交通、大模型、出行、仿真

Large Language Models (LLMs), capable of handling multi-modal input and outputs such as text, voice, images, and video, are transforming the way we process information. Beyond just generating textual responses to prompts, they can integrate with different software platforms to offer comprehensive solutions across diverse applications. In this paper, we present ChatSUMO, an LLM-based agent that integrates language processing skills to generate abstract and real-world simulation scenarios in the widely-used traffic simulator - Simulation of Urban MObility (SUMO). Our methodology begins by leveraging the LLM for user input, which adapts it to relevant keywords needed to run python scripts. These scripts are designed to convert specified regions into coordinates, fetch data from OpenStreetMap, transform it into a road network, and subsequently run SUMO simulations with the designated traffic conditions. The outputs of the simulations are then interpreted by the LLM resulting in informative comparisons and summaries. Users can continue the interaction and generate a variety of customized scenarios without prior traffic simulation expertise. Any city available from OpenStreetMap can be imported, and for demonstration, we created a real-world simulation for the city of Albany. ChatSUMO also allows simulation customization capabilities of edge edit, traffic light optimization, and vehicle edit by users through the web interface.

No abstract available

The growing complexity of urban mobility systems has made traffic simulation indispensable for evidence-based transportation planning and policy evaluation. However, despite the analytical capabilities of platforms such as the Simulation of Urban MObility (SUMO), their application remains largely confined to domain experts. Developing realistic simulation scenarios requires expertise in network construction, origin-destination modeling, and parameter configuration for policy experimentation, creating substantial barriers for non-expert users such as policymakers, urban planners, and city officials. Moreover, the requests expressed by these users are often incomplete and abstract-typically articulated as high-level objectives, which are not well aligned with the imperative, sequential workflows employed in existing language-model-based simulation frameworks. To address these challenges, this study proposes AgentSUMO, an agentic framework for interactive simulation scenario generation via large language models. AgentSUMO departs from imperative, command-driven execution by introducing an adaptive reasoning layer that interprets user intents, assesses task complexity, infers missing parameters, and formulates executable simulation plans. The framework is structured around two complementary components, the Interactive Planning Protocol, which governs reasoning and user interaction, and the Model Context Protocol, which manages standardized communication and orchestration among simulation tools. Through this design, AgentSUMO converts abstract policy objectives into executable simulation scenarios. Experiments on urban networks in Seoul and Manhattan demonstrate that the agentic workflow achieves substantial improvements in traffic flow metrics while maintaining accessibility for non-expert users, successfully bridging the gap between policy goals and executable simulation workflows.

Realistic and controllable traffic simulation is a core capability that is necessary to accelerate autonomous vehicle (AV) development. However, current approaches for controlling learning-based traffic models require significant domain expertise and are difficult for practitioners to use. To remedy this, we present CTG++, a scene-level conditional diffusion model that can be guided by language instructions. Developing this requires tackling two challenges: the need for a realistic and controllable traffic model backbone, and an effective method to interface with a traffic model using language. To address these challenges, we first propose a scene-level diffusion model equipped with a spatio-temporal transformer backbone, which generates realistic and controllable traffic. We then harness a large language model (LLM) to convert a user's query into a loss function, guiding the diffusion model towards query-compliant generation. Through comprehensive evaluation, we demonstrate the effectiveness of our proposed method in generating realistic, query-compliant traffic simulations.

Building upon prior work in using generative AI for traffic scenario generation, this study exploits fine-tuning of Large Language Models (LLMs) to generate road geometries for Simulation of Urban MObility (SUMO). While previous applications focused on generating traffic agents, simulation configurations, and high-level scenarios, the current work shifts focus to infrastructure-level generation. We fine-tuned a foundation language model on a curated dataset of SUMO road network configurations, including XML-based definitions and geometric layouts. The resulting model is capable of producing syntactically valid and semantically meaningful road networks, supporting some configurations of road segments. We evaluated the outputs generated for validity, diversity, and utility in downstream simulation tasks. This approach demonstrates a scalable pathway for procedurally generating traffic environments, allowing rapid prototyping, testing, and training for autonomous vehicle simulations and intelligent transportation systems.

Evaluating autonomous vehicles with controllability enables scalable testing in counterfactual or structured settings, enhancing both efficiency and safety. We introduce LangTraj, a language-conditioned scene-diffusion model that simulates the joint behavior of all agents in traffic scenarios. By conditioning on natural language inputs, LangTraj provides flexible and intuitive control over interactive behaviors, generating nuanced and realistic scenarios. Unlike prior approaches that depend on domain-specific guidance functions, LangTraj incorporates language conditioning during training, facilitating more intuitive traffic simulation control. We propose a novel closed-loop training strategy for diffusion models, explicitly tailored to enhance stability and realism during closed-loop simulation. To support language-conditioned simulation, we develop Inter-Drive, a large-scale dataset with diverse and interactive labels for training language-conditioned diffusion models. Our dataset is built upon a scalable pipeline for annotating agent-agent interactions and single-agent behaviors, ensuring rich and varied supervision. Validated on the Waymo Open Motion Dataset, LangTraj demonstrates strong performance in realism, language controllability, and language-conditioned safety-critical simulation, establishing a new paradigm for flexible and scalable autonomous vehicle testing. Project Website: https://langtraj.github.io/

This paper explores the potential of using large language models (LLMs), such as GPT-5.2 and Gemini 3, in the transportation industry through applications in vehicle design, autonomous navigation, traffic control, and other areas. Special attention is given to augmented sampling generation and multimodal processing. Key issues discussed include safety certification, model transparency, and ethical considerations of their implementation. Purpose: to investigate the prospects for using AI agents based on LLMs in the transport industry. Results: this paper considers the use of augmented sampling generation and multimodal data processing, along with examples, including traffic light control using AI, simulation scenario generation and driver fatigue analysis. Theoretical Significance: this paper concludes that the synergy between AI and transportation will inevitably lead to increased safety and efficiency, and that LLMs will play a significant role in future intelligent adaptive transport systems.

Microscopic traffic simulation has become an important tool for autonomous driving training and testing. Although recent data-driven approaches advance realistic behavior generation, their learning still relies primarily on a single real-world dataset, which limits their diversity and thereby hinders downstream algorithm optimization. In this paper, we propose DriveGen, a novel traffic simulation framework with large models for more diverse traffic generation that supports further customized designs. DriveGen consists of two internal stages: the initialization stage uses a large language model and retrieval technique to generate map and vehicle assets; the rollout stage outputs trajectories with selected waypoint goals from a visual language model and a specifically designed diffusion planner. Through this two-staged process, DriveGen fully utilizes large models’ high-level cognition and reasoning of driving behavior, obtaining greater diversity beyond datasets while maintaining high realism. To support effective downstream optimization, we additionally develop DriveGen-CS, an automatic corner case generation pipeline that uses failures of the driving algorithm as additional prompt knowledge for large models without the need for retraining or fine-tuning. Experiments show that our generated scenarios and corner cases have superior performance compared to state-of-the-art baselines. Downstream experiments further verify that the synthesized traffic of DriveGen provides better optimization of the performance of typical driving algorithms, demonstrating the effectiveness of our framework.

No abstract available

No abstract available

Conventional traffic flow simulations struggle to reconcile macroscopic statistical trends with the causal logic of microscopic events, a limitation that hinders their application in advanced intelligent transportation systems. To address this gap, this study introduces a novel hybrid framework that integrates data-driven statistical modeling, knowledge-based causal reasoning, and the narrative capabilities of Large Language Models (LLMs). Our method first establishes a simulation baseline by capturing periodic traffic features with a statistical model. An LLM is then used to automatically construct a domain-specific knowledge base, which provides logical constraints for simulating stochastic events. Finally, the LLM generates coherent event narratives, and their causal impacts are dynamically integrated into the baseline traffic flow. Experiments show that our framework produces diverse traffic scenarios that are statistically realistic, logically coherent, and narratively consistent. This research presents a new approach for generating complex traffic data with both high fidelity and high-level semantics, thereby bridging a critical gap in the field.

Simulation forms the backbone of modern self-driving development. Simulators help develop, test, and improve driving systems without putting humans, vehicles, or their environment at risk. However, simulators face a major challenge: They rely on realistic, scalable, yet interesting content. While recent advances in rendering and scene reconstruction make great strides in creating static scene assets, modeling their layout, dynamics, and behaviors remains challenging. In this work, we turn to language as a source of supervision for dynamic traffic scene generation. Our model, LCTGen, combines a large language model with a transformer-based decoder architecture that selects likely map locations from a dataset of maps, and produces an initial traffic distribution, as well as the dynamics of each vehicle. LCTGen outperforms prior work in both unconditional and conditional traffic scene generation in terms of realism and fidelity. Code and video will be available at https://ariostgx.github.io/lctgen.

No abstract available

Traffic simulation aims to learn a policy for traffic agents that, when unrolled in closed-loop, faithfully recovers the joint distribution of trajectories observed in the real world. Inspired by large language models, tokenized multi-agent policies have recently become the state-of-the-art in traffic simulation. However, they are typically trained through open-loop behavior cloning, and thus suffer from covariate shift when executed in closed-loop during simulation. In this work, we present Closest Among Top-K (CAT-K) rollouts, a simple yet effective closed-loop fine-tuning strategy to mitigate covariate shift. CAT-K fine-tuning only requires existing trajectory data, without reinforcement learning or generative adversarial imitation. Concretely, CAT-K fine-tuning enables a small 7M-parameter tokenized traffic simulation policy to outperform a 102M-parameter model from the same model family, achieving the top spot on the Waymo Sim Agent Challenge leaderboard at the time of submission. The code is available at https://github.com/NVlabs/catk.

Achieving diverse pedestrian behaviors in traffic simulation is crucial for enhancing the realism and complexity necessary for autonomous vehicle (AV) testing. Human drivers adjust behavior based on context, balancing caution with errors that may contribute to accidents. Real-world scenarios — such as congestion, overtaking, lane changes, and interactions between human-driven and AVs — often create adversarial conditions that challenge autonomous driving (AD), especially in heterogeneous and densely populated environments. While recent studies have advanced AV capabilities in perception, risk assessment, adaptive control, and decision-making, a significant gap remains in the realistic and context-sensitive simulation of pedestrian behaviors, particularly in pedestrian-dense settings. To address this, we propose a large language model (LLM)-enhanced traffic editing strategy that dynamically configures diverse pedestrian behaviors in AD simulations. By harnessing the generative and interpretive capacities of LLMs, our approach facilitates real-time, prompt-based customization of pedestrian scenarios, thereby enhancing the adaptability and complexity of simulations. Broadly compatible with existing simulators, this strategy elevates the rigor of AV testing. Experimental results in CARLA and Udacity simulators indicate that AVs tested in LLM-enhanced environments demonstrate higher collision rates under autopilot compared to default scenarios, highlighting the increased realism and adversarial conditions introduced. Once implemented, the approach manages pedestrian behaviors through the LLM, substantially improving the efficiency of handling complex traffic patterns and fostering robust AD testing. The primary source code of this project is publicly available at https://aizierjiang.github.io/LLM4TrafficEditor .

Autonomous driving significantly benefits from datadriven deep neural networks that require large-scale and highquality datasets for training. To achieve higher performance or conduct critical testing, trajectories with game-theoretic properties are necessary. However, defining such scenarios and collecting large-scale data from reality or simulation is hard and costly. In this work, we propose a novel two-step method that generates traffic trajectories with game-theoretic properties, based on the Large Language Model (LLM). We apply the method to generate a challenging traffic trajectory dataset, which is validated on many downstream tasks, including perception and prediction. The experimental results show that the generated trajectories can achieve high-difficulty testing and advanced training. We conclude that our method successfully generates the challenging traffic trajectories, which contribute to collecting such critical data.

Large Language Models (LLMs) offer significant promise for intelligent traffic management; however, current chain-based systems like TrafficGPT are hindered by sequential task execution, high token usage, and poor scalability, making them inefficient for complex, real-world scenarios. To address these limitations, we propose GraphTrafficGPT, a novel graph-based architecture, which fundamentally redesigns the task coordination process for LLM-driven traffic applications. GraphTrafficGPT represents tasks and their dependencies as nodes and edges in a directed graph, enabling efficient parallel execution and dynamic resource allocation. The main idea behind the proposed model is a Brain Agent that decomposes user queries, constructs optimized dependency graphs, and coordinates a network of specialized agents for data retrieval, analysis, visualization, and simulation. By introducing advanced context-aware token management and supporting concurrent multi-query processing, the proposed architecture handles interdependent tasks typical of modern urban mobility environments. Experimental results demonstrate that GraphTrafficGPT reduces token consumption by 50.2% and average response latency by 19.0% compared to TrafficGPT, while supporting simultaneous multi-query execution with up to 23.0% improvement in efficiency.

Abstract This study summarizes the first stage in the implementation of an agent-based freight modeling system that has a global representation of agents and detailed modeling of a large-scale transportation network. The model is used to evaluate the transportation and energy impacts of goods movement across urban and national scales. The framework is implemented within POLARIS, a C++-based Planning and Operations Language for Agent-based Regional Integrated Simulation, which consists of an activity-based modeling (ABM) and dynamic traffic assignment (DTA) system that has robust features for passenger travel. This platform provides a tool to model interactions among consumers, producers, and the transportation system. The main objective of this initial implementation is to implement a freight model within POLARIS following an agent-based paradigm with behavioral and simulation methods. This paper presents the initial framework and illustrates the application of the model. Building upon earlier works, a parcel location assignment algorithm for business establishments in the population is documented, along with a method for estimating establishment production and consumption volumes. In addition to population generation, other features of the model include push-pull supply chains, multimodal path choice, choice of transportation logistics node, and dynamic traffic assignment. A module with e-commerce supply and demand was also developed to analyze the effects of e-commerce delivery on last-mile energy use and congestion.

We present ChatScene, a Large Language Model (LLM)-based agent that leverages the capabilities of LLMs to gener-ate safety-critical scenarios for autonomous vehicles. Given unstructured language instructions, the agent first generates textually described traffic scenarios using LLMs. These scenario descriptions are subsequently broken down into several sub-descriptions for specified details such as behaviors and locations of vehicles. The agent then distinctively transforms the textually described sub-scenarios into domain-specific languages, which then generate actual code for prediction and control in simulators, facilitating the creation of diverse and complex scenarios within the CARLA simulation environment. A key part of our agent is a comprehensive knowledge retrieval component, which efficiently translates specific textual descriptions into corresponding domain-specific code snippets by training a knowledge database containing the scenario description and code pairs. Extensive experimental results underscore the efficacy of ChatScene in improving the safety of autonomous vehicles. For instance, the scenarios generated by ChatScene show a 15% increase in collision rates compared to state-of-the-art baselines when tested against different reinforcement learning-based ego vehicles. Furthermore, we show that by using our generated safety-critical scenarios to fine-tune different RL-based autonomous driving models, they can achieve a 9% reduction in collision rates, surpassing current SOTA methods. ChatScene effectively bridges the gap between textual descriptions of traffic scenarios and practical CARLA simulations, providing a unified way to conveniently generate safety-critical scenarios for safety testing and improvement for AVs. The code is available at https://github.com/javyduck/ChatScene.

City scene generation has gained significant attention in autonomous driving, smart city development, and traffic simulation. It helps enhance infrastructure planning and monitoring solutions. Existing methods have employed a two-stage process involving city layout generation, typically using Variational Autoencoders (VAEs), Generative Adversarial Networks (GANs), or Transformers, followed by neural rendering. These techniques often exhibit limited diversity and noticeable artifacts in the rendered city scenes. The rendered scenes lack variety, resembling the training images, resulting in monotonous styles. Additionally, these methods lack planning capabilities, leading to less realistic generated scenes. In this paper, we introduce CityCraft, an innovative framework designed to enhance both the diversity and quality of urban scene generation. Our approach integrates three key stages: initially, a diffusion transformer (DiT) model is deployed to generate diverse and controllable 2D city layouts. Subsequently, a Large Language Model(LLM) is utilized to strategically make land-use plans within these layouts based on user prompts and language guidelines. Based on the generated layout and city plan, we utilize the asset retrieval module and Blender for precise asset placement and scene construction. Furthermore, we contribute two new datasets to the field: 1)CityCraft-OSM dataset including 2D semantic layouts of urban areas, corresponding satellite images, and detailed annotations. 2) CityCraft-Buildings dataset, featuring thousands of diverse, high-quality 3D building assets. CityCraft achieves state-of-the-art performance in generating realistic 3D cities.

Understanding traveler behavior and accurately predicting travel mode choice are at the heart of transportation planning and policy-making. This study proposes TransMode-LLM, an innovative framework that integrates statistical methods with LLM-based techniques to predict travel modes from travel survey data. The framework operates through three phases: (1) statistical analysis identifies key behavioral features, (2) natural language encoding transforms structured data into contextual descriptions, and (3) LLM adaptation predicts travel mode through multiple learning paradigms including zero-shot and one/few-shot learning and domain-enhanced prompting. We evaluate TransMode-LLM using both general-purpose models (GPT-4o, GPT-4o-mini) and reasoning-focused models (o3-mini, o4-mini) with varying sample sizes on real-world travel survey data. Extensive experiment results demonstrate that the LLM-based approach achieves competitive accuracy compared to state-of-the-art baseline classifiers models. Moreover, few-shot learning significantly improves prediction accuracy, with models like o3-mini showing consistent improvements of up to 42.9\% with 5 provided examples. However, domain-enhanced prompting shows divergent effects across LLM architectures. In detail, it is helpful to improve performance for general-purpose models with GPT-4o achieving improvements of 2.27% to 12.50%. However, for reasoning-oriented models (o3-mini, o4-mini), domain knowledge enhancement does not universally improve performance. This study advances the application of LLMs in travel behavior modeling, providing promising and valuable insights for both academic research and transportation policy-making in the future.

