optimize traffic prediction in the target domain by adaptively integrating knowledge from multiple sources. Finally, the performance of the Multi-TLSTGCN model is evaluated under various levels of target data scarcity and compared with models that do not incorporate source domain knowledge. The experimental results demonstrate several key insights: 1) Models fine-tuned with a single-cluster pre-trained source model where the clusters are formed based on similar traffic patterns are more effective at minimizing negative knowledge transfer than those fine-tuned with single-city pre-trained source models. 2) The proposed Multi-TLSTGCN outperforms baseline models in bicycle traffic prediction, showing promise for accurate predictions in data-scarce environments; and 3) The Multi-TLSTGCN model remains robust across varying levels of data scarcity, exhibiting only a slight decrease in accuracy as the availability of target data decreases, in contrast to models relying solely on target domain data. These findings highlight the Multi-TLSTGCN model as an effective and promising solution for bicycle traffic prediction with limited data availability.

Mobility is vital for societal wellbeing, economic growth, social inclusion, and access to essential amenities. However, the current system faces significant challenges, including environmental impact, unequal access, and safety concerns. […]

Traffic flow variables are essential for understanding, analysing, and optimising how pedestrians move in urban areas. In this chapter, we introduce the variables that can be used to describe a pedestrian flow. These variables can be microscopic (individual pedestrians), macroscopic (aggregate level), or mesoscopic (distributions of microscopic quantities). We start with the most detailed level of description, a trajectory. The trajectory describes the dynamics of an individual pedestrian as a function of time by its (two-dimensional) position in space. Based on the trajectory, we derive the most relevant microscopic variables, velocity and acceleration. Then, we give different definitions of density, one of the key macroscopic flow variables, by taking a snapshot of a pedestrian traffic situation at a time instant. Density is used to express crowdedness and level-of-service. Next, we look at the space-mean velocity of a pedestrian flow, followed by the third key macroscopic variable, the flow rate. The flow rate is the number of pedestrians passing a cross-section during a certain time period. We end this chapter with the generalised definitions of flow, density, and velocity for a time-space region, and show how they are related.

Accurate short-term predictions of active mode traffic are crucial for effective urban traffic control and management, helping to reduce delays, stops, and improve travel time reliability, and optimize travel route choice. While most methods focus on motorized traffic, active modes like walking and cycling have been overlooked due to their complex dynamics and sensitivity to external factors like weather and individual choices, making them inherently less predictable. To address this, we propose a Dynamic Attention-based Spatial-Temporal Graph Convolutional Network (DyASTGCN) model that incorporates the impact of weather on graph spatial correlations within the active mode traffic network. Additionally, we introduce a fusion approach to integrate various heterogeneous spatial correlations, aiming to represent the optimal spatial correlations within the active mode network. Experimental results demonstrate that weather changes have a lagging effect on traffic network spatial correlations. Specifically, active mode traffic demonstrates significant sensitivity to precipitation, with notable changes in spatial correlations occurring within 5 minutes. Conversely, it takes approximately 20 minutes for spatial correlations to respond to wind speed influences. By incorporating both precipitation and wind speed with a 20-minute lag, our model outperforms those using only one feature, achieving the best traffic prediction performance. Given the uncertain traffic state and highly sparse nature of active mode data, our fusion approach adeptly captures the essential spatial correlations required for accurate traffic flow prediction. This allows our model to better understand complex graph correlations and traffic patterns, improving prediction accuracy and offering valuable insights into active mode network dynamics.