The advent of large language models (LLMs) presents new opportunities for travel demand modeling. However, behavioral misalignment between LLMs and humans presents obstacles for the usage of LLMs, and existing alignment methods are frequently inefficient or impractical given the constraints of typical travel demand data. This paper introduces a novel framework for aligning LLMs with human travel choice behavior, tailored to the current travel demand data sources. Our framework uses a persona inference and loading process to condition LLMs with suitable prompts to enhance alignment. The inference step establishes a set of base personas from empirical data, and a learned persona loading function driven by behavioral embeddings guides the loading process. We validate our framework on the Swissmetro mode choice dataset, and the results show that our proposed approach significantly outperformed baseline choice models and LLM-based simulation models in predicting both aggregate mode choice shares and individual choice outcomes. Furthermore, we showcase that our framework can generate insights on population behavior through interpretable parameters. Overall, our research offers a more adaptable, interpretable, and resource-efficient pathway to robust LLM-based travel behavior simulation, paving the way to integrate LLMs into travel demand modeling practice in the future.

Route choice modeling (RCM) underpins travel behavior analysis, traffic policy evaluation, and network performance forecasting. Despite advances from discrete choice models to deep learning, existing methods struggle to integrate implicit contextual factors—such as spatial semantics—beyond explicit link attributes. Here, we propose a generalized link-based RCM framework that fuses structural knowledge from classical models with Large Language Model (LLM)-inferred context to both streamline the modeling process and boost predictive performance. Our approach tokenizes link transitions, enabling the seamless integration of structured features with LLM-derived semantics to enhance expressiveness and behavioral fidelity. We identify three key mechanisms: (1) granularity alignment between links and tokens for structured-context fusion; (2) implicit information compensation through LLMs' encoded environmental information; and (3) controlled generalization anchored by domain constraints and trajectory data. Experiments on diverse OD transfer tasks show that our LLM-enhanced models reduce path flow Jensen-Shannon divergence by up to 45 % compared to traditional Recursive Logit (RL), and outperform deep RL (DRL) models by $\text{1 0 - 2 0 \%}$ on average. These results advance a unified paradigm for integrating data, structural knowledge, and contextual information to improve traveler behavior modeling.

Effective modeling of how human travelers learn and adjust their travel behavior from interacting with transportation systems is critical for system assessment and planning. However, this task is also difficult due to the complex cognition and decision-making involved in such behavior. Recent research has begun to leverage Large Language Model (LLM) agents for this task. Building on this, we introduce a novel dual-agent framework that enables continuous learning and alignment between LLM agents and human travelers on learning and adaptation behavior from online data streams. Our approach involves a set of LLM traveler agents, equipped with a memory system and a learnable persona, which serve as simulators for human travelers. To ensure behavioral alignment, we introduce an LLM calibration agent that leverages the reasoning and analytical capabilities of LLMs to train the personas of these traveler agents. Working together, this dual-agent system is designed to track and align the underlying decision-making mechanisms of travelers and produce realistic, adaptive simulations. Using a real-world dataset from a day-to-day route choice experiment, we show our approach significantly outperforms existing LLM-based methods in both individual behavioral alignment and aggregate simulation accuracy. Furthermore, we demonstrate that our method moves beyond simple behavioral mimicry to capture the evolution of underlying learning processes, a deeper alignment that fosters robust generalization. Overall, our framework provides a new approach for creating adaptive and behaviorally realistic agents to simulate travelers'learning and adaptation that can benefit transportation simulation and policy analysis.

Travel choice analysis is crucial for understanding individual travel behavior to develop appropriate transport policies and recommendation systems in Intelligent Transportation Systems (ITS). Despite extensive research, this domain faces two critical challenges: a) modeling with limited survey data, and b) simultaneously achieving high model explainability and accuracy. In this paper, we introduce a novel prompt-learning-based Large Language Model(LLM) framework that significantly improves prediction accuracy and provides explicit explanations for individual predictions. This framework involves three main steps: transforming input variables into textual form; building of demonstrations similar to the object, and applying these to a well-trained LLM. We tested the framework's efficacy using two widely used choice datasets: London Passenger Mode Choice (LPMC) and Optima-Mode collected in Switzerland. The results indicate that the LLM significantly outperforms state-of-the-art deep learning methods and discrete choice models in predicting people's choices. Additionally, we present a case of explanation illustrating how the LLM framework generates understandable and explicit explanations at the individual level.

Agent-based modeling approaches represent the state-of-art in modeling travel demand and transportation system dynamics and are valuable tools for transportation planning. However, established agent-based approaches in transportation rely on multi-hierarchical mathematical models to simulate travel behavior, which faces theoretical and practical limitations. The advent of large language models (LLM) provides a new opportunity to refine agent-based modeling in transportation. LLM agents, which have impressive reasoning and planning abilities, can serve as a proxy of human travelers and be integrated into the modeling framework. However, despite evidence of their behavioral soundness, no existing studies have assessed the impact and validity of LLM-agent-based simulations from a system perspective in transportation. This paper aims to address this issue by designing and integrating LLM agents with human-traveler-like characteristics into a simulation of a transportation system and assessing its performance based on existing benchmarks. Using the classical transportation setting of the morning commute, we find that not only do the agents exhibit fine behavioral soundness, but also produce system dynamics that align well with standard benchmarks. Our analysis first verifies the effectiveness and potential of LLM-agent-based modeling for transportation planning on the system level.

Understanding travelers'route choices can help policymakers devise optimal operational and planning strategies for both normal and abnormal circumstances. However, existing choice modeling methods often rely on predefined assumptions and struggle to capture the dynamic and adaptive nature of travel behavior. Recently, Large Language Models (LLMs) have emerged as a promising alternative, demonstrating remarkable ability to replicate human-like behaviors across various fields. Despite this potential, their capacity to accurately simulate human route choice behavior in transportation contexts remains doubtful. To satisfy this curiosity, this paper investigates the potential of LLMs for route choice modeling by introducing an LLM-empowered agent,"LLMTraveler."This agent integrates an LLM as its core, equipped with a memory system that learns from past experiences and makes decisions by balancing retrieved data and personality traits. The study systematically evaluates the LLMTraveler's ability to replicate human-like decision-making through two stages of day-to-day (DTD) congestion games: (1) analyzing its route-switching behavior in single origin-destination (OD) pair scenarios, where it demonstrates patterns that align with laboratory data but cannot be fully explained by traditional models, and (2) testing its capacity to model adaptive learning behaviors in multi-OD scenarios on the Ortuzar and Willumsen (OW) network, producing results comparable to Multinomial Logit (MNL) and Reinforcement Learning (RL) models. These experiments demonstrate that the framework can partially replicate human-like decision-making in route choice while providing natural language explanations for its decisions. This capability offers valuable insights for transportation policymaking, such as simulating traveler responses to new policies or changes in the network.

Traditional cognitive travel modeling typically employs a unified cognitive model to simulate representative travel behaviors, which may usually result in a weak characterization of user heterogeneity in paths, modes, and other factors. Large language model (LLM), by contrast, has significantly enhanced the anthropomorphic and personalized features of intelligent systems. To integrate their advantages, this article proposes LLM-driven cognitive modeling to generate more diverse and personalized travel demands. The new method sufficiently exploits LLM such as the llama as a basis and provides personalized travel plans so that more heterogenous travel demands could be generated. Additionally, introducing LLM into cognitive modeling can significantly reduce the time of model development, thus accelerating the research or engineering deployment. By calibrating and testing with one month’s data from public transportation (buses and subways) in Beijing, our method, compared to traditional cognitive models, not only achieves better accuracy in reproducing typical travel patterns, but also generates more diverse ones, providing a more comprehensive input for computational experiments on traffic management and control strategies.

This paper introduces the concept of travel behavior embeddings, a method for re-representing discrete variables that are typically used in travel demand modeling, such as mode, trip purpose, education level, family type or occupation. This re-representation process essentially maps those variables into a latent space called the \emph{embedding space}. The benefit of this is that such spaces allow for richer nuances than the typical transformations used in categorical variables (e.g. dummy encoding, contrasted encoding, principal components analysis). While the usage of latent variable representations is not new per se in travel demand modeling, the idea presented here brings several innovations: it is an entirely data driven algorithm; it is informative and consistent, since the latent space can be visualized and interpreted based on distances between different categories; it preserves interpretability of coefficients, despite being based on Neural Network principles; and it is transferrable, in that embeddings learned from one dataset can be reused for other ones, as long as travel behavior keeps consistent between the datasets. The idea is strongly inspired on natural language processing techniques, namely the word2vec algorithm. Such algorithm is behind recent developments such as in automatic translation or next word prediction. Our method is demonstrated using a model choice model, and shows improvements of up to 60\% with respect to initial likelihood, and up to 20% with respect to likelihood of the corresponding traditional model (i.e. using dummy variables) in out-of-sample evaluation. We provide a new Python package, called PyTre (PYthon TRavel Embeddings), that others can straightforwardly use to replicate our results or improve their own models. Our experiments are themselves based on an open dataset (swissmetro).

Personal travel planning is a challenging task that aims to find a feasible plan that not only satisfies diverse constraints but also meets the demands of the user’s explicit and implicit preferences. In this paper, we study how to integrate the user’s implicit preference into the progress of travel planning. We introduce Re-alTravel, an augmented version of the Trav-elPlanner by incorporating real user reviews and point-of-interest metadata from Google Local. Based on RealTravel, we propose Personal Travel Solver (PTS), an integrated system that combines LLMs with numerical solvers to generate travel plans that satisfy both explicit constraints and implicit user preferences. PTS employs a novel architecture that seamlessly connects explicit constraint validation with implicit preference modeling through five specialized modules. The experimental results demonstrate the system’s effectiveness, achieving better performance than baseline methods, and improvement in the level of personalization. Our data and code are available at PersonalTravelSolver.

The topic of intermodal passenger mobility has become more important during the last 20 years. As mobility options increase in number and flexibility, it gets more and more attractive to combine multiple modes on single trips. In addition, intermodal travel behavior is expected to contribute to less car dependent mobility and transport sector’s reduction of greenhouse gas emissions. Creating and improving the conditions for such a behavior requires planning with knowledge about influencing factors and highest resistances. Empirical evidence and behavioral models can support decisions on measures improving intermodal travel supply. This work presents an agent-based model approach containing intermodal travel behavior with regard to its most important decisions. It enables the combination of a multitude of modes and can be extended to even more modes. By combining many decisions and influences it is comprehensible and adaptable to different surveys and circumstances. We show that results are realistic and impacts are valid to be able to forecast effects of potential measures. 2020 The Authors. Published by Elsevier B.V. his is an open access article under the license (http://creativeco ons.org/licenses/by-nc-nd/4.0/) i i ilit of the Conference Program Chairs.

Large language models (LLMs) are increasingly used as behavioral proxies for self-interested travelers in agent-based traffic models. Although more flexible and generalizable than conventional models, the practical use of these approaches remains limited by scalability due to the cost of calling one LLM for every traveler. Moreover, it has been found that LLM agents often make opaque choices and produce unstable day-to-day dynamics. To address these challenges, we propose to model each homogeneous traveler group facing the same decision context with a single representative LLM agent who behaves like the population's average, maintaining and updating a mixed strategy over routes that coincides with the group's aggregate flow proportions. Each day, the LLM reviews the travel experience and flags routes with positive reinforcement that they hope to use more often, and an interpretable update rule then converts this judgment into strategy adjustments using a tunable (progressively decaying) step size. The representative-agent design improves scalability, while the separation of reasoning from updating clarifies the decision logic while stabilizing learning. In classic traffic assignment settings, we find that the proposed approach converges rapidly to the user equilibrium. In richer settings with income heterogeneity, multi-criteria costs, and multi-modal choices, the generated dynamics remain stable and interpretable, reproducing plausible behavioral patterns well-documented in psychology and economics, for example, the decoy effect in toll versus non-toll road selection, and higher willingness-to-pay for convenience among higher-income travelers when choosing between driving, transit, and park-and-ride options.

Modeling subrational agents, such as humans or economic households, is inherently challenging due to the difficulty in calibrating reinforcement learning models or collecting data that involves human subjects. Existing work highlights the ability of Large Language Models (LLMs) to address complex reasoning tasks and mimic human communication, while simulation using LLMs as agents shows emergent social behaviors, potentially improving our comprehension of human conduct. In this paper, we propose to investigate the use of LLMs to generate synthetic human demonstrations, which are then used to learn subrational agent policies though Imitation Learning. We make an assumption that LLMs can be used as implicit computational models of humans, and propose a framework to use synthetic demonstrations derived from LLMs to model subrational behaviors that are characteristic of humans (e.g., myopic behavior or preference for risk aversion). We experimentally evaluate the ability of our framework to model sub-rationality through four simple scenarios, including the well-researched ultimatum game and marshmallow experiment. To gain confidence in our framework, we are able to replicate well-established findings from prior human studies associated with the above scenarios. We conclude by discussing the potential benefits, challenges and limitations of our framework.

Travel behavior prediction is a core problem in transportation demand management and is traditionally addressed using numerical models calibrated on observed data. With recent advances in large language models (LLMs), new opportunities have emerged to model human decision-making through natural language reasoning. This study explores the use of LLMs for travel behavior prediction through two complementary frameworks. The first framework employs a zero-shot prompting strategy, where the prediction task, traveler attributes, and relevant domain knowledge are described in text, enabling the LLM to directly generate predictions without task-specific training data. The second framework uses LLM-generated text embeddings as high-level representations of travel scenarios, which are then combined with conventional supervised learning models to support prediction in small-sample settings. Empirical results show that both approaches achieve performance comparable to, and in some cases competitive with, classical models such as multinomial logit, random forest, and neural networks. These findings suggest that LLMs offer a flexible and data-efficient alternative for travel behavior prediction.

No abstract available

No abstract available

No abstract available

Location-based services (LBS) have accumulated extensive human mobility data on diverse behaviors through check-in sequences. These sequences offer valuable insights into users' intentions and preferences. Yet, existing models analyzing check-in sequences fail to consider the semantics contained in these sequences, which closely reflect human visiting intentions and travel preferences, leading to an incomplete comprehension. Drawing inspiration from the exceptional semantic understanding and contextual information processing capabilities of large language models (LLMs) across various domains, we present Mobility-LLM, a novel framework that leverages LLMs to analyze check-in sequences for multiple tasks. Since LLMs cannot directly interpret check-ins, we reprogram these sequences to help LLMs comprehensively understand the semantics of human visiting intentions and travel preferences. Specifically, we introduce a visiting intention memory network (VIMN) to capture the visiting intentions at each record, along with a shared pool of human travel preference prompts (HTPP) to guide the LLM in understanding users' travel preferences. These components enhance the model's ability to extract and leverage semantic information from human mobility data effectively. Extensive experiments on four benchmark datasets and three downstream tasks demonstrate that our approach significantly outperforms existing models, underscoring the effectiveness of Mobility-LLM in advancing our understanding of human mobility data within LBS contexts.

As the applicability of Large Language Models (LLMs) extends beyond traditional text processing tasks, there is a burgeoning interest in their potential to excel in planning and reasoning assignments, realms traditionally reserved for System 2 cognitive competencies. Despite their perceived versatility, the research community is still unraveling effective strategies to harness these models in such complex domains. The recent discourse introduced by the paper on LLM Modulo marks a significant stride, proposing a conceptual framework that enhances the integration of LLMs into diverse planning and reasoning activities. This workshop paper delves into the practical application of this framework within the domain of travel planning, presenting a specific instance of its implementation. We are using the Travel Planning benchmark by the OSU NLP group, a benchmark for evaluating the performance of LLMs in producing valid itineraries based on user queries presented in natural language. While popular methods of enhancing the reasoning abilities of LLMs such as Chain of Thought, ReAct, and Reflexion achieve a meager 0%, 0.6%, and 0% with GPT3.5-Turbo respectively, our operationalization of the LLM-Modulo framework for TravelPlanning domain provides a remarkable improvement, enhancing baseline performances by 4.6x for GPT4-Turbo and even more for older models like GPT3.5-Turbo from 0% to 5%. Furthermore, we highlight the other useful roles of LLMs in the planning pipeline, as suggested in LLM-Modulo, which can be reliably operationalized such as extraction of useful critics and reformulator for critics.

Large Language Models (LLMs) have the potential to fundamentally change the way people engage in computer programming. Agent-based modeling (ABM) has become ubiquitous in natural and social sciences and education, yet no prior studies have explored the potential of LLMs to assist it. We designed NetLogo Chat to support the learning and practice of NetLogo, a programming language for ABM. To understand how users perceive, use, and need LLM-based interfaces, we interviewed 30 participants from global academia, industry, and graduate schools. Experts reported more perceived benefits than novices and were more inclined to adopt LLMs in their workflow. We found significant differences between experts and novices in their perceptions, behaviors, and needs for human-AI collaboration. We surfaced a knowledge gap between experts and novices as a possible reason for the benefit gap. We identified guidance, personalization, and integration as major needs for LLM-based interfaces to support the programming of ABM.

Travel planning with large language models (LLMs) offers promising solutions for creating personalized itineraries, but existing systems face some limitations: they rely on static user profiles for personalization rather than historical mobility patterns, lack real-world deployment and validation, and evaluate using previous-generation LLM models. We present P2P, a comprehensive LLM-powered travel planning system that addresses these gaps. Our system incorporates users' historical activity patterns and behaviors to generate personalized single-day and multi-day itineraries, moving beyond profile-based personalization. Through a production-ready Android application, users can create and iteratively refine real travel plans using natural language feedback, providing insight into LLM reliability in authentic usage scenarios. We compare three models: two open-source (Llama-4, DeepSeek-R1) and one proprietary (GPT-4.1) for itinerary generation, and are the first to employ GPT-5 as a judge alongside Llama-4 to assess plan quality and personalization effectiveness. Our findings reveal significant variation in planning capabilities between models, with models able to partially satisfy planning constraints but showing decreased performance as constraint complexity increases. Although LLMs demonstrate promising potential for personalized travel planning, models struggle with balancing personalization against other constraints, indicating substantial room for improvement in achieving robust deployment in real-world applications. We also reveal that the evaluation results vary based on the LLM model used for the evaluation, demonstrating how differences in their reasoning abilities affect the quality of the evaluation. The source code and the Android application are available at https://github.com/hasaansworld/travel-planner.

Human mobility is inextricably linked to social issues such as traffic congestion, energy consumption, and public health; however, privacy concerns restrict access to mobility data. Recently, research have utilized Large Language Models (LLMs) for human mobility generation, in which the challenge is how LLMs can understand individuals' mobility behavioral differences to generate realistic trajectories conforming to real world contexts. This study handles this problem by presenting an LLM agent-based framework (MobAgent) composing two phases: understanding-based mobility pattern extraction and reasoning-based trajectory generation, which enables generate more real travel diaries at urban scale, considering different individual profiles. MobAgent extracts reasons behind specific mobility trendiness and attribute influences to provide reliable patterns; infers the relationships between contextual factors and underlying motivations of mobility; and based on the patterns and the recursive reasoning process, MobAgent finally generates more authentic and personalized mobilities that reflect both individual differences and real-world constraints. We validate our framework with 0.2 million travel survey data, demonstrating its effectiveness in producing personalized and accurate travel diaries. This study highlights the capacity of LLMs to provide detailed and sophisticated understanding of human mobility through the real-world mobility data.

In transportation system demand modeling and simulation, agent-based models and microsimulations are current state-of-the-art approaches. However, existing agent-based models still have some limitations on behavioral realism and resource demand that limit their applicability. In this study, leveraging the emerging technology of large language models (LLMs) and LLM-based agents, we propose a general LLM-agent-based modeling framework for transportation systems. We argue that LLM agents not only possess the essential capabilities to function as agents but also offer promising solutions to overcome some limitations of existing agent-based models. Our conceptual framework design closely replicates the decision-making and interaction processes and traits of human travelers within transportation networks, and we demonstrate that the proposed systems can meet critical behavioral criteria for decision-making and learning behaviors using related studies and a demonstrative example of LLM agents' learning and adjustment in the bottleneck setting. Although further refinement of the LLM-agent-based modeling framework is necessary, we believe that this approach has the potential to improve transportation system modeling and simulation.

This study examines the change in activities and associated travel during the beginning of COVID-19 pandemic in Indonesia. This study is particularly interested in analyzing the role of attitudes, descriptive norms, protective behaviors toward COVID-19, travel frequency before the pandemic, and spatial and individual characteristics on activity-travel behavior changes in relation to information and communication technology (ICT) use. Data were obtained from 1062 respondents using a web-based questionnaire survey. Structural equation modeling was used to examine the complex relationships among variables. This study found that descriptive norms positively affected the frequency of travel during the COVID-19 pandemic. Teleworking and e-learning and attitudes toward COVID-19 directly affected activity-travel behavior changes. On the contrary, teleshopping did not contribute to reducing out-of-home activities during the COVID-19 pandemic. Experience of ICT influenced a decline in travel frequency and ride-hailing use. Furthermore, although personal attributes insignificantly influenced activity-travel behavior change, these attributes directly affected ICT use. Meanwhile, people living outside of Java Island had a higher travel frequency during the beginning of COVID-19 pandemic than their counterparts. Based on our findings, this study recommends that the very initial step in an emergency caused by a disaster be to massively socialize or educate people about the risk of the pandemic and to continue with a policy to minimize travel by encouraging teleworking and e-learning. Empowering ICT to support activities from home will beneficially minimize the spread of the pandemic.

In recent years, researchers have made notable advancements in various disciplines using large-scale foundation models. However, foundation models in the transportation system have not received adequate attention. To address this gap, we propose the Generative Graph Transformer (GGT), a transportation foundation model (TFM) that leverages graph structure and dynamic graph generation algorithms. The primary objective of our TFM is to capture participant behavior and interaction in the transportation system, at various scales, and establish a large-scale neural network to comprehend the entire system. The GGT-based TFM can overcom challenges of structural complexity and model accuracy in conventional traffic models. This approach holds promise for addressing complex traffic issues by utilizing up-to-date real traffic data. To demonstrate the capabilities of GGT, a simulation experiment was conducted.

Traffic simulation is a crucial tool for transportation decision-making and policy development. However, achieving realistic simulations in the face of the high dimensionality and heterogeneity of traffic environments is a longstanding challenge. In this paper, we present TransWordNG, a traffic simulator that uses Data-driven algorithms and Graph Computing techniques to learn traffic dynamics from real data. The functionality and structure of TransWorldNG are introduced, which utilize a foundation model for transportation management and control. The results demonstrate that TransWorldNG can generate more realistic traffic patterns compared to traditional simulators. Additionally, TransWorldNG exhibits better scalability, as it shows linear growth in computation time as the scenario scale increases. To the best of our knowledge, this is the first traffic simulator that can automatically learn traffic patterns from real-world data and efficiently generate accurate and realistic traffic environments.

Traditional agent-based urban mobility simulations often rely on rigid rulebased systems that struggle to capture the complexity, adaptability, and behavioral diversity inherent in human travel decision making. Inspired by recent advancements in large language models and AI agent technologies, we introduce GATSim, a novel framework that leverages these advancements to simulate urban mobility using generative agents with dedicated cognitive structures. GATSim agents are characterized by diverse socioeconomic profiles, individual lifestyles, and evolving preferences shaped through psychologically informed memory systems and lifelong learning. The main contributions of this work are: 1) a comprehensive architecture that integrates urban mobility foundation model with agent cognitive systems and transport simulation environment; 2) a hierarchical memory designed for efficient retrieval of contextually relevant information, incorporating spatial and temporal associations; 3) planning and reactive mechanisms for modeling adaptive mobility behaviors which integrate a multi-scale reflection process to transform specific travel experiences into generalized behavioral insights. Experiments indicate that generative agents perform competitively with human annotators in role-playing scenarios, while naturally producing realistic macroscopic traffic patterns. The code for the prototype implementation is publicly available at https://github.com/qiliuchn/gatsim.

No abstract available

Intelligent and efficient energy supply management lays an essential foundation for urban governance and electric vehicle (EV) industry. Specifically, battery swapping is a novel mode of power supply for EVs. However, the new way of energy supply complicates the action policies of EVs, especially when the number of power supply facilities is limited. To address this issue, this paper proposes a multilevel deep reinforcement learning (DRL) method to coordinate the action of EVs within the battery charging and swapping station (BCSS) environment. Firstly, an action-driven simulation framework is developed to simulate the BCSS environment and obtain the EVs’ attributes. Then the multilevel algorithm is proposed to drive the EVs to obtain charging strategies. In the multilevel algorithm, the initial decision for EVs is provided by a DRL-based model. Then the advantage value function is utilized to adjust the initial decision of EVs to meet the constraints of limited charging and swapping equipment. Besides, unlike traditional DRL-based methods, the proposed model is driven by the rewards obtained from EV actions. Finally, extensive experiments have shown that the proposed multilevel DRL-based method has superior performance over existing methods in resolving coordinated battery charging and swapping actions. In particular, the proposed model can provide a suggested and reasonable price range for the practical battery swapping mode operation.

This paper proposes a reinforcement learning–based approach toward intelligent supply chains capable of adaptive and data-driven decision-making in dynamic environments. Traditional supply chain optimization methods often rely on static assumptions and predefined rules, which limit their effectiveness under uncertainty. In contrast, the proposed framework models supply chain operations as a sequential decision-making process, allowing an agent to learn optimal policies through continuous interaction with the environment. Key operational components, including inventory control, transportation planning, and demand fulfillment, are integrated into a unified reinforcement learning model. Simulation-based experiments demonstrate that the proposed approach outperforms conventional optimization and rule-based methods in terms of total operational cost, service level, and system adaptability. The results indicate that reinforcement learning provides a promising foundation for building intelligent supply chains that can autonomously respond to changing operational conditions. Each supply chain entity, such as suppliers, warehouses, and distributors, is modeled as an autonomous learning agent that interacts with other agents and the shared environment. Through cooperative learning, the agents gradually develop coordinated policies that improve overall system performance. The proposed approach addresses the limitations of centralized decision-making by enabling decentralized yet coordinated control. The learning-based framework enables supply chain systems to adapt decisions dynamically without explicit mathematical modeling of uncertainties. Experimental results obtained from simulated logistics scenarios show that deep reinforcement learning significantly improves decision quality compared to traditional heuristics, particularly in volatile environments.

To address the inefficiency and delayed response in multi-agent coordination during vessel entry and departure scheduling, this study proposes an Extended Critical Path Method (Extended-CPM)–based optimization model. From the perspective of time optimization, the model abstracts the entire port operation process into a multi-agent activity network, integrating sequential, parallel, resource, and institutional constraints. A coordination coefficient and a compressibility parameter are introduced to quantify process performance under different levels of information sharing. Parameterized simulation results indicate that as coordination levels increase, total process duration can be reduced by approximately 15–25%, with bottlenecks in approval and dispatching significantly alleviated and the critical path shifting toward berthing operations. The results demonstrate that the proposed model effectively captures the impact of coordination mechanisms on scheduling efficiency and provides a theoretical foundation for process optimization under integrated port–shipping operations.

The spatial differentiation of land use induces traffic demand and guides the construction of traffic supply; traffic conditions are an important influencing factor in determining the nature of land use, and there is a close interaction between the two. This study uses a neural network-based approach at the urban grid level to portray representative phenomena of urban development and analyze the interaction between transportation and land use. The results reflect the model's effective simulation of urban laws, and the case study reveals the differences in the laws of different cities, to guide the benign development of cities and transportation. This article firstly conducts a study on the theoretical foundation; compares the development history, planning, and design methods and practical experience of road planning and resilient planning; summarizes the experience of resilient road system design; and analyzes the future development trend, based on the above basic theoretical research, to develop research ideas and methods. Secondly, the scenario analysis method is explicitly applied to analyze various scenarios that may occur in the future development process of simulated urban roads and rank the scenarios based on the probability of occurrence. For the impact of traffic on land use, the concepts of vitality and potential are introduced, and a multidimensional long and short-term memory network (MDLSTM) model is established. The model takes into account land use lags and potential transfer and has relatively higher prediction accuracy. The results show that larger cities with urban dominant industries and tertiary industries also have higher land use potential and the more significantly influenced by traffic.

This research aims to model system dynamics for mixed traffic flow consisting of Connected and Automated Vehicles (CAVs) and Human-driven Vehicles (HVs). It quantifies the impact of CAVs’ speed change on the overall traffic state on a real-time basis. The model describes the impedance of CAVs’ speed reduction on traffic flow and considers the impact of potential additional lane change induced by the speed reduction. To validate the effectiveness of the proposed model, a VISSIM based microscopic simulation evaluation is performed. The results confirm that the accuracy of the proposed model is generally over 80% with the CAVs’ speed reduction constrained within 20 km/h. Sensitivity analysis is conducted in terms of various CAV penetration rates and congestion levels. The proposed model demonstrates consistently good performance across all CAV penetration rates and congestion levels. A showcase is presented to show the effect of the system dynamics in active traffic management. The proposed model could serve as the foundation of CAV based traffic management applications, such as variable speed limit and speed harmonization.

This study tackles the inefficiencies of the prevalent electronic toll collection (ETC) systems on expressways, which are hindered by limited communication ranges and notable processing delays that significantly reduce actual transportation throughput compared to theoretical capacities. To address these issues, we have developed an innovative free-flow tolling system that integrates cutting-edge 5G communication and high-precision localization technologies, aiming to streamline the tolling process and boost efficiency.We constructed a comprehensive traffic flow model that factors in diverse vehicle classifications and distinct driving patterns. This model serves as the foundation for simulating the tolling operations and evaluating the performance of the proposed system. Employing a cellular automaton framework, our simulation study meticulously assesses the service level of the tolling system, enabling us to fine-tune the design of tolling channels and enhance traffic organization schemes.The findings indicate that the proposed system has the potential to significantly improve traffic flow, minimize congestion, and elevate the overall efficiency of expressway transportation networks. Our research provides robust theoretical support and practical insights for the deployment of future expressway free-flow tolling systems.

No abstract available

No abstract available

Connected and automated vehicle (CAV) technologies need comprehensive testing and evaluation before actual implementation in the real world. However, many inherent technical challenges exist due to the complexity of CAVs. An integrated evaluation platform is needed with vehicle and traffic simulation tools from different domains and X-in-the-loop (XIL) components to fully evaluate all aspects of CAV technologies. In this work, a multi-resolution XIL simulation framework named Real-Sim is presented to support inclusive testing and evaluation of CAVs. Real-Sim approach is simply defined as nearly any part of a system can be "in the loop", either physically or virtually. Oak Ridge National Laboratory (ORNL) has adapted Real-Sim into all its XIL capable laboratories, as well as much of its simulation and model-based design. The foundation of this approach is the system-of-systems concept which has become more relevant as much of transportation research has expanded beyond the vehicle into the traffic networks and traffic control devices. Real-Sim allows researchers to bring real, tangible hardware and software into simulated environments. It applies to a wide variety of applications, including Real controller – Simulated vehicle, Real vehicle – Simulated virtual environment, Real signal controllers – Simulated traffic, etc.

No abstract available

No abstract available

This study explores how the integration of generative artificial intelligence, multi-objective evolutionary optimization, and reinforcement learning can enable sustainable and cost-effective decision-making in supply chain strategy. Using real-world retail demand data enriched with synthetic sustainability attributes, we trained a Variational Autoencoder (VAE) to generate plausible future demand scenarios. These were used to seed a Non-Dominated Sorting Genetic Algorithm (NSGA-II) aimed at identifying Pareto-optimal sourcing strategies that balance delivery cost and CO2 emissions. The resulting Pareto frontier revealed favorable trade-offs, enabling up to 50% emission reductions for only a 10–15% cost increase. We further deployed a deep Q-learning (DQN) agent to dynamically manage weekly shipments under a selected balanced strategy. The reinforcement learning policy achieved an additional 10% emission reduction by adaptively switching between green and conventional transport modes in response to demand and carbon pricing. Importantly, the agent also demonstrated resilience during simulated supply disruptions by rerouting decisions in real time. This research contributes a novel AI-based decision architecture that combines generative modeling, evolutionary search, and adaptive control to support sustainability in complex and uncertain supply chains.

Generative AI can be used in multi-actor environments with purposes ranging from social science modeling to interactive narrative and AI evaluation. Supporting this diversity of use cases -- which we classify as Simulationist, Dramatist, and Evaluationist -- demands a flexible scenario definition framework. We argue here that a good approach is to take inspiration from tabletop role-playing games (TTRPGs), where a Game Master (GM) is responsible for the environment and generates all parts of the story not directly determined by the voluntary actions of player characters. We argue that the Entity-Component architectural pattern is useful here. In such a system, the GM is not a hardcoded computer game but is itself a configurable entity, composed of components just like any other actor. By design, the approach allows for a separation between the underlying implementation details handled by an engineer, the creation of reusable components, and their composition and configuration managed by a designer who constructs entities from the components. This separation of concerns is instrumental for achieving rapid iteration, maintaining modularity, and ultimately to ensure scalability. We describe the ongoing evolution of the Concordia library in terms of this philosophy, demonstrating how it allows users to effectively configure scenarios that align with their specific goals.

The subway power supply system, as a critical component of urban rail transit infrastructure, plays a pivotal role in ensuring operational efficiency and safety. However, current systems remain heavily dependent on manual interventions for fault diagnosis and recovery, limiting their ability to meet the growing demand for automation and efficiency in modern urban environments. While the concept of “self-healing” has been successfully implemented in power grids and distribution networks, adapting these technologies to subway power systems presents distinct challenges. This review introduces an innovative approach by integrating multi-agent systems (MASs) with advanced artificial intelligence (AI) algorithms, focusing on their potential to create fully autonomous self-healing control architectures for subway power networks. The novel contribution of this review lies in its hybrid model, which combines MASs with the IEC 61850 communication standard to develop fault diagnosis, isolation, and recovery mechanisms specifically tailored for subway systems. Unlike traditional methods, which rely on centralized control, the proposed approach leverages distributed decision-making capabilities within MASs, enhancing fault detection accuracy, speed, and system resilience. Through a thorough review of the state of the art in self-healing technologies, this work demonstrates the unique benefits of applying MASs and AI to address the specific challenges of subway power systems, offering significant advancement over existing methodologies in the field.

The integration of emerging uncrewed aerial vehicle (UAV) with artificial intelligence (AI) and ground-embedded robots (GERs) has transformed emergency rescue operations in unknown environments. However, the high computational demands of such missions often exceed the capacity of a single UAV, making it difficult for the system to continuously and stably provide high-level services. To address these challenges, this paper proposes a novel cooperation framework involving UAVs, GERs, and airships. This framework enables resource pooling through UAV-to-GER (U2G) and UAV-to-airship (U2A) communications, providing computing services for UAV offloaded tasks. Specifically, we formulate the multi-objective optimization problem of task assignment and exploration optimization in UAVs as a dynamic long-term optimization problem. Our objective is to minimize task completion time and energy consumption while ensuring system stability over time. To achieve this, we first employ the Lyapunov optimization method to transform the original problem, with stability constraints, into a per-slot deterministic problem. We then propose an algorithm named HG-MADDPG, which combines the Hungarian algorithm with a generative diffusion model (GDM)-based multi-agent deep deterministic policy gradient (MADDPG) approach, to jointly optimize exploration and task assignment decisions. In HG-MADDPG, we first introduce the Hungarian algorithm as a method for exploration area selection, enhancing UAV efficiency in interacting with the environment. We then innovatively integrate the GDM and multi-agent deep deterministic policy gradient (MADDPG) to optimize task assignment decisions, such as task offloading and resource allocation. Simulation results demonstrate the effectiveness of the proposed approach, with significant improvements in task offloading efficiency, latency reduction, and system stability compared to baseline methods.