Car dominance in urban landscapes poses environmental, health, and congestion challenges. This comprehensive study examines the potential of shared mobility in car-free areas. Specifically, it investigates the mobility behaviour of inner-city older adult residents (50 + ), traditionally heavy car users through a case study of small-medium-sized Dutch cities and a stated preference experiment. This study applies a Latent Class model to analyse the heterogeneity in passengers’ preferences, identifying four distinct groups: Price Sensitive & Private Car Enthusiasts, Time-Conscious Travellers, Pro-Cycling & Conventional travellers, and Micromobility Enthusiasts. The model predicts class membership based on travel behaviour data from the stated choice experiment and examines the role of key factors such as travel cost, travel time, and walking distance in shaping mode choices across five transport options: bike, e-bike, e-scooter, e-Brommobiel, and e-car. The findings reveal that a significant portion of travellers recognise the value of shared mobility options in reducing private car dependency, underscoring the need for targeted interventions to address barriers and enhance accessibility to promote shared mobility adoption. Based on these distinct passenger segments, the study proposes specific policy measures that not only enhance transport planning but also address existing challenges and user concerns in sustainable urban mobility.

Walking is the most fundamental form of human transportation. It is also the most sustainable mode, and very frequently used in a variety of situations. As such, pedestrian behavior influences the design of our infrastructure, the efficiency of transit systems, and even the safety of large crowds. Despite its ubiquity and importance, the study of pedestrian and crowd dynamics remains an evolving field within transport science. This book aims to give readers a comprehensive overview of pedestrian dynamics, covering fundamental theories, modeling approaches, empirical findings, and applications. All in relation to pedestrian planning. This book is divided into two parts. The first part lays the foundations of pedestrian movement dynamics and behavior, while the second part explores applications of the foundational knowledge presented in the first part. By integrating pedestrian flow analysis, human behavior insights, modeling techniques, and applications, this book contributes to a deeper understanding of how to create walkable, resilient, and human-centered urban environments.

Driving heterogeneity significantly influences traffic performance, contributing to traffic disturbances, increased crash risks, and inefficient fuel use and emissions. With the growing availability of driving behaviour data, Machine Learning (ML) techniques have become widely used for analysing driving behaviour and identifying heterogeneity. This paper presents a systematic review of current ML-based methods for driving heterogeneity identification. The review organises key concepts and categorisations of driving heterogeneity, highlights strengths and drawbacks of various methods, and outlines applications of identification analysis. Based on the literature review, we propose a structured framework that guides the ML-based identification process. The framework starts with an extensive data collection and rigorous pre-processing process, followed by feature selection techniques that identify features most indicative of driving behaviours. Sophisticated models including supervised, unsupervised, semi-supervised, and reinforcement learning techniques are discussed with multi-perspective performance evaluation. This paper provides a comprehensive reference for researchers and practitioners to understand driving heterogeneity, supporting the development of data-driven solutions for improving traffic management and road safety.

In public spaces such as city centers, train stations, airports, shopping malls, and multi-modal hubs, accurately predicting pedestrian flow is crucial for effective crowd management e.g. congestion prevention and evacuation planning. Traditional microscopic simulation models offer fine-grained insights by simulating each pedestrian individually, but they are computationally intensive and typically used at the planning and design stage, making them unsuitable for real-time interventions in high-demand scenarios. Macroscopic models, on the other hand, reduce computational cost by aggregating pedestrian behavior and solving partial differential equations, but they typically require estimates of traffic states such as density and speed — quantities that are difficult to measure accurately in practice. Additionally, as the complexity of these physics-based models increases, their computational feasibility for real-time use becomes even more limited. Data-driven (machine learning) models provide a computationally efficient alternative, enhancing real-time prediction capabilities. However, they often require large historical datasets to generalize well, and their performance can degrade under out-of-distribution (OOD) conditions. Moreover, most black-box learning models lack interpretability and domain-specific insights, limiting their practical adoption. In this paper, we propose a novel pedestrian flow prediction model based on the theory of crowd diffusion. Our method estimates flow rates directly from sensor-observed data and infers both Origin–Destination (OD) demand and route choice probabilities to support real-time operations. To address the OOD challenge, we incorporate an online learning mechanism that continuously calibrates model parameters based on incoming observations.

As automated vehicles (AVs) become more common, it is important to understand how human-driven vehicles (HDVs) would interact with them. This research investigated HDV gap acceptance behavior in mixed traffic with AVs at a priority intersection, focusing on how mixed traffic factors affect this behavior and overall traffic efficiency. Using a driving simulator, four scenarios were tested by varying AV driving style (less defensive, more defensive, and HDV-like) and AV recognizability (distinguishable or not from HDVs). Gap acceptance models were estimated based on the collected trajectory data. These models were then integrated into the SUMO microscopic traffic simulation platform, where a T-intersection network was set up. Simulation runs varied based on AV driving style, recognizability, penetration rate (0-75% in 25% increments), and whether HDV behavioral adaptation was considered. The results indicated increased minor road vehicle delays with higher AV penetration rates. Recognizable less defensive AVs, and more defensive AVs with high penetration rates caused the largest delays for minor road vehicles compared to other conditions. Ignoring behavioral adaptation led to a delay underestimation of up to 75% for minor road vehicles. In conclusion, there is behavioral adaptation in gap acceptance of HDVs in mixed traffic environments. Taking into account the behavioral adaptation is essential for accurately assessing traffic efficiency in mixed traffic conditions, and guiding AV deployment policies.

Large Language Models zijn AI-systemen die menselijke taal begrijpen en zich er ook in kunnen uiten. Ze zijn de basis onder populaire applicaties als ChatGPT, Gemini en Copilot. Maar inmiddels is de technologie zó breed inzetbaar dat ze ook doordringt in de mobiliteitssector. Hoe werken de Large Language Models? Hoe kunnen ze van nut zijn in ons vakgebied? En wat zijn de mitsen en maren

In urban centers, cycling is increasingly popular as an eco-friendly transportation mode and a short-distance transport option, driving higher demand for accurate bicycle travel time estimation. Policymakers need to understand bicycle traffic for urban traffic management and sustainable transport promotion, while cyclists benefit from better route planning and improved network efficiency. However, urban bicycle travel time estimation has not received as much attention as car traffic estimation and presents several challenges: 1) Limited availability of structural cycling data, which can be inaccessible due to privacy concerns and/or severely biased by user demographics. 2) The diverse and complex behaviors of cyclists. 3) The lack of strict road constraints for cyclists and frequent rule violations, complicating the model definition of a comprehensive cycling infrastructure network. This paper presents the first study on urban bicycle travel time estimation using GPS tracking data. Leveraging graph-based deep learning's ability to learn from topological network information, we introduce the Dual Graph-based approach for bicycles (DG4b), which employs two parallel encode-process-decode pipelines: one for a shared undirected road network graph to capture intrinsic road characteristics, and another for a directed trip-specific graph reflecting unique trip features. The outputs are combined to estimate road segment speeds and overall trip travel time. When applied to a real-world dataset from Berlin, our method shows superior accuracy and reliability compared to baseline models, while maintaining low complexity. Our approach provides a novel perspective on integrating bicycling-specific characteristics and aims to inspire more future research in bicycle-related traffic estimation.

Crowd management plays a vital role in urban planning and emergency response. Accurate crowd prediction is important for venue operators to respond effectively to adverse crowd dynamics during large gatherings. Although many studies have tried to predict crowd densities or movement dynamics with data-driven predictive models, their validation is often limited to data within the same scenario. As a result, the predictability of the data-driven model in unseen scenarios, such as evacuation scenarios, remains unknown due to the challenges of collecting out-of-distribution data regarding emergency conditions. To address this problem, we present an evaluation pipeline to evaluate different kinds of data-driven models. A method is proposed to generate realistic scenarios by simulation and collect synthetic data from these scenarios to acquire a comprehensive dataset. With these synthetic data, we evaluated different predictive models, from traditional machine learning methods to deep learning time-series prediction models, to explore their generalizability. Furthermore, we propose a weighted average metric, which is better suited to determine the performance of forecasting algorithms under adverse conditions. Through extensive experimentation, we showcase the heterogeneity and diversity of the simulation dataset. The evaluation results also revealed that all the data-driven models performed poorly in unseen scenarios, highlighting the urgent need to develop a robust and generalizable model for predicting crowd flow in indoor spaces.