Multi-Agent Path Finding (MAPF) is the challenging problem of finding collision-free paths for multiple agents, which has a wide range of applications, such as automated warehouses, smart manufacturing, and traffic management. Recently, machine learning-based approaches have become popular in addressing MAPF problems in a decentralized and potentially generalizing way. Most learning-based MAPF approaches use reinforcement and imitation learning to train agent policies for decentralized execution under partial observability. However, current state-of-the-art approaches suffer from a prevalent bias to micro-aspects of particular MAPF problems, such as congestions in corridors and potential delays caused by single agents, leading to tight specializations through extensive engineering via oversized models, reward shaping, path finding algorithms, and communication. These specializations are generally detrimental to the sample efficiency, i.e., the learning progress given a certain amount of experience, and generalization to previously unseen scenarios. In contrast, curriculum learning offers an elegant and much simpler way of training agent policies in a step-by-step manner to master all aspects implicitly without extensive engineering. In this paper, we propose a generative curriculum approach to learning-based MAPF using Variational Autoencoder Utilized Learning of Terrains (VAULT). We introduce a two-stage framework to (I) train the VAULT via unsupervised learning to obtain a latent space representation of maps and (II) use the VAULT to generate curricula in order to improve sample efficiency and generalization of learning-based MAPF methods. For the second stage, we propose a bi-level curriculum scheme by combining our VAULT curriculum with a low-level curriculum method to improve sample efficiency further. Our framework is designed in a modular and general way, where each proposed component serves its purpose in a black-box manner without considering specific micro-aspects of the underlying problem. We empirically evaluate our approach in maps of the public MAPF benchmark set as well as novel artificial maps generated with the VAULT. Our results demonstrate the effectiveness of the VAULT as a map generator and our VAULT curriculum in improving sample efficiency and generalization of learning-based MAPF methods compared to alternative approaches. We also demonstrate how data pruning can further reduce the dependence on available maps without affecting the generalization potential of our approach.

We discuss the emerging new opportunity for building feedback‐rich computational models of social systems using generative artificial intelligence. Referred to as generative agent‐based models (GABMs), such individual‐level models utilize large language models to represent human decision‐making in social settings. We provide a GABM case in which human behavior can be incorporated into simulation models by coupling a mechanistic model of human interactions with a pre‐trained large language model. This is achieved by introducing a simple GABM of social norm diffusion in an organization. For educational purposes, the model is intentionally kept simple. We examine a wide range of scenarios and the sensitivity of the results to several changes in the prompt. We hope the article and the model serve as a guide for building useful dynamic models of various social systems that include realistic human reasoning and decision‐making. © 2024 System Dynamics Society.

Generative Artificial Intelligence (AI) is boosting anticipatory forms of governance, through which state actors seek to predict the future and strategically intervene in the present. In this context, city brains represent an emerging type of generative AI currently employed in urban governance and public policy in a growing number of cities. City brains are large-scale AIs residing in vast digital urban platforms, which manage multiple urban domains including transport, safety, health, and environmental monitoring. They use Large Language Models (LLMs) to generate visions of urban futures: visions that are in turn used by policymakers to generate new urban policies. In this paper, we advance a twofold contribution. Theoretically, we develop a critical theory of anticipatory governance in the age of generative AI. More specifically, we focus on technocratic approaches to anticipatory governance, to explain how the act of governing extends into the future by means of predictive AI technology. Our approach is critical in order to expose the dangers that the use of AI (generative AI, in particular) in urban governance poses, and to identify their causes. These dangers include the formation of a policy process that, under the influence of unintelligible LLMs, risks losing transparency and thus accountability, and the marginalization of human stakeholders (citizens, in particular) as the role of AI in the management of cities keeps growing and governance begins to turn posthuman. Empirically, we critically examine an existing city brain project under development in China and ground our critical theory in a real-life example.

Agent-based models have gained wide acceptance in transportation planning because with increasing computational power, large-scale people-centric mobility simulations are possible. Several modeling efforts have been reported in the literature on the demand side (with sophisticated activity-based models that focus on an individual’s day activity patterns) and on the supply side (with detailed representation of network dynamics through simulation-based dynamic traffic assignment models). This paper proposes an extension to a state-of-the-art integrated agent-based demand and supply model—SimMobility—for the design and evaluation of autonomous vehicle systems. SimMobility integrates various mobility-sensitive behavioral models in a multiple time-scale structure comprising three simulation levels: (a) a long-term level that captures land use and economic activity, with special emphasis on accessibility; (b) a midterm level that handles agents’ activities and travel patterns; and (c) a short-term level that simulates movement of agents, operational systems, and decisions at a microscopic granularity. In that context, this paper proposes several extensions at the short-term and midterm levels to model and simulate autonomous vehicle systems and their effects on travel behavior. To showcase these features, the first-cut results of a hypothetical on-demand service with autonomous vehicles in a car-restricted zone of Singapore are presented. SimMobility was successfully used in an integrated manner to test and assess the performance of different autonomous vehicle fleet sizes and parking station configurations and to uncover changes in individual mobility patterns, specifically in regard to modal shares, routes, and destinations.

No abstract available

Autonomous Driving Systems (ADS) are advancing rapidly due to progress in deep learning, yet critical challenges remain, particularly in the realm of safety verification. As safety-critical systems, ADS must undergo rigorous testing across diverse scenarios. Real-world data, while valuable, are inherently inflexible for interaction and scenario customization. In contrast, simulator-generated synthetic scenarios provide a platform that enables interaction, control, editability, and adaptability to specific needs. However, current simulation approaches are limited—either relying on costly, manually crafted, overly templated scenarios or generating unconditioned trivial behaviors based on learned distributions. In this work, we introduce LASER, an innovative framework that leverages large language models (LLMs) to conduct traffic simulations based on natural language inputs. The framework operates in two phases. First, it generates scripts from user-provided descriptions. Second, it executes these scripts by guiding autonomous agents within the CARLA simulator to perform tasks in real-time. This method effectively decomposes tasks, allocates controls, and integrates interactive elements to create dynamic and scalable simulations that align with user requirements. By using LASER, we overcome the rigid constraints of traditional simulation methods, enabling the creation of complex, diverse, flexible and on-demand driving scenarios. The approach significantly enhances the process of generating ADS training and testing data, addressing the scalability and diversity issues associated with previous simulation models. The code and all demos are available anonymously at https://njudeepengine.github.io/LASER/.

No abstract available

Testing and validating Autonomous Vehicle (AV) performance in safety-critical and diverse scenarios is crucial before real-world deployment. However, manually creating such scenarios in simulation remains a significant and time-consuming challenge. This work introduces a novel method that generates dynamic temporal scene graphs corresponding to diverse traffic scenarios, on-demand, tailored to user-defined preferences, such as AV actions, sets of dynamic agents, and criticality levels. A temporal Graph Neural Network (GNN) model learns to predict relationships between ego-vehicle, agents, and static structures, guided by real-world spatiotemporal interaction patterns and constrained by an ontology that restricts predictions to semantically valid links. Our model consistently outperforms the baselines in accurately generating links corresponding to the requested scenarios. We render the predicted scenarios in simulation to further demonstrate their effectiveness as testing environments for AV agents.

This article presents an agent-based travel demand model, where agents react to transport supply across all mobility choices. Long-term choices include mobility tool ownership and work locations. Daily travel patterns are simulated at the individual level by sequentially combining activity frequency, activity durations and destinations as well as a rule-based time-of-day scheduling.  A key to success in this novel approach is balancing individual preferences of travelers with system constraints. The model incorporates two types of constraints: 1) capacity constraints of the transport infrastructure. 2) natural time and space constraints during the execution of individual 24-hour day plans. Model results are validated against empirical observations of travel demand in Switzerland. The article concludes with a perspective for further research and development in the field of applied agent-based modeling.

Abstract Shared Autonomous Vehicles (SAV), or robotaxis, are expected to be commercially available within this decade. This new transport mode has the potential to revolutionize travel, offering a level of service comparable to traditional taxis with much lower prices. This may attract travelers currently using other modes, impacting the economic sustainability of public transport as well as car ownership levels. We investigate this potential demand using a scalable SAV simulation framework. We do not establish a future equilibrium considering the interaction between all users on a detailed road network, but establish the potential demand for a large metropolitan area. Travelers can choose between their current mode and the new SAV mode, with fare and waiting times which depend on real-time demand. For our input data we train a statistical model on a large transport survey from Germany for an urban region, allowing us to generate a large number of trips with realistic characteristics. We conduct a sensitivity analysis to study the effect of several key parameters on the modal shift. We find that SAVs can be attractive to many active mode and public transport users unless regulations are put in place. Our results also show that due to SAV fleet constraints, changes in incentives for travelers currently using cars may have significant consequences on the behavior of other travelers. We further calculate key economic indicators for the fleet, which can inform the discussion on the fleet size and fare level that operators are likely to choose when maximizing their own profit.

Autonomous mobility-on-demand (AMoD) systems, powered by advances in robotics, control, and machine learning (ML), offer a promising paradigm for future urban transportation. AMoD offers fast and personalized travel services by leveraging centralized control of autonomous vehicle fleets to optimize operations and enhance service performance. However, the rapid growth of this field has outpaced the development of standardized practices for evaluating and reporting results, leading to significant challenges in reproducibility. As AMoD control algorithms become increasingly complex and data-driven, a lack of transparency in modeling assumptions, experimental setups, and algorithmic implementation hinders scientific progress and undermines confidence in the results. This article presents a systematic study of reproducibility in AMoD research. We identify key components across the research pipeline, spanning system modeling, control problems, simulation design, algorithm specification, and evaluation, and analyze common sources of irreproducibility. We survey prevalent practices in the literature, highlight gaps, and propose a structured framework to assess and improve reproducibility. While focused on AMoD, the principles and practices we advocate generalize to a broader class of cyber-physical systems that rely on networked autonomy and data-driven control. This work aims to lay the foundation for a more transparent and reproducible research culture in the design and deployment of intelligent mobility systems.

Shared autonomous vehicles (SAVs) are a fleet of autonomous taxis that provide point-to-point transportation services for travellers, and have the potential to reshape the nature of the transportation market in terms of operational costs, environmental outcomes, increased tolling efficiency, etc. However, the number of waiting passengers could become arbitrarily large when the fleet size is too small for travel demand, which could cause an unstable network. An unstable network will make passengers impatient and some people will choose some other alternative travel modes, such as metro or bus. To achieve stable and reliable SAV services, this study designs a dynamic queueing model for waiting passengers and provides a fast maximum stability dispatch policy for SAVs when the average number of waiting for passengers is bounded in expectation, which is analytically proven by the Lyapunov drift techniques. After that, we expand the stability proof to a more realistic scenario accounting for the existence of exiting passengers. Unlike previous work, this study considers exiting passengers in stability analyses for the first time. Moreover, the maximum stability of the network doesn't require a planning horizon based on the proposed dispatch policy. The simulation results show that the proposed dispatch policy can ensure the waiting queues and the number of exiting passengers remain bound in several experimental settings.

The First Mile–Last Mile (FMLM) connectivity is crucial for improving public transit accessibility and efficiency, particularly in sprawling suburban regions where traditional fixed-route transit systems are often inadequate. Autonomous on-Demand Shuttles (AODS) hold promise as an option for FMLM connections because of their cost-effectiveness and improved safety features, thereby enhancing user convenience and reducing reliance on personal vehicles. A critical issue in AODS service design is the optimization of travel paths, for which realistic traffic network assignment combined with optimal routing offers a viable solution. In this study, we have designed an AODS controller that integrates a mesoscopic simulation-based Dijkstra shortest path algorithm with a greedy and exhaustive insertion heuristics approach to optimize the travel routes of the shuttles. The controller also considers the charging infrastructure/strategies and the impact of the shuttles on regular traffic flow for routes and fleet-size planning. The controller is implemented in the Aimsun Next traffic simulator considering Lake Nona in Orlando, Florida as a case study. We show that, under the present demand based on 1% of total trips as transit riders, a fleet of three autonomous shuttles can serve about 80% of FMLM trip requests on an on-demand basis with an average waiting time below 4 min. Additional power sources have a significant effect on service quality as the inactive waiting time for charging would increase the fleet size. We also show that low-speed autonomous shuttles would have negligible impact on regular vehicle flow, making them suitable for suburban areas. These findings have important implications for sustainable urban planning and public transit operations.

Multi-agent coordination is critical for next-generation autonomous vehicle (AV) systems, yet naive implementations of communication-based rerouting can lead to catastrophic performance degradation. This study investigates a fundamental problem in decentralized multi-agent navigation: routing loops, where vehicles without persistent obstacle memory become trapped in cycles of inefficient path recalculation. Through systematic simulation experiments involving 72 unique configurations across varying vehicle densities (15, 35, 55 vehicles) and obstacle frequencies (6, 20 obstacles), we demonstrate that memory-less reactive rerouting increases average travel time by up to 682% compared to baseline conditions. To address this, we introduce Object Memory Management (OMM), a lightweight mechanism enabling agents to retain and share knowledge of previously encountered obstacles. OMM operates by maintaining a distributed blacklist of blocked nodes, which each agent consults during Dijkstra-based path recalculation, effectively preventing redundant routing attempts. Our results show that OMM-enabled coordination reduces average travel time by 75.7% and wait time by 88% compared to memory-less systems, while requiring only 1.67 route recalculations per vehicle versus 9.83 in memory-less scenarios. This work provides empirical evidence that persistent, shared memory is not merely beneficial but essential for robust multi-agent coordination in dynamic environments. The findings have implications beyond autonomous vehicles, informing the design of decentralized systems in robotics, network routing, and distributed AI. We provide a comprehensive experimental analysis, including detailed scenario breakdowns, scalability assessments, and visual documentation of the routing loop phenomenon, demonstrating OMM's critical role in preventing detrimental feedback cycles in cooperative multi-agent systems.

As the populations continue to age across many nations, ensuring accessible and efficient transportation options for older adults has become an increasingly important concern. Autonomous Mobility-on-Demand (AMoD) systems have emerged as a potential solution to address the needs faced by older adults in their daily mobility. However, estimation of older adult mobility needs, and how they vary over space and time, is crucial for effective planning and implementation of such service, and conventional four-step approaches lack the granularity to fully account for these needs. To address this challenge, we propose an agent-based model of older adults mobility demand in Winnipeg, Canada. The model is built for 2022 using primarily open data, and is implemented in the Multi-Agent Transport Simulation toolkit. After calibration to accurately reproduce observed travel behaviours, a new AMoD service is tested in simulation and its potential adoption among Winnipeg older adults is explored. The model can help policy makers to estimate the needs of the elderly populations for door-to-door transportation and can guide the design of AMoD transport systems.

Traffic simulation is important for transportation optimization and policy making. While existing simulators such as SUMO and MATSim offer fully-featured platforms and utilities, users without too much knowledge about these platforms often face significant challenges when conducting experiments from scratch and applying them to their daily work. To solve this challenge, we propose TrafficSimAgent, an LLM-based agent framework that serves as an expert in experiment design and decision optimization for general-purpose traffic simulation tasks. The framework facilitates execution through cross-level collaboration among expert agents: high-level expert agents comprehend natural language instructions with high flexibility, plan the overall experiment workflow, and invoke corresponding MCP-compatible tools on demand; meanwhile, low-level expert agents select optimal action plans for fundamental elements based on real-time traffic conditions. Extensive experiments across multiple scenarios show that TrafficSimAgent effectively executes simulations under various conditions and consistently produces reasonable outcomes even when user instructions are ambiguous. Besides, the carefully designed expert-level autonomous decision-driven optimization in TrafficSimAgent yields superior performance when compared with other systems and SOTA LLM based methods.

This study investigates the implementation of semi-on-demand (SoD) hybrid-route services using Shared Autonomous Vehicles (SAVs) on existing transit lines. SoD services combine the cost efficiency of fixed-route buses with the flexibility of on-demand services. SAVs first serve all scheduled fixed-route stops, then drop off and pick up passengers in the pre-determined flexible-route portion, and return to the fixed route. This study addresses four key questions: optimal fleet and vehicle sizes for peak-hour fixed-route services with SAVs and during transition (from drivers to autonomous vehicles), optimal off-peak SoD service planning, and suitable use cases. The methodology combines analytical modeling for service planning with agent-based simulation for operational analysis. We examine ten bus routes in Munich, Germany, considering full SAV and transition scenarios with varying proportions of drivers. Our findings demonstrate that the lower operating costs of SAVs improve service quality through increased frequency and smaller vehicles, even in transition scenarios. The reduced headway lowers waiting time and also favors more flexible-route operation in SoD services. The optimal SoD settings range from fully flexible to hybrid routes, where higher occupancy from the terminus favors shorter flexible routes. During the transition phase, limited fleet size and higher headways constrain the benefits of flexible-route operations. The simulation results corroborate the SoD benefits of door-to-door convenience, attracting more passengers without excessive detours and operator costs at moderate flexible-route lengths, and validate the analytical model.

The way in which autonomous transport will be adopted is likely to determine the net social benefits delivered by the technology and the sustainability of the transport system. Autonomous vehicles (AVs) will change travel behavior due to reduction in the effort needed for humans to drive a vehicle, the need for them to find a parking space, and the costs related to vehicle operation. The AVs’ benefits are likely to increase their adoption compared to conventional human-driven vehicles, possibly leading to more vehicle kilometers travelled (VKT) and consequently weakening their benefits in large cities particularly if they are used in competition with public transport (PT). This paper evaluates the interplay between AVs and PT, and how road network pricing can be used to influence behavioral changes when personal autonomous vehicles (PAVs) are highly available. An agent-based demand model framework is proposed to estimate the mode share of PAVs and PT based on their perceived travelling costs on a real transport network in Melbourne, Australia. The modelling results suggest that convenient and affordable PAVs could compete with traditional PT and reduce overall PT patronage by up to 10%. However, through considering road network pricing schemes, the role of PAVs could be shifted from competing with PT to a complementary first-and-last-mile service that increases PT share by almost 17%. The results also show that road pricing policies can be used as effective interventions to manage PAV operations by reducing empty vehicle trips by 20%.

Operational Design Domain (ODD) refers to the specific condition set under which an autonomous driving system is designed to operate. Currently, ODD deployment primarily focuses on safety factors, such as lighting and road surface conditions. However, the impact of ODD deployment on network efficiency is equally significant. To efficiently optimize ODD deployment in terms of network efficiency, we propose a surrogate model that incorporates multi-user structural information and integrate it into an efficient simulation-based optimization framework. The proposed surrogate model can accurately estimate the complex simulation model and construct a hyperplane containing the efficient descent direction for simulation-based optimization. And we compare two common deployment forms, namely, Dedicated Zone (DZ) and Shared Zone (SZ). The results reveal that DZ and SZ have different deployment tendencies. And DZ facilitates the spatial reallocation of travel demand, effectively alleviating congestion, while SZ reduces the temporal continuity of congestion, thus accelerating traffic flow.