Identifying driving heterogeneity plays an important role in improving traffic safety and efficiency. This paper proposes a novel framework to identify driving heterogeneity from the underlying characteristics of driving behaviour. The framework includes three processes: Action phase extraction, Action pattern calibration, and Action pattern classification. The concepts of Action phase and Action patterns are proposed to decipher and interpret driving behaviours. Action phases are extracted by rule-based segmentation methods and Action patterns are calibrated based on an unsupervised learning approach. The extraction and calibration processes provide a rigorous labelling approach for the attention-based LSTM Action pattern classification process. Evaluation of the framework on a large-scale naturalistic driving dataset reveals six distinct Action patterns. The implementation of the attention mechanism to LSTM models significantly enhanced both the accuracy and time efficiency of Action pattern identification. The proposed framework offers benefits in detecting and reducing variability in driving behaviour through ITS applications such as user-based traffic management, personalised Advanced Driver Assistance Systems (ADAS), and advanced autonomous vehicles (AV) design, thereby enhancing road safety and traffic efficiency.

The field of transportation research addresses the complexities of a particular sociotechnical system. Its usual focus is on human transportation systems, but non-human systems that effect transportation are also abundant in nature. This paper draws an analogy between modern human transportation systems and mycorrhizal networks (MN), the underground networks formed by fungi and plants for resource transportation. By examining MN, the study aims to extract insights applicable to human transport and to explore potential reciprocal learnings about natural systems. The research emphasizes an interdisciplinary approach that acknowledges both the technical and social dimensions of transport. The primary focus is to propose improvements to human transportation by learning from the natural efficiency of MN, thereby fostering a more holistic understanding and implementation of transport solutions.

Physical distancing has been an important asset in limiting the SARS-CoV-2 virus spread during the COVID-19 pandemic. This study aims to assess compliance with physical distancing and to evaluate the combination of observed and self-reported data used. This research shows that it is difficult to operationalize new rules, that context affects compliance, that there needs to be a need for compliance, and that rules require upkeep. From a methodological point of view, this study found that the combined methods provide a comprehensive picture of compliance behaviour, that it is challenging but essential to mitigate response fatigue in long-term monitoring studies, and that it would be interesting in future research to learn how actual behaviour is influenced by personal narratives.

To promote urban sustainability, many cities are adopting bicycle-friendly policies, leveraging GPS trajectories as a vital data source. However, the inherent errors in GPS data necessitate a critical preprocessing step known as map-matching. Due to GPS device malfunction, road network ambiguity for cyclists, and inaccuracies in publicly accessible streetmaps, existing map-matching methods face challenges in accurately selecting the best-mapped route. In urban settings, these challenges are exacerbated by high buildings, which tend to attenuate GPS accuracy, and by the increased complexity of the road network. To resolve this issue, this work introduces a map-matching method tailored for cycling travel data in urban areas. The approach introduces two main innovations: a reliable classification of road availability for cyclists, with a particular focus on the main road network, and an extended multi-objective map-matching scoring system. This system integrates penalty, geometric, topology, and temporal scores to optimize the selection of mapped road segments, collectively forming a complete route. Rotterdam, the second-largest city in the Netherlands, is selected as the case study city, and real-world data is used for method implementation and evaluation. Hundred trajectories were manually labelled to assess the model performance and its sensitivity to parameter settings, GPS sampling interval, and travel time. The method is able to unveil variations in cyclist travel behavior, providing municipalities with insights to optimize cycling infrastructure and improve traffic management, such as by identifying high-traffic areas for targeted infrastructure upgrades and optimizing traffic light settings based on cyclist waiting times.