This study introduces a framework to maximize societal benefits associated with the autonomous vehicle (AV)‐dedicated lane implementation at large‐scale transportation networks, considering the travel time savings and the required investments to prepare the infrastructure for their deployment. To this end, a bi‐level optimization problem is formulated. The upper level determines the links for dedicated lane deployment, while at the lower level, a mesoscopic traffic simulation tool is employed to enable a realistic representation of these vehicles in a mixed traffic. The problem is solved using the genetic algorithm. To further reduce the computational burden, this study adopts a clustering method based on the snake algorithm to group the candidate links and reduce the size of the solution space. The proposed framework is successfully applied to the case study of Chicago downtown network, considering various demand levels, AV market penetration rates, and implementation approaches. The results highlight the need for optimizing the placement of AV‐dedicated lanes (AVDLs) to ensure the economically beneficial adoption of this strategy across different scenarios. This study provides transportation planners with key operational insights to facilitate the effective adoption of AVDLs during the transitional phase from human‐driven vehicles to a fully AV environment.

The electrification and automation of mobility are reshaping how cities operate on-demand transport systems. Managing Electric Autonomous Mobility-on-Demand (EAMoD) fleets effectively requires coordinating dispatch, rebalancing, and charging decisions under multiple uncertainties, including travel demand, travel time, energy consumption, and charger availability. We address this challenge with a combined stochastic and robust model predictive control (MPC) framework. The framework integrates spatio-temporal Bayesian neural network forecasts with a multi-stage stochastic optimization model, formulated as a large-scale mixed-integer linear program. To ensure real-time applicability, we develop a tailored Nested Benders Decomposition that exploits the scenario tree structure and enables efficient parallelized solution. Stochastic optimization is employed to anticipate demand and infrastructure variability, while robust constraints on energy consumption and travel times safeguard feasibility under worst-case realizations. We evaluate the framework using high-fidelity simulations of San Francisco and Chicago. Compared with deterministic, reactive, and robust baselines, the combined stochastic and robust approach reduces median passenger waiting times by up to 36% and 95th-percentile delays by nearly 20%, while also lowering rebalancing distance by 27% and electricity costs by more than 35%. We also conduct a sensitivity analysis of battery size and vehicle efficiency, finding that energy-efficient vehicles maintain stable performance even with small batteries, whereas less efficient vehicles require larger batteries and greater infrastructure support. Our results emphasize the importance of jointly optimizing predictive control, vehicle capabilities, and infrastructure planning to enable scalable, cost-efficient EAMoD operations.

The significantly higher level of detail of agent-based travel demand models (ABM) compared to aggregate models stands in contrast to their high simulation times. Once fast model responses are necessary, the application of ABMs may pose run time challenges. For such purposes, a condensation of ABMs’ sensitivities and saturation e ff ects regarding travel mode choice is required, which - as shown in this paper - is achieved by a multiple discrete continuous extreme value model (MDCEV). The application case of this method is a serious game which enables decision-makers and citizens low-threshold access to a better understanding of interrelations within the urban mobility system. The mode choice model for the serious game is estimated on the basis of the simulated travel demand patterns of an ABM. The comparison of the model application with the ABM reference results attests a good ability to reproduce the current travel demand, even for di ff erent types of tra ffi c. Also modal shift reactions on policy interventions reveal a high level of consistency

The emergence and rapid development of autonomous driving technology have created new transportation options for travelers, including autonomous vehicles (AVs). However, significant uncertainties persist concerning travelers’ willingness to adopt this emerging transportation mode, despite its potential benefits. This study adopts an agent-based simulation approach to investigate travel mode choice behavior for AVs, accounting for the impact of travel experience and social interaction on traveler preferences. The results demonstrate that both travel experience and social interactions exert significant influences on mode choice decisions. The travel experience aligns travelers’ decisions with a neutral preference outcome, whereas social interaction prompts travelers to select modes similar to their peers. Additionally, social interaction encourages the preference for modes with intermediate utility values. Ultimately, this research offers data support for automotive companies to identify potential AV users and informs policy development for governments. The findings also contribute to understanding the role of social interaction and travel experience in shaping travel mode choice behaviors, bridging the gap between behavioral economics and transportation research. This study provides valuable insights for designing user-centered transportation policies and marketing strategies.

Shared autonomous vehicles (SAVs) combine autonomous driving and sharing mobility, offering potential to reshape future travel modes. Their self-driving attribute is expected to significantly reduce parking demand and alter the parking landscape. Previous studies have primarily focused on the quantity of parking demand, with parking choice based solely on parking prices. The spatial distribution of parking demand remains unclear, especially during the transition period with both conventional vehicles and SAVs mixed on roads. An agent-based simulation model was developed to evaluate the impact of SAVs on parking demand from quantity and spatial distribution perspectives. A generalized cost function considered not only parking prices, but also the road toll and energy fee, was developed to alleviate the negative effect of SAVs’ endless cruising. We also explored the trend of such effects with varying SAVs’ market penetration rates. The results indicated a substantial decrease in parking demand (nearly 80%), spreading from the central business district (CBD) to the periphery, leading to a significant increase in vehicle miles traveled (VMT) within the entire network. We also found that the additional VMT is mainly due to SAVs’ empty travel during providing continuous services. To relieve the poor utilization of parking lots in CBD due to outspreading demand, a parking policy which can dynamically adjust parking price based on its real-time occupancy rate was proposed and verified. Results suggested that it would increase of the average utilization rate of overall parking lots by 8% and reduce the VMT caused by SAVs’ parking by 29.5% if the policy applied with a large pricing change coefficient.

This paper presents the coupling of a state-of-the-art ride-pooling fleet simulation package with the mobiTopp travel demand modeling framework. The coupling of both models enables a detailed agent- and activity-based demand model, in which travelers have the option to use ride-pooling based on real-time offers of an optimized ride-pooling operation. On the one hand, this approach allows the application of detailed mode-choice models based on agent-level attributes coming from mobiTopp functionalities. On the other hand, existing state-of-the-art ride-pooling optimization can be applied to utilize the full potential of ride-pooling. The introduced interface allows mode choice based on real-time fleet information and thereby does not require multiple iterations per simulated day to achieve a balance of ride-pooling demand and supply. The introduced methodology is applied to a case study of an example model where in total approximately 70,000 trips are performed. Simulations with a simplified mode-choice model with varying fleet size (0–150 vehicles), fares, and further fleet operators’ settings show that (i) ride-pooling can be a very attractive alternative to existing modes and (ii) the fare model can affect the mode shifts to ride-pooling. Depending on the scenario, the mode share of ride-pooling is between 7.6% and 16.8% and the average distance-weighed occupancy of the ride-pooling fleet varies between 0.75 and 1.17.

Shared mobility solutions such as bike sharing services play a key role to reduce greenhouse gas emissions in urban areas. In this paper, we present an approach to model station-based bike sharing in the multi-modal agent-based travel demand model mobiTopp. We compare di ff erent implementations of how agents choose their bike pick-up and drop-o ff stations. In addition to two variations of distance minimization, we also present a gravity approach to represent the reliability of a system. By also comparing di ff erent behavioral attitudes of the agents towards walking, a total of six scenarios were implemented and tested. The presented approach allows to easily test scenarios with a varying number of bikes and stations. We apply our algorithm to a model for the city of Hamburg, Germany, where the mobility behavior of a total of 1.9 million agents is modeled. Our simulations show plausible results. The average distances, utilization shares of each station, and other parameters match with values from the actual service. While the di ff erent strategies result in significantly di ff erent access times, and provide further new valuable insights and options for parameterization, di ff erences in resulting demand are small. Overall, this model provides new methods to simulate bike sharing in travel demand models, thus helps to simulate an important mode of transport of the future.

The goal of this project is the development of a large-scale agent-based traffic simulation system for Amsterdam urban area, validated on sensor data and adjusted for decision support in critical situations and for policy making in sustainable city development, emission control and electric car research. In this paper we briefly describe the agent-based simulation workflow and give the details of our data- driven approach for (1) modeling the road network of Amsterdam metropolitan area extended by major national roads, (2) recreating the car owners population distribution from municipality demographic data, (3) modeling the agent activity based on travel survey, and (4) modeling the inflow and outflow boundary conditions based on the traffic sensor data. The models are implemented in scientific Python and MATSim agent-based freeware. Simulation results of 46.5 thousand agents -with travel plans sampled from the model distributions- show that travel demand model is consistent, but should be improved to correspond with sensor data. The next steps in our project are: extensive validation, calibration and testing of large-scale scenarios, including critical events like the major power outage in the Netherlands (doi:10.1016/j.procs.2015.11.039), and modelling emissions and heat islands caused by traffic jams.

No abstract available

No abstract available

Shared mobility solutions improve traffic efficiency in cities. This paper addresses the design of an urban demand-responsive transit system operated by a fleet of autonomous modules as a public transportation service. We introduce a discrete event-based multi-agent simulation to reflect system dynamics at an operational level. Vehicles, customers, and intelligent stops are modeled as software agents. Each entity has its own decision-making and planning component. Vehicles compete for a limited number of order requests, and decentralized stops manage the matching between a customer and a prospective vehicle. Scheduling and routing events take place highly dynamically. A simulation study evaluates a demand-responsive operation as a replacement for a fixed-route bus service. Results show that a total demand of 2.3 million rides can be fulfilled with a fleet of around 43,000 six-seater modules. In comparison with the current bus system, the total energy consumption and the system’s utilization are improved, while transport times and distances result in values below the bus system’s performance and therefore must be optimized in further research.

The growing interest in Automated Mobility on Demand (AMoD) services in passenger transportation necessitates accurate forecasting for successful deployment. However, the paucity of real-world data is a significant challenge. In this study, we present a unique technique for developing a synthetic user population tailored to AMoD car services. We identify possible passengers using selection criteria such as age, gender, activity status, and income, and then utilize a multi-agent simulation tool to define passenger movements within the AMoD service and plan out daily journeys. Additionally, a spatiotemporal analysis reveals use patterns that are well captured by Machine Learning models such as Random Forest, XGBoost, Neural Networks (NN), and Linear Regression (LR). Finally, by estimating spatio-temporal demand for automated cars, our model gives critical insights into the ideal allocation of fleet resources, thereby advancing the progress of AMoD transportation systems.

The goal of this study was to analyze the impact of private autonomous vehicles (PAVs), specifically their near-activity location travel patterns, on vehicle miles traveled (VMT). The study proposes an integrated mode choice and simulation-based parking assignment model, along with an iterative solution approach, to analyze the impacts of PAVs on VMT, mode choice, parking lot usage, and other system performance measures. The dynamic simulation-based parking assignment model determines the parking location choice of each traveler as a function of the spatial–temporal demand for parking from the mode choice model, whereas the multinomial logit mode choice model determines mode splits based on the costs and service quality of each travel mode coming, in part, from the parking assignment model. The paper presents a case study to illustrate the power of the modeling framework. The case study varies the percentage of persons with a private vehicle (PV) who own a PAV versus a private conventional vehicle (PCV). The results indicated that PAV owners traveled an extra 0.11 to 1.51 mi compared with PCV owners on average, and the PV mode share was significantly higher for PAV owners. Therefore, as PCVs are converted into PAVs in the future, the results indicate substantial increases in VMT near activity destinations. However, the results also indicated that adjusting parking fees and redistributing parking lot capacities could reduce VMT. The significant increase in VMT from PAVs implies that planners should develop policies to reduce PAV deadheading miles near activity locations, as the automated era comes closer.

Autonomous mobility-on-demand (AMoD) systems represent a rapidly developing mode of transportation wherein travel requests are dynamically handled by a coordinated fleet of robotic, self-driving vehicles. Given a graph representation of the transportation network - one where, for example, nodes represent areas of the city, and edges the connectivity between them - we argue that the AMoD control problem is naturally cast as a node-wise decision-making problem. In this paper, we propose a deep reinforcement learning framework to control the rebalancing of AMoD systems through graph neural networks. Crucially, we demonstrate that graph neural networks enable reinforcement learning agents to recover behavior policies that are significantly more transferable, generalizable, and scalable than policies learned through other approaches. Empirically, we show how the learned policies exhibit promising zero-shot transfer capabilities when faced with critical portability tasks such as inter-city generalization, service area expansion, and adaptation to potentially complex urban topologies.

In recent years, the rapid growth of on-demand delivery services, especially in food deliveries, has spurred the exploration of innovative mobility solutions. In this context, lightweight autonomous vehicles have emerged as a potential alternative. However, their fleet-level behavior remains largely unexplored. To address this gap, we have developed an agent-based model and an environmental impact study assessing the fleet performance of lightweight autonomous food delivery vehicles. This model explores critical factors such as fleet sizing, service level, operational strategies, and environmental impacts. We have applied this model to a case study in Cambridge, MA, USA, where results indicate that there could be significant environmental benefits in replacing traditional car-based deliveries with shared lightweight autonomous vehicle fleets. Lastly, we introduce an interactive platform that offers a user-friendly means of comprehending the model’s performance and potential trade-offs, which can help inform decision-makers in the evolving landscape of food delivery innovation.

In future smart cities supported by cyber-physical social intelligence, autonomous behavioral decision for vehicular agents is going to become a general demand. Despite much progress achieved in autonomous behavioral decision of vehicular agents, the existing works can just be used in scenarios of short-distance behavioral decision. Naturally, they are not well suitable for long-distance behavioral decision tasks, posing much challenge in realistic cyber-physical environment. To bridge the existing gaps, this article proposes an autonomous behavioral decision framework for vehicular agents using cyber-physical social intelligence. First, it is expected to establish a dynamic planning model with multiple objectives and constraints. This can be embedded into the control unit of a vehicular agent to endow it with proper social intelligence. On this basis, an iterative search algorithm is specifically designed for it to find the optimal solutions from the whole solution space. Finally, two typical situation cases are implemented with use of simulation modeling to display the working architecture of the proposed method. In addition, a universal optimization search algorithm is selected as the baseline to be compared with the proposed method. The comparison results reveal both planning utility and running efficiency of the proposed method.

The increase in the number of vehicles in highly dense areas with a limited road network has become a challenge for main cities worldwide. Recent decades demographic and economic growth has generated a gradual increase in traffic congestion and the commutation times of inhabitants, which directly affects their productivity and quality of life. To counteract these impacts, governments have implemented Travel Demand Management (TDM) that restrict the circulation of private vehicles to reduce traffic congestion, while discouraging their acquisition and promoting the use of public transport. However, due to their coercive characteristics, the effectiveness of TDM to discourage the acquisition of private vehicles and their long-term effect on controlling the growth of the vehicle fleet has been harshly questioned. This research presents the recent scientific methodologies to estimate variations in car fleet growth and shows estimation variables in Medellín city using publicly available databases, as well as one method proper for the selection of the proper evaluation model. These set up scenarios for projecting the vehicle fleet based on the implementation of public policies for demand management of demand such as the restriction due to license plate-based (L&P) use and scrapping. This allows for evaluating the relationship between vehicle age allowed for the circulation of vehicles on the road network, demographic growth, and the trend increase for new vehicles.

The development of autonomous vehicles is dramatically reshaping the transportation landscape, bringing new challenges and opportunities in human-machine interaction. As autonomous vehicles evolve, understanding and responding to human intent becomes significant, and therefore require new ways of human-autonomy teaming. A human digital twin (HDT) is a virtual representation of an individual driver, capturing their preferences, behaviors, and physiological states, enabling machines to better understand and predict human needs and responses. In this paper, we explore how large language models (LLMs), like GPT-4 and LLaMA, together with HDTs are changing the way humans team up with autonomous vehicles. These LLMs help make our conversations with vehicles more natural and intuitive. By pairing them in HDTs, we can get real-time feedback and smarter responses. This combination offers not just easier control but also safer driving experiences. We will break down how this works, why it matters, and what we might expect in the future.

In the rapidly evolving landscape of digital twins (DT) and 6G networks, the integration of large language models (LLMs) presents a novel approach to network management. This paper explores the application of LLMs in managing 6G-empowered DT networks, with a focus on optimizing data retrieval and communication efficiency in smart city scenarios. The proposed framework leverages LLMs for intelligent DT problem analysis and radio resource management (RRM) in fully autonomous way without any manual intervention. Our proposed framework — LINKs, builds up a lazy loading strategy which can minimize transmission delay by selectively retrieving the relevant data. Based on the data retrieval plan, LLMs transform the retrieval task into an numerical optimization problem and utilizing solvers to build an optimal RRM, ensuring efficient communication across the network. Simulation results demonstrate the performance improvements in data planning and network management, highlighting the potential of LLMs to enhance the integration of DT and 6G technologies.

This paper presents a novel design of a multi-agent system framework that applies large language models (LLMs) to automate the parametrization of simulation models in digital twins. This framework features specialized LLM agents tasked with observing, reasoning, decision-making, and summarizing, enabling them to dynamically interact with digital twin simulations to explore parametrization possibilities and determine feasible parameter settings to achieve an obj ective. The proposed approach enhances the usability of simulation model by infusing it with knowledge heuristics from LLM and enables autonomous search for feasible parametrization to solve a user task. Furthermore, the system has the potential to increase user-friendliness and reduce the cognitive load on human users by assisting in complex decision-making processes. The effectiveness and functionality of the system are demonstrated through a case study, and the visualized demos and codes are available at a GitHub Repository: https://github.comlYuchenXia/LLMDrivenSimulation

No abstract available

Rapid urbanization, alongside escalating resource depletion and ecological degradation, underscores the critical need for innovative urban development solutions. In response, sustainable smart cities are increasingly turning to cutting-edge technologies—such as Generative Artificial Intelligence (GenAI), Foundation Models (FMs), and Urban Digital Twin (UDT) frameworks—to transform urban planning and design practices. These transformative tools provide advanced capabilities to analyze complex urban systems, optimize resource management, and enable evidence-based decision-making. Despite recent progress, research on integrating GenAI and FMs into UDT frameworks remains scant, leaving gaps in our ability to capture complex urban flows and multimodal dynamics essential to achieving environmental sustainability goals. Moreover, the lack of a robust theoretical foundation and real-world operationalization of these tools hampers comprehensive modeling and practical adoption. This study introduces a pioneering Large Flow Model (LFM), grounded in a robust foundational framework and designed with GenAI capabilities. It is specifically tailored for integration into UDT systems to enhance predictive analytics, adaptive learning, and complex data management functionalities. To validate its applicability and relevance, the Blue City Project in Lausanne City is examined as a case study, showcasing the ability of the LFM to effectively model and analyze urban flows—namely mobility, goods, energy, waste, materials, and biodiversity—critical to advancing environmental sustainability. This study highlights how the LFM addresses the spatial challenges inherent in current UDT frameworks. The LFM demonstrates its novelty in comprehensive urban modeling and analysis by completing impartial city data, estimating flow data in new locations, predicting the evolution of flow data, and offering a holistic understanding of urban dynamics and their interconnections. The model enhances decision-making processes, supports evidence-based planning and design, fosters integrated development strategies, and enables the development of more efficient, resilient, and sustainable urban environments. This research advances both the theoretical and practical dimensions of AI-driven, environmentally sustainable urban development by operationalizing GenAI and FMs within UDT frameworks. It provides sophisticated tools and valuable insights for urban planners, designers, policymakers, and researchers to address the complexities of modern cities and accelerate the transition towards sustainable urban futures.

This paper presents an innovative large model framework for optimizing the task offloading efficiency in vehicular edge networks, with a focus on ultra-reliable low-latency communication. We introduce a comprehensive model that integrates quantum computing with a deep reinforcement learning (DRL) model, supported by long short-term memory (LSTM) networks and a digital twin framework. This integration is designed to address the complexities of distributed vehicular edge computing networks, targeting efficient latency, energy, and quality-of-service management. Our model utilizes the parallel processing capabilities of quantum computing to enhance the DRL algorithm, effectively handling high-dimensional decision spaces. LSTM networks provide predictive insights into future network states in a digital twin framework and ensure real-time synchronization and adaptive strategy optimization. We employ a multi-agent framework, encompassing vehicles, unmanned aerial vehicles, and base stations, each utilizing a Nash equilibrium-based strategy for optimal decision-making, supplemented by incentive and penalty functions for reward optimization. Simulation results demonstrate notable improvements in task offloading efficiency, highlighting the model's efficacy over conventional DRL models.

No abstract available

This study explores a novel approach to reevaluating urban safety using Vision Large Language Models (VLLMs) integrated with digital twin technology. Our methodology involves randomly selecting street views across various U.S. cities and employing VLLMs to detect and analyze street safety. We incorporate existing user-reported data through an API to validate our findings. Preliminary results indicate that this new approach significantly enhances the accuracy and reliability of urban safety assessments. The integration of VLLMs with digital twin frameworks presents a promising avenue for urban planners and policymakers to achieve more dynamic and real-time insights into city safety, ultimately contributing to smarter and safer urban environments. Our findings suggest that this method holds substantial potential for broader applications in digital twin initiatives, facilitating more informed decision-making processes.

The next generation (6G) wireless networks are under intensive research and envisioned to realize the interconnection of everything and ubiquitous intelligence. One of the major challenges faced by 6G networks is to deliver fully intelligent, automated network control, and customized services, given the vast complexity, scale, and dynamics of the network environment. Digital Twin (DT) technology emerges as a promising solution. By replicating a virtual 6G network system, the DT can effectively monitor the system status, and provide robust planning and optimization functionalities to achieve efficient and intelligent control of 6G networks and applications. To enable intelligent and automated control of 6G networks, we leverage Large Generative Models (LGMs) and propose an innovative LGM-enabled DTs framework. In this framework, a real-time duplicate of the 6G network system is maintained within the DT, updated through continuous synchronization with the physical network. LGMs analyze network contexts and situations to generate intelligent control strategies for network optimization and automation. A case study on the LGM-enabled DT-supported connected autonomous driving is conducted with an incentive scheme designed for efficient DTs co-evolution. Preliminary results demonstrate the feasibility and superior performance of the proposed framework and schemes.

No abstract available

No abstract available

No abstract available

The growth of transportation networks and their increasing interconnections, although positive, has the downside effect of an increasing complexity which make them difficult to use, to assess and limits their efficiency. On average in the UK, 23% of travel time is lost in connections for trips with more than one mode and the lack of synchronization decreases very slowly with population size. This lack of synchronization between modes induces differences between the theoretical quickest trip and the ‘time-respecting’ path, which takes into account waiting times at interconnection nodes. We analyse here the statistics of these paths on the multilayer, temporal network of the entire, multimodal british public transportation system. We propose a statistical decomposition – the ‘anatomy’ – of trips in urban areas, in terms of riding, waiting and walking times and which shows how the temporal structure of trips varies with distance and allows us to compare different cities. Weaknesses in systems can be either insufficient transportation speed or service frequency, but the key parameter controlling their global efficiency is the total number of stop events per hour for all modes. This analysis suggests the need for better optimization strategies, adapted to short, long unimodal or multimodal trips.

The escalating demands of urbanization and globalization have placed immense pressure on existing transport infrastructures, necessitating the evolution of smarter, integrated, and sustainable mobility solutions. This study explores the integration of Artificial Intelligence (AI)-driven multimodal transport systems as a transformative approach to optimize real-time urban and intercity mobility. At the broader level, the research investigates how AI technologies—such as machine learning, computer vision, and natural language processing—can intelligently manage data from diverse transportation modes, including buses, trains, ride-shares, bicycles, and pedestrian pathways. The convergence of these modes through centralized, intelligent platforms enables seamless trip planning, efficient resource allocation, and improved commuter experiences. Focusing on the operational dimension, the study emphasizes real-time optimization capabilities powered by AI models that dynamically respond to traffic patterns, service disruptions, weather conditions, and passenger demands. Reinforcement learning algorithms are deployed to support adaptive routing, demand forecasting, and congestion management across multiple networks. Furthermore, the integration of geospatial data, Internet of Things (IoT) sensors, and edge computing enhances responsiveness and situational awareness, facilitating predictive analytics and automated decision-making in transport control centers. A case-based evaluation in metropolitan and intercity corridors demonstrates significant improvements in travel time reduction, fuel efficiency, and service reliability. The framework also supports inclusive mobility by integrating accessibility-focused solutions and equitable service distribution. This study concludes that AI-integrated multimodal transport systems represent a crucial pillar in the future of smart mobility, enabling governments and stakeholders to transition from fragmented transit systems to coordinated, intelligent ecosystems that are resilient, sustainable, and user-centric.

This paper addresses the pressing challenge of urban mobility in the context of growing urban populations, changing demand patterns for urban mobility, and emerging technologies like Mobility-on-Demand (MoD) platforms and Autonomous Vehicle (AV). As urban areas swell and demand pattern changes, the integration of Autonomous Mobility-on-Demand (AMoD) systems with existing public transit (PT) networks presents great opportunities to enhancing urban mobility. We propose a novel optimization framework for solving the Transit-Centric Multimodal Urban Mobility with Autonomous Mobility-on-Demand (TCMUM-AMoD) at scale. The system operator (public transit agency) determines the network design and frequency settings of the PT network, fleet sizing and allocations of AMoD system, and the pricing for using the multimodal system with the goal of minimizing passenger disutility. Passengers' mode and route choice behaviors are modeled explicitly using discrete choice models. A first-order approximation algorithm is introduced to solve the problem at scale. Using a case study in Chicago, we showcase the potential to optimize urban mobility across different demand scenarios. To our knowledge, ours is the first paper to jointly optimize transit network design, fleet sizing, and pricing for the multimodal mobility system while considering passengers' mode and route choices.

No abstract available

This study presents a Deep Reinforcement Learning (DRL) approach to address the multimodal mobility behavior of daily commuters, focusing specifically on students' home-university multimodal trips. The proposed mesoscopic model addresses key limitations of recent macro and microscopic models by balancing individual mobility preferences with significant group-level student factors. At its core, the model employs a Proximal Policy Optimization (PPO)-based agent that learns to match student navigation behavior in a multimodal transportation network, considering his group mobility factors such as vehicle ownership and origin-destination regions. Experiments conducted on a SUMO (Simulation of Urban MObility) simulated dataset of a university students' trips in the Toulouse metropolitan area demonstrate the model's performance in both unimodal and multimodal transportation scenarios. The resulting policy offers potential applications in predicting future multimodal mobility behavior, optimizing resource allocation for communities with regular travel needs, and developing more efficient and environment friendly urban transportation systems.

No abstract available

No abstract available

ABSTRACT This paper proposes a method for calculating transport networks capacity in dealing with multimodal transfers. Multimodal networks are represented by a modified ‘supernetwork’, while the passenger’s travels are defined as ‘superpaths’. Within this framework, the relation between the travel demand from O-D matrices and the resulting link flows in the supernetwork is modelled as a relationship matrix to describe urban mobility by using a logit-based stochastic user equilibrium. Based on this relationship matrix, an approximate iteration algorithm (AIA) is developed. Our numerical results show that the AIA performs better than the sensitivity analysis-based algorithm (SAB) and genetic algorithm (GA) regarding the execution-time, and that the capacity of multimodal transport networks can be underestimated if the combined travels are neglected.

No abstract available

Understanding the carbon emissions of multimodal travel—comprising walking, metro, bus, cycling, and ride-hailing—is essential for promoting sustainable urban mobility. However, most existing studies focus on single-mode travel, while underlying spatiotemporal and behavioral determinants remain insufficiently explored due to the lack of fine-grained data and interpretable analytical frameworks. This study proposes a novel integration of high-frequency, real-world mobility trajectory data with interpretable machine learning to systematically identify the key drivers of carbon emissions at the individual trip level. Firstly, multimodal travel chains are reconstructed using continuous GPS trajectory data collected in Beijing. Secondly, a model based on Calculate Emissions from Road Transport (COPERT) is developed to quantify trip-level CO2 emissions. Thirdly, four interpretable machine learning models based on gradient boosting—XGBoost, GBDT, LightGBM, and CatBoost—are trained using transportation and built environment features to model the relationship between CO2 emissions and a set of explanatory variables; finally, Shapley Additive exPlanations (SHAP) and partial dependence plots (PDPs) are used to interpret the model outputs, revealing key determinants and their non-linear interaction effects. The results show that transportation-related features account for 75.1% of the explained variance in emissions, with bus usage being the most influential single factor (contributing 22.6%). Built environment features explain the remaining 24.9%. The PDP analysis reveals that substantial emission reductions occur only when the shares of bus, metro, and cycling surpass threshold levels of approximately 40%, 40%, and 30%, respectively. Additionally, travel carbon emissions are minimized when trip origins and destinations are located within a 10 to 11 km radius of the central business district (CBD). This study advances the field by establishing a scalable, interpretable, and behaviorally grounded framework to assess carbon emissions from multimodal travel, providing actionable insights for low-carbon transport planning and policy design.

The integration of real-time multimodal traffic data fusion is critical for advancing Intelligent Transportation Systems (ITS) in urban environments. This paper presents a novel camera-based traffic data fusion system deployed at the Munich Mobility Research Campus (MORE) as a real-world laboratory. The system captures multimodal trajectories in real-time, which are processed using a state-of-the-art data fusion engine, integrating both low- and high-level fusion functionalities. The evaluation focuses on challenging traffic scenarios, including mixed-traffic environments, groups of vulnerable road users, and adjacent cycling pathways. Performance assessment for aggregated traffic counts is conducted using Root Mean Square Error and paired t-tests, while isolated detection is evaluated using True Positive Rate, Positive Predictive Value, and F1-Score. Results indicate high accuracy in detecting and classifying vehicles and bicycles, with minor underestimation biases observed in pedestrian detection within dense groups. The findings demonstrate the potential of real-time trajectory data fusion for multimodal ITS applications, supporting enhanced traffic planning, management, and control. Future work will extend this research by analyzing system latencies, refining speed accuracy, and exploring real-time anomaly detection. The proposed system contributes to the deployment and evaluation of scalable ITS solutions for connected, automated, and multimodal mobility.

Vision transformer capabilities for images have increased significantly in recent years. Multimodal vision transformers are now able to generate accurate captions for images and demonstrate strong capabilities in understanding this visual input. More recently, these models have been built to handle videos, with or without audio. However, these trans-formers have seldom been trained on datasets related to accessibility. In this study, we focus on generating navigation instructions for individuals with visual impairment in the context of outdoor, urban environments. We use the spatial-temporal vision language model (VLM), VideoLLaMA3, to process videos and generate a series of instructions based on a prompt specifically designed for individuals with visual impairments. With our approach, we were able to surpass the performance of the GPT-40 model. In the future, we anticipate this approach being extended through the use of landmark detection and improved fine-tuning. In this work, we investigate the use of VLMs as a backbone within a pipeline that incorporates prompting, postprocessing, and other techniques to develop spatially and temporally accurate instructions.

This paper aims at exploring the potentiality of the multimodal fusion of remote sensing imagery with information coming from mobility demand data in the framework of land-use mapping in urban areas. After a discussion on the function of mobility demand data, a probabilistic fusion framework is developed to take advantage of remote sensing and transport data, and their joint use for urban land-use and land-cover applications in urban and surrounding areas. Two different methods are proposed within this framework, the first based on pixelwise probabilistic decision fusion and the second on the combination with a region-based multiscale Markov random field. The experimental validation is conducted on a case study associated with the city of Genoa, Italy.

No abstract available

No abstract available

Traditional multimodal public transportation recommenders often overlook in-vehicle crowding, a critical factor that causes passenger discomfort and leads to an inefficient distribution of people across the network that affects its reliability. To address this, we propose a proof of concept for a novel framework that directly integrates crowding into its optimization process, balancing it with user preferences such as travel habits, travel time, and line changes. Built on the Behavior-Enabled IoT (BeT) paradigm, our system is designed to manage the crucial QoE and QoS trade-off inherent in smart mobility. We validate our balanced strategy using real-world data from Lyon, comparing it against two baselines: a QoE-driven model that prioritizes user habits and a QoS-driven model that focuses solely on network efficiency. Our Wilcoxon-based statistical analysis demonstrates that a balanced strategy is the most effective approach for substantially mitigating public transit crowding. Our Wilcoxon-based statistical analysis demonstrates that a balanced strategy is the most effective approach for mitigating public transit crowding, since it leads to a substantial decrease in crowding. Despite a potential increase in travel times, our solution respects user habits and avoids excessive transfers, providing significant operational improvements without compromising passenger convenience.

Multimodal urban transportation offers an efficient and reliable solution for urban mobility. This paper proposes a novel mathematical formulation, based on Mixed Integer Linear Programming (MILP), specifically addressing the optimization of formal public transportation modes within urban settings. Unlike existing models, our approach focuses exclusively on the formal transport sector while incorporating relevant operational constraints. The study begins with a concise review of the literature on optimization and organizational challenges in multimodal urban transport. Computational experiments are performed using an optimization solver to evaluate the performance and effectiveness of the proposed model.

No abstract available

Predicting human mobility holds significant practical value, with applications ranging from enhancing disaster risk planning to simulating epidemic spread. In this paper, we present the GeoFormer, a decoder-only transformer model adapted from the GPT architecture to forecast human mobility. Our proposed model is rigorously tested in the context of the HuMob Challenge 2023---a competition designed to evaluate the performance of prediction models on standardized datasets to predict human mobility. The challenge leverages two datasets encompassing urban-scale data of 25,000 and 100,000 individuals over a longitudinal period of 75 days. Geo-Former stands out as a top performer in the competition, securing a place in the top-3 ranking. Its success is underscored by performing well on both performance metrics chosen for the competition---the GEO-BLEU and the Dynamic Time Warping (DTW) measures. The performance of the GeoFormer on the HuMob Challenge 2023 underscores its potential to make substantial contributions to the field of human mobility prediction, with far-reaching implications for disaster preparedness, epidemic control, and beyond.

Human mobility generation plays a critical role in urban transportation planning. Existing human mobility generation models often fall short of understanding travelers' demographics and integrating multimodal information, including activity purposes, destination choices and transport mode preferences. Recently, mobility generation models leveraging Large Language Models (LLMs) have gained significant attention, while they are limited in directly reproducing spatial information in human mobility profiles. To address these challenges, this paper proposes the Mobility Generative Language Model (MobGLM), a novel approach for generating synthetic human mobility data to support urban planning, transport management, energy consumption and epidemic control. MobGLM addresses these limitations by capturing the complex relationships between agents' mobility patterns and individual demographics. By incorporating personal information, activity types, locations and traffic modes as encoders, MobGLM uniquely identifies and replicates features of human mobility. Our framework is evaluated using a large, real-world mobility dataset and benchmarked against state-of-the-art personal mobility generation techniques. The results demonstrate the effectiveness of MobGLM in producing accurate and reliable synthetic mobility data, highlighting its potential applications in various urban mobility contexts.

Predicting human daily behavior is challenging due to the complexity of routine patterns and short-term fluctuations. While data-driven models have improved behavior prediction by leveraging empirical data from various platforms and devices, the reliance on sensitive, large-scale user data raises privacy concerns and limits data availability. Synthetic data generation has emerged as a promising solution, though existing methods are often limited to specific applications. In this work, we introduce BehaviorGen, a framework that uses large language models (LLMs) to generate high-quality synthetic behavior data. By simulating user behavior based on profiles and real events, BehaviorGen supports data augmentation and replacement in behavior prediction models. We evaluate its performance in scenarios such as pertaining augmentation, fine-tuning replacement, and fine-tuning augmentation, achieving significant improvements in human mobility and smartphone usage predictions, with gains of up to 18.9%. Our results demonstrate the potential of BehaviorGen to enhance user behavior modeling through flexible and privacy-preserving synthetic data generation.

Cellular traffic data hold significant potential for applications such as network planning, traffic prediction, mobility modeling, and personalized recommendations. However, limited data accessibility hinders more open data-driven research. Previous studies have explored data synthesis, while exhibiting flexible limitations in supporting conditional traffic synthesis, and are vulnerable to multidimensional data modeling. In this paper, we present LLMCell, a flexible and effective framework that leverages the arbitrary conditioning and contextual understanding capabilities of the large language model (LLM) to generate high-quality synthetic cellular traffic data. The LLMCell comprises three key components: i) a textual encoder for converting raw cellular traffic data into textual representations, ii) a generative model learner to fine-tune pre-trained LLM based on encoded textual representation for cellular traffic generation, and iii) a synthetic data sampling module for final synthetic data sampling and textual-to-data transformation. Experiments conducted on a large-scale dataset demonstrate the superior fidelity and utility of LLMCell over state-of-the-art baselines, and the synthetic data can effectively preserve user privacy. We release our synthetic dataset to the public to benefit future research in the wireless network community1.

This study presents an innovative approach to urban mobility simulation by integrating a Large Language Model (LLM) with Agent-Based Modeling (ABM). Unlike traditional rule-based ABM, the proposed framework leverages LLM to enhance agent diversity and realism by generating synthetic population profiles, allocating routine and occasional locations, and simulating personalized routes. Using real-world data, the simulation models individual behaviors and large-scale mobility patterns in Taipei City. Key insights, such as route heat maps and mode-specific indicators, provide urban planners with actionable information for policy-making. Future work focuses on establishing robust validation frameworks to ensure accuracy and reliability in urban planning applications.

Human mobility modeling is critical for urban planning and transportation management, yet existing approaches often lack the integration capabilities needed to handle diverse data sources. We present a foundation model framework for universal human mobility patterns that leverages cross-domain data fusion and large language models to address these limitations. Our approach integrates multi-modal data of distinct nature and spatio-temporal resolution, including geographical, mobility, socio-demographic, and traffic information, to construct a privacy-preserving and semantically enriched human travel trajectory dataset. Our framework demonstrates adaptability through domain transfer techniques that ensure transferability across diverse urban contexts, as evidenced in case studies of Los Angeles (LA) and Egypt. The framework employs LLMs for semantic enrichment of trajectory data, enabling comprehensive understanding of mobility patterns. Quantitative evaluation shows that our generated synthetic dataset accurately reproduces mobility patterns observed in empirical data. The practical utility of this foundation model approach is demonstrated through large-scale traffic simulations for LA County, where results align well with observed traffic data. On California's I-405 corridor, the simulation yields a Mean Absolute Percentage Error of 5.85% for traffic volume and 4.36% for speed compared to Caltrans PeMS observations, illustrating the framework's potential for intelligent transportation systems and urban mobility applications.

Urban mobility analytics increasingly depend on large-scale trajectory data to support tasks such as route monitoring, operational auditing, and anomaly detection. However, taxi trajectory datasets contain highly sensitive spatio-temporal information, creating a fundamental trade-off between analytical utility and user privacy. To address this challenge, this paper proposes DiffTraj–LM TAD, an integrated privacy-preserving framework for synthetic taxi trajectory generation and anomaly probing. The framework combines a diffusion-based generative model to synthesize high-fidelity taxi trajectories with a transformer-based language model for unsupervised trajectory anomaly detection. An iterative refinement mechanism is introduced to enrich anomaly diversity and in terpretability without relying on manually labeled data, while differential privacy and trajectory-level anonymization provide formal privacy guarantees. Extensive experiments conducted on large-scale taxi datasets from Beijing, Chengdu, and Xi’an demonstrate that models trained on synthetic data achieve anomaly detection performance comparable to real data (F1-score ≈ 0.8), preserve strong statistical and spatial fidelity (Jensen–Shannon divergence < 0.05), and reduce adversarial inference success rates by more than 70%. The proposed framework enables reproducible, privacy-compliant mobility analytics and offers a practical foundation for developing and benchmarking trajectory anomaly detection systems without exposing sensitive user information.

Large language models (LLMs) have shown promise in automating travel planning, yet they often fall short in addressing nuanced spatiotemporal rationality. While existing benchmarks focus on basic plan validity, they neglect critical aspects such as route efficiency, POI appeal, and real-time adaptability. This paper introduces TP-RAG, the first benchmark tailored for retrieval-augmented, spatiotemporal-aware travel planning. Our dataset includes 2,348 real-world travel queries, 85,575 fine-grain annotated POIs, and 18,784 high-quality travel trajectory references sourced from online tourist documents, enabling dynamic and context-aware planning. Through extensive experiments, we reveal that integrating reference trajectories significantly improves spatial efficiency and POI rationality of the travel plan, while challenges persist in universality and robustness due to conflicting references and noisy data. To address these issues, we propose EvoRAG, an evolutionary framework that potently synergizes diverse retrieved trajectories with LLMs'intrinsic reasoning. EvoRAG achieves state-of-the-art performance, improving spatiotemporal compliance and reducing commonsense violation compared to ground-up and retrieval-augmented baselines. Our work underscores the potential of hybridizing Web knowledge with LLM-driven optimization, paving the way for more reliable and adaptive travel planning agents.

The widespread adoption of location-based services has led to the generation of vast amounts of mobility data, providing significant opportunities to model user movement dynamics within urban environments. Recent advancements have focused on adapting Large Language Models (LLMs) for mobility analytics. However, existing methods face two primary limitations: inadequate semantic representation of locations (i.e., discrete IDs) and insufficient modeling of mobility signals within LLMs (i.e., single templated instruction fine-tuning). To address these issues, we propose QT-Mob, a novel framework that significantly enhances LLMs for mobility analytics. QT-Mob introduces a location tokenization module that learns compact, semantically rich tokens to represent locations, preserving contextual information while ensuring compatibility with LLMs. Furthermore, QT-Mob incorporates a series of complementary fine-tuning objectives that align the learned tokens with the internal representations in LLMs, improving the model's comprehension of sequential movement patterns and location semantics. The proposed QT-Mob framework not only enhances LLMs' ability to interpret mobility data but also provides a more generalizable approach for various mobility analytics tasks. Experiments on three real-world dataset demonstrate the superior performance in both next-location prediction and mobility recovery tasks, outperforming existing deep learning and LLM-based methods.

In order to closely simulate the real network scenario thereby verify the effectiveness of protocol designs, it is necessary to model the traffic flows carried over realistic networks. Extensive studies [1] showed that the actual traffic in access and local area networks (e.g., those generated by ftp and video streams) exhibits the property of self-similarity and long-range dependency (LRD) [2]. In this appendix we briefly introduce the property of self-similarity and suggest a practical approach for modeling self-similar traces with specified traffic intensity.

Adaptive-Cruise Control (ACC) automatically accelerates or decelerates a vehicle to maintain a selected time gap, to reach a desired velocity, or to prevent a rear-end collision. To this end, the ACC sensors detect and track the vehicle ahead for measuring the actual distance and speed difference. Together with the own velocity, these input variables are exactly the same as in car-following models. The focus of this contribution is: What will be the impact of a spreading of ACC systems on the traffic dynamics? Do automated driving strategies have the potential to improve the capacity and stability of traffic flow or will they necessarily increase the heterogeneity and instability? How does the result depend on the ACC equipment level? We discuss microscopic modeling aspects for human and automated (ACC) driving. By means of microscopic traffic simulations, we study how a variable percentage of ACC-equipped vehicles influences the stability of traffic flow, the maximum flow under free traffic conditions until traffic breaks down, and the dynamic capacity of congested traffic. Furthermore, we compare different percentages of ACC with respect to travel times in a specific congestion scenario. Remarkably, we find that already a small amount of ACC equipped cars and, hence, a marginally increased free and dynamic capacity, leads to a drastic reduction of traffic congestion.

This paper introduces a fuzzy cellular model of road traffic that was intended for on-line applications in traffic control. The presented model uses fuzzy sets theory to deal with uncertainty of both input data and simulation results. Vehicles are modelled individually, thus various classes of them can be taken into consideration. In the proposed approach, all parameters of vehicles are described by means of fuzzy numbers. The model was implemented in a simulation of vehicles queue discharge process. Changes of the queue length were analysed in this experiment and compared to the results of NaSch cellular automata model.

Artificial intelligence is reshaping scientific exploration, but most methods automate procedural tasks without engaging in scientific reasoning, limiting autonomy in discovery. We introduce Materials Agents for Simulation and Theory in Electronic-structure Reasoning (MASTER), an active learning framework where large language models autonomously design, execute, and interpret atomistic simulations. In MASTER, a multimodal system translates natural language into density functional theory workflows, while higher-level reasoning agents guide discovery through a hierarchy of strategies, including a single agent baseline and three multi-agent approaches: peer review, triage-ranking, and triage-forms. Across two chemical applications, CO adsorption on Cu-surface transition metal (M) adatoms and on M-N-C catalysts, reasoning-driven exploration reduces required atomistic simulations by up to 90% relative to trial-and-error selection. Reasoning trajectories reveal chemically grounded decisions that cannot be explained by stochastic sampling or semantic bias. Altogether, multi-agent collaboration accelerates materials discovery and marks a new paradigm for autonomous scientific exploration.

This paper discusses two main themes. First, it investigates the formation of a spatiotemporal cognitive map (mental image) of a road network in travelers memory, which entails the travelers global conceptual understanding of congestion or the degree of crowding of the network. Second, it tries to investigate how latent learning of travelers from previous experiences shapes parts of the mental image, even for the parts of the network with which the travelers are unfamiliar. An experiment of route choice experiences was conducted among 90 participants in order to gain insight into the formation of a cognitive map and latent learning. In this experiment, the following independent variables are connected to the formation of a mental image of the network and the quality of the generalization of the unfamiliar parts of the network: (i) dispersion of the links travel time throughout the network, (ii) number of trips the traveler makes, (iii) travelers gender, (iv) travelers driving experiences, (v) travelers natural level of optimism or pessimism, (vi) salient or noticeable features on the network, and (vii) the presence of traffic signals. Several nonparametric (distribution-free) tests are employed to test the hypotheses. The results indicate that apart from the travelers gender and salient or noticeable features on the network, the considered factors significantly affect the degree of the recognizability of the network elements by travelers.

There is a clear desire to model and comprehend human behavior. Trends in research covering this topic show a clear assumption that many view human reasoning as the presupposed standard in artificial reasoning. As such, topics such as game theory, theory of mind, machine learning, etc. all integrate concepts which are assumed components of human reasoning. These serve as techniques to attempt to both replicate and understand the behaviors of humans. In addition, next generation autonomous and adaptive systems will largely include AI agents and humans working together as teams. To make this possible, autonomous agents will require the ability to embed practical models of human behavior, which allow them not only to replicate human models as a technique to "learn", but to to understand the actions of users and anticipate their behavior, so as to truly operate in symbiosis with them. The main objective of this paper it to provide a succinct yet systematic review of the most important approaches in two areas dealing with quantitative models of human behaviors. Specifically, we focus on (i) techniques which learn a model or policy of behavior through exploration and feedback, such as Reinforcement Learning, and (ii) directly model mechanisms of human reasoning, such as beliefs and bias, without going necessarily learning via trial-and-error.

Travel demand forecasting is an essential part of transportation planning and management. The four-step travel model is the traditional and most-common procedure utilized for travel demand forecasting, and many models have been proposed in the literature to describe each step separately. However, there is still a lack of a unified modeling framework that can successfully describe the collective choice behavior of travelers interacting with each other at different steps. This study uses the free utility model, whose objective function is mathematically consistent with the Helmholtz free energy in physics, to separately and simultaneously describe travelers' mode, destination, and route choice behaviors. The free utility model's basic assumption is that the travelers will trade off the expected utility and information-processing cost to maximize their own utility. This model provides not only a unified modeling framework for traveler choice behavior, but also provides a new perspective for understanding the user equilibrium model in transportation science and the potential game model in game theory.

As we discussed in Part I of this topic, there is a clear desire to model and comprehend human behavior. Given the popular presupposition of human reasoning as the standard for learning and decision-making, there have been significant efforts and a growing trend in research to replicate these innate human abilities in artificial systems. In Part I, we discussed learning methods which generate a model of behavior from exploration of the system and feedback based on the exhibited behavior as well as topics relating to the use of or accounting for beliefs with respect to applicable skills or mental states of others. In this work, we will continue the discussion from the perspective of methods which focus on the assumed cognitive abilities, limitations, and biases demonstrated in human reasoning. We will arrange these topics as follows (i) methods such as cognitive architectures, cognitive heuristics, and related which demonstrate assumptions of limitations on cognitive resources and how that impacts decisions and (ii) methods which generate and utilize representations of bias or uncertainty to model human decision-making or the future outcomes of decisions.

Although researchers increasingly adopt machine learning to model travel behavior, they predominantly focus on prediction accuracy, ignoring the ethical challenges embedded in machine learning algorithms. This study introduces an important missing dimension - computational fairness - to travel behavior analysis. We first operationalize computational fairness by equality of opportunity, then differentiate between the bias inherent in data and the bias introduced by modeling. We then demonstrate the prediction disparities in travel behavior modeling using the 2017 National Household Travel Survey (NHTS) and the 2018-2019 My Daily Travel Survey in Chicago. Empirically, deep neural network (DNN) and discrete choice models (DCM) reveal consistent prediction disparities across multiple social groups: both over-predict the false negative rate of frequent driving for the ethnic minorities, the low-income and the disabled populations, and falsely predict a higher travel burden of the socially disadvantaged groups and the rural populations than reality. Comparing DNN with DCM, we find that DNN can outperform DCM in prediction disparities because of DNN's smaller misspecification error. To mitigate prediction disparities, this study introduces an absolute correlation regularization method, which is evaluated with synthetic and real-world data. The results demonstrate the prevalence of prediction disparities in travel behavior modeling, and the disparities still persist regarding a variety of model specifics such as the number of DNN layers, batch size and weight initialization. Since these prediction disparities can exacerbate social inequity if prediction results without fairness adjustment are used for transportation policy making, we advocate for careful consideration of the fairness problem in travel behavior modeling, and the use of bias mitigation algorithms for fair transport decisions.

Machine learning has proved to be very successful for making predictions in travel behavior modeling. However, most machine-learning models have complex model structures and offer little or no explanation as to how they arrive at these predictions. Interpretations about travel behavior models are essential for decision makers to understand travelers' preferences and plan policy interventions accordingly. Therefore, this paper proposes to apply and extend the model distillation approach, a model-agnostic machine-learning interpretation method, to explain how a black-box travel mode choice model makes predictions for the entire population and subpopulations of interest. Model distillation aims at compressing knowledge from a complex model (teacher) into an understandable and interpretable model (student). In particular, the paper integrates model distillation with market segmentation to generate more insights by accounting for heterogeneity. Furthermore, the paper provides a comprehensive comparison of student models with the benchmark model (decision tree) and the teacher model (gradient boosting trees) to quantify the fidelity and accuracy of the students' interpretations.

Discovering patterns and detecting anomalies in individual travel behavior is a crucial problem in both research and practice. In this paper, we address this problem by building a probabilistic framework to model individual spatiotemporal travel behavior data (e.g., trip records and trajectory data). We develop a two-dimensional latent Dirichlet allocation (LDA) model to characterize the generative mechanism of spatiotemporal trip records of each traveler. This model introduces two separate factor matrices for the spatial dimension and the temporal dimension, respectively, and use a two-dimensional core structure at the individual level to effectively model the joint interactions and complex dependencies. This model can efficiently summarize travel behavior patterns on both spatial and temporal dimensions from very sparse trip sequences in an unsupervised way. In this way, complex travel behavior can be modeled as a mixture of representative and interpretable spatiotemporal patterns. By applying the trained model on future/unseen spatiotemporal records of a traveler, we can detect her behavior anomalies by scoring those observations using perplexity. We demonstrate the effectiveness of the proposed modeling framework on a real-world license plate recognition (LPR) data set. The results confirm the advantage of statistical learning methods in modeling sparse individual travel behavior data. This type of pattern discovery and anomaly detection applications can provide useful insights for traffic monitoring, law enforcement, and individual travel behavior profiling.

Understanding and modeling individual travel behavior responses is crucial for urban mobility regulation and policy evaluation. The Markov decision process (MDP) provides a structured framework for dynamic travel behavior modeling at the individual level. However, solving an MDP in this context is highly data-intensive and faces challenges of data quantity, spatial-temporal coverage, and situational diversity. To address these, we propose a group-effect-enhanced generative adversarial imitation learning (gcGAIL) model that improves the individual behavior modeling efficiency by leveraging shared behavioral patterns among passenger groups. We validate the gcGAIL model using a public transport fare-discount case study and compare against state-of-the-art benchmarks, including adversarial inverse reinforcement learning (AIRL), baseline GAIL, and conditional GAIL. Experimental results demonstrate that gcGAIL outperforms these methods in learning individual travel behavior responses to incentives over time in terms of accuracy, generalization, and pattern demonstration efficiency. Notably, gcGAIL is robust to spatial variation, data sparsity, and behavioral diversity, maintaining strong performance even with partial expert demonstrations and underrepresented passenger groups. The gcGAIL model predicts the individual behavior response at any time, providing the basis for personalized incentives to induce sustainable behavior changes (better timing of incentive injections).

Agent-based simulations have been used in modeling transportation systems for traffic management and passenger flows. In this work, we hope to shed light on the complex factors that influence transportation mode decisions within developing countries, using Colombia as a case study. We model an ecosystem of human agents that decide at each time step on the mode of transportation they would take to work. Their decision is based on a combination of their personal satisfaction with the journey they had just taken, which is evaluated across a personal vector of needs, the information they crowdsource from their prevailing social network, and their personal uncertainty about the experience of trying a new transport solution. We simulate different network structures to analyze the social influence for different decision-makers. We find that in low/medium connected groups inquisitive people actively change modes cyclically over the years while imitators cluster rapidly and change less frequently.

Transport mode detection is a classification problem aiming to design an algorithm that can infer the transport mode of a user given multimodal signals (GPS and/or inertial sensors). It has many applications, such as carbon footprint tracking, mobility behaviour analysis, or real-time door-to-door smart planning. Most current approaches rely on a classification step using Machine Learning techniques, and, like in many other classification problems, deep learning approaches usually achieve better results than traditional machine learning ones using handcrafted features. Deep models, however, have a notable downside: they are usually heavy, both in terms of memory space and processing cost. We show that a small, optimized model can perform as well as a current deep model. During our experiments on the GeoLife and SHL 2018 datasets, we obtain models with tens of thousands of parameters, that is, 10 to 1,000 times less parameters and operations than networks from the state of the art, which still reach a comparable performance. We also show, using the aforementioned datasets, that the current preprocessing used to deal with signals of different lengths is suboptimal, and we provide better replacements. Finally, we introduce a way to use signals with different lengths with the lighter Convolutional neural networks, without using the heavier Recurrent Neural Networks.

With the proliferation of electric vehicles (EVs), the transportation network and power grid become increasingly interdependent and coupled via charging stations. The concomitant growth in charging demand has posed challenges for both networks, highlighting the importance of charging coordination. Existing literature largely overlooks the interactions between power grid security and traffic efficiency. In view of this, we study the en-route charging station (CS) recommendation problem for EVs in dynamically coupled transportation-power systems. The system-level objective is to maximize the overall traffic efficiency while ensuring the safety of the power grid. This problem is for the first time formulated as a constrained Markov decision process (CMDP), and an online prediction-assisted safe reinforcement learning (OP-SRL) method is proposed to learn the optimal and secure policy by extending the PPO method. To be specific, we mainly address two challenges. First, the constrained optimization problem is converted into an equivalent unconstrained optimization problem by applying the Lagrangian method. Second, to account for the uncertain long-time delay between performing CS recommendation and commencing charging, we put forward an online sequence-to-sequence (Seq2Seq) predictor for state augmentation to guide the agent in making forward-thinking decisions. Finally, we conduct comprehensive experimental studies based on the Nguyen-Dupuis network and a large-scale real-world road network, coupled with IEEE 33-bus and IEEE 69-bus distribution systems, respectively. Results demonstrate that the proposed method outperforms baselines in terms of road network efficiency, power grid safety, and EV user satisfaction. The case study on the real-world network also illustrates the applicability in the practical context.

The augmented scale and complexity of urban transportation networks have significantly increased the execution time and resource requirements of vehicular network simulations, exceeding the capabilities of sequential simulators. The need for a parallel and distributed simulation environment is inevitable from a smart city perspective, especially when the entire city-wide information system is expected to be integrated with numerous services and ITS applications. In this paper, we present a conceptual model of an Integrated Distributed Connected Vehicle Simulator (IDCVS) that can emulate real-time traffic in a large metro area by incorporating hardware-in-the-loop simulation together with the closed-loop coupling of SUMO and OMNET++. We also discuss the challenges, issues, and solution approaches for implementing such a parallel closed-loop transportation network simulator by addressing transportation network partitioning problems, synchronization, and scalability issues. One unique feature of the envisioned integrated simulation tool is that it utilizes the vehicle traces collected through multiple roadway sensors-DSRC onboard unit, magnetometer, loop detector, and video detector. Another major feature of the proposed model is the incorporation of hybrid parallelism in both transportation and communication simulation platforms. We identify the challenges and issues involved in IDCVS to incorporate this multi-level parallelism. We also discuss the approaches for integrating hardware-in-the-loop simulation, addressing the steps involved in preprocessing sensor data, filtering, and extrapolating missing data, managing large real-time traffic data, and handling different data formats.

To make effective decisions in novel environments with long-horizon goals, it is crucial to engage in hierarchical reasoning across spatial and temporal scales. This entails planning abstract subgoal sequences, visually reasoning about the underlying plans, and executing actions in accordance with the devised plan through visual-motor control. We propose Compositional Foundation Models for Hierarchical Planning (HiP), a foundation model which leverages multiple expert foundation model trained on language, vision and action data individually jointly together to solve long-horizon tasks. We use a large language model to construct symbolic plans that are grounded in the environment through a large video diffusion model. Generated video plans are then grounded to visual-motor control, through an inverse dynamics model that infers actions from generated videos. To enable effective reasoning within this hierarchy, we enforce consistency between the models via iterative refinement. We illustrate the efficacy and adaptability of our approach in three different long-horizon table-top manipulation tasks.

This article proposes a methodology to model and simulate complex systems, based on IRM4MLS, a generic agent-based meta-model able to deal with multi-level systems. This methodology permits the engineering of dynamic multi-level agent-based models, to represent complex systems over several scales and domains of interest. Its goal is to simulate a phenomenon using dynamically the lightest representation to save computer resources without loss of information. This methodology is based on two mechanisms: (1) the activation or deactivation of agents representing different domain parts of the same phenomenon and (2) the aggregation or disaggregation of agents representing the same phenomenon at different scales.

Multi-agent reinforcement learning focuses on training the behaviors of multiple learning agents that coexist in a shared environment. Recently, MARL models, such as the Multi-Agent Transformer (MAT) and ACtion dEpendent deep Q-learning (ACE), have significantly improved performance by leveraging sequential decision-making processes. Although these models can enhance performance, they do not explicitly consider the importance of the order in which agents make decisions. In this paper, we propose an Agent Order of Action Decisions-MAT (AOAD-MAT), a novel MAT model that considers the order in which agents make decisions. The proposed model explicitly incorporates the sequence of action decisions into the learning process, allowing the model to learn and predict the optimal order of agent actions. The AOAD-MAT model leverages a Transformer-based actor-critic architecture that dynamically adjusts the sequence of agent actions. To achieve this, we introduce a novel MARL architecture that cooperates with a subtask focused on predicting the next agent to act, integrated into a Proximal Policy Optimization based loss function to synergistically maximize the advantage of the sequential decision-making. The proposed method was validated through extensive experiments on the StarCraft Multi-Agent Challenge and Multi-Agent MuJoCo benchmarks. The experimental results show that the proposed AOAD-MAT model outperforms existing MAT and other baseline models, demonstrating the effectiveness of adjusting the AOAD order in MARL.

There are more than 7,000 public transit agencies in the U.S. (and many more private agencies), and together, they are responsible for serving 60 billion passenger miles each year. A well-functioning transit system fosters the growth and expansion of businesses, distributes social and economic benefits, and links the capabilities of community members, thereby enhancing what they can accomplish as a society. Since affordable public transit services are the backbones of many communities, this work investigates ways in which Artificial Intelligence (AI) can improve efficiency and increase utilization from the perspective of transit agencies. This book chapter discusses the primary requirements, objectives, and challenges related to the design of AI-driven smart transportation systems. We focus on three major topics. First, we discuss data sources and data. Second, we provide an overview of how AI can aid decision-making with a focus on transportation. Lastly, we discuss computational problems in the transportation domain and AI approaches to these problems.

Airline disruption management traditionally seeks to address three problem dimensions: aircraft scheduling, crew scheduling, and passenger scheduling, in that order. However, current efforts have, at most, only addressed the first two problem dimensions concurrently and do not account for the propagative effects that uncertain scheduling outcomes in one dimension can have on another dimension. In addition, existing approaches for airline disruption management include human specialists who decide on necessary corrective actions for airline schedule disruptions on the day of operation. However, human specialists are limited in their ability to process copious amounts of information imperative for making robust decisions that simultaneously address all problem dimensions during disruption management. Therefore, there is a need to augment the decision-making capabilities of a human specialist with quantitative and qualitative tools that can rationalize complex interactions amongst all dimensions in airline disruption management, and provide objective insights to the specialists in the airline operations control center. To that effect, we provide a discussion and demonstration of an agnostic and systematic paradigm for enabling expeditious simultaneously-integrated recovery of all problem dimensions during airline disruption management, through an intelligent multi-agent system that employs principles from artificial intelligence and distributed ledger technology. Results indicate that our paradigm for simultaneously-integrated recovery executes in polynomial time and is effective when all the flights in the airline route network are disrupted.

Autonomous Vehicles (AVs) i.e., self-driving cars, operate in a safety critical domain, since errors in the autonomous driving software can lead to huge losses. Statistically, road intersections which are a part of the AVs operational design domain (ODD), have some of the highest accident rates. Hence, testing AVs to the limits on road intersections and assuring their safety on road intersections is pertinent, and thus the focus of this paper. We present a situation coverage-based (SitCov) AV-testing framework for the verification and validation (V&V) and safety assurance of AVs, developed in an open-source AV simulator named CARLA. The SitCov AV-testing framework focuses on vehicle-to-vehicle interaction on a road intersection under different environmental and intersection configuration situations, using situation coverage criteria for automatic test suite generation for safety assurance of AVs. We have developed an ontology for intersection situations, and used it to generate a situation hyperspace i.e., the space of all possible situations arising from that ontology. For the evaluation of our SitCov AV-testing framework, we have seeded multiple faults in our ego AV, and compared situation coverage based and random situation generation. We have found that both generation methodologies trigger around the same number of seeded faults, but the situation coverage-based generation tells us a lot more about the weaknesses of the autonomous driving algorithm of our ego AV, especially in edge-cases. Our code is publicly available online, anyone can use our SitCov AV-testing framework and use it or build further on top of it. This paper aims to contribute to the domain of V&V and development of AVs, not only from a theoretical point of view, but also from the viewpoint of an open-source software contribution and releasing a flexible/effective tool for V&V and development of AVs.

Challenged by urbanization and increasing travel needs, existing transportation systems need new mobility paradigms. In this article, we present the emerging concept of autonomous mobility-on-demand, whereby centrally orchestrated fleets of autonomous vehicles provide mobility service to customers. We provide a comprehensive review of methods and tools to model and solve problems related to autonomous mobility-on-demand systems. Specifically, we first identify problem settings for their analysis and control, from both operational and planning perspectives. We then review modeling aspects, including transportation networks, transportation demand, congestion, operational constraints, and interactions with existing infrastructure. Thereafter, we provide a systematic analysis of existing solution methods and performance metrics, highlighting trends and trade-offs. Finally, we present various directions for further research.

Autonomous driving has attracted significant research interests in the past two decades as it offers many potential benefits, including releasing drivers from exhausting driving and mitigating traffic congestion, among others. Despite promising progress, lane-changing remains a great challenge for autonomous vehicles (AV), especially in mixed and dynamic traffic scenarios. Recently, reinforcement learning (RL), a powerful data-driven control method, has been widely explored for lane-changing decision makings in AVs with encouraging results demonstrated. However, the majority of those studies are focused on a single-vehicle setting, and lane-changing in the context of multiple AVs coexisting with human-driven vehicles (HDVs) have received scarce attention. In this paper, we formulate the lane-changing decision making of multiple AVs in a mixed-traffic highway environment as a multi-agent reinforcement learning (MARL) problem, where each AV makes lane-changing decisions based on the motions of both neighboring AVs and HDVs. Specifically, a multi-agent advantage actor-critic network (MA2C) is developed with a novel local reward design and a parameter sharing scheme. In particular, a multi-objective reward function is proposed to incorporate fuel efficiency, driving comfort, and safety of autonomous driving. Comprehensive experimental results, conducted under three different traffic densities and various levels of human driver aggressiveness, show that our proposed MARL framework consistently outperforms several state-of-the-art benchmarks in terms of efficiency, safety and driver comfort.

Closed-loop evaluation is increasingly critical for end-to-end autonomous driving. Current closed-loop benchmarks using the CARLA simulator rely on manually configured traffic scenarios, which can diverge from real-world conditions, limiting their ability to reflect actual driving performance. To address these limitations, we introduce a simple yet challenging closed-loop evaluation framework that closely integrates real-world driving scenarios into the CARLA simulator with infrastructure cooperation. Our approach involves extracting 800 dynamic traffic scenarios selected from a comprehensive 100-hour video dataset captured by high-mounted infrastructure sensors, and creating static digital twin assets for 15 real-world intersections with consistent visual appearance. These digital twins accurately replicate the traffic and environmental characteristics of their real-world counterparts, enabling more realistic simulations in CARLA. This evaluation is challenging due to the diversity of driving behaviors, locations, weather conditions, and times of day at complex urban intersections. In addition, we provide a comprehensive closed-loop benchmark for evaluating end-to-end autonomous driving models. Project URL: \href{https://github.com/AIR-THU/DriveE2E}{https://github.com/AIR-THU/DriveE2E}.

We present CEMA: Causal Explanations in Multi-Agent systems; a framework for creating causal natural language explanations of an agent's decisions in dynamic sequential multi-agent systems to build more trustworthy autonomous agents. Unlike prior work that assumes a fixed causal structure, CEMA only requires a probabilistic model for forward-simulating the state of the system. Using such a model, CEMA simulates counterfactual worlds that identify the salient causes behind the agent's decisions. We evaluate CEMA on the task of motion planning for autonomous driving and test it in diverse simulated scenarios. We show that CEMA correctly and robustly identifies the causes behind the agent's decisions, even when a large number of other agents is present, and show via a user study that CEMA's explanations have a positive effect on participants' trust in autonomous vehicles and are rated as high as high-quality baseline explanations elicited from other participants. We release the collected explanations with annotations as the HEADD dataset.

The capability to learn and adapt to changes in the driving environment is crucial for developing autonomous driving systems that are scalable beyond geo-fenced operational design domains. Deep Reinforcement Learning (RL) provides a promising and scalable framework for developing adaptive learning based solutions. Deep RL methods usually model the problem as a (Partially Observable) Markov Decision Process in which an agent acts in a stationary environment to learn an optimal behavior policy. However, driving involves complex interaction between multiple, intelligent (artificial or human) agents in a highly non-stationary environment. In this paper, we propose the use of Partially Observable Markov Games(POSG) for formulating the connected autonomous driving problems with realistic assumptions. We provide a taxonomy of multi-agent learning environments based on the nature of tasks, nature of agents and the nature of the environment to help in categorizing various autonomous driving problems that can be addressed under the proposed formulation. As our main contributions, we provide MACAD-Gym, a Multi-Agent Connected, Autonomous Driving agent learning platform for furthering research in this direction. Our MACAD-Gym platform provides an extensible set of Connected Autonomous Driving (CAD) simulation environments that enable the research and development of Deep RL- based integrated sensing, perception, planning and control algorithms for CAD systems with unlimited operational design domain under realistic, multi-agent settings. We also share the MACAD-Agents that were trained successfully using the MACAD-Gym platform to learn control policies for multiple vehicle agents in a partially observable, stop-sign controlled, 3-way urban intersection environment with raw (camera) sensor observations.

Travel planning (TP) agent has recently worked as an emerging building block to interact with external tools and resources for travel itinerary generation, ensuring enjoyable user experience. Despite its benefits, existing studies rely on hand craft prompt and fixed agent workflow, hindering more flexible and autonomous TP agent. This paper proposes DeepTravel, an end to end agentic reinforcement learning framework for building autonomous travel planning agent, capable of autonomously planning, executing tools, and reflecting on tool responses to explore, verify, and refine intermediate actions in multi step reasoning. To achieve this, we first construct a robust sandbox environment by caching transportation, accommodation and POI data, facilitating TP agent training without being constrained by real world APIs limitations (e.g., inconsistent outputs). Moreover, we develop a hierarchical reward modeling system, where a trajectory level verifier first checks spatiotemporal feasibility and filters unsatisfied travel itinerary, and then the turn level verifier further validate itinerary detail consistency with tool responses, enabling efficient and precise reward service. Finally, we propose the reply augmented reinforcement learning method that enables TP agent to periodically replay from a failures experience buffer, emerging notable agentic capacity. We deploy trained TP agent on DiDi Enterprise Solutions App and conduct comprehensive online and offline evaluations, demonstrating that DeepTravel enables small size LLMs (e.g., Qwen3 32B) to significantly outperform existing frontier LLMs such as OpenAI o1, o3 and DeepSeek R1 in travel planning tasks.

Digital twins are models of real-world systems that can simulate their dynamics in response to potential actions. In complex settings, the state and action variables, and available data and knowledge relevant to a system can constantly change, requiring digital twins to continuously update with these changes to remain relevant. Current approaches struggle in this regard, as they require fixed, well-defined modelling environments, and they cannot adapt to novel variables without re-designs, or incorporate new information without re-training. To address this, we frame digital twinning as an in-context learning problem using large language models, enabling seamless updates to the twin at inference time. We develop CALM-DT, a Context-Adaptive Language Model-based Digital Twin that can accurately simulate across diverse state-action spaces using in-context learning alone by utilising fine-tuned encoders for sample retrieval. We empirically demonstrate CALM-DT's competitive performance with existing digital twin approaches, and its unique ability to adapt to changes in its modelling environment without parameter updates.

As a promising technology, vehicular edge computing (VEC) can provide computing and caching services by deploying VEC servers near vehicles. However, VEC networks still face challenges such as high vehicle mobility. Digital twin (DT), an emerging technology, can predict, estimate, and analyze real-time states by digitally modeling objects in the physical world. By integrating DT with VEC, a virtual vehicle DT can be created in the VEC server to monitor the real-time operating status of vehicles. However, maintaining the vehicle DT model requires ongoing attention from the VEC server, which also needs to offer computing services for the vehicles. Therefore, effective allocation and scheduling of VEC server resources are crucial. This study focuses on a general VEC network with a single VEC service and multiple vehicles, examining the two types of delays caused by twin maintenance and computational processing within the network. By transforming the problem using satisfaction functions, we propose an optimization problem aimed at maximizing each vehicle's resource utility to determine the optimal resource allocation strategy. Given the non-convex nature of the issue, we employ multi-agent Markov decision processes to reformulate the problem. Subsequently, we propose the twin maintenance and computing task processing resource collaborative scheduling (MADRL-CSTC) algorithm, which leverages multi-agent deep reinforcement learning. Through experimental comparisons with alternative algorithms, it demonstrates that our proposed approach is effective in terms of resource allocation.

Digital Twins (DTs) are rapidly emerging as a fundamental brick of engineering cyber-physical systems, but their notion is still mostly bound to specific business domains (e.g. manufacturing), goals (e.g. product design), or application domains (e.g. the Internet of Things). As such, their value as general purpose engineering abstractions is yet to be fully revealed. In this paper, we relate DTs with agents and multiagent systems, as the latter are arguably the most rich abstractions available for the engineering of complex socio-technical and cyber-physical systems, and the former could both fill in some gaps in agent-oriented engineering and benefit from an agent-oriented interpretation -- in a cross-fertilisation journey.

Digital twins, as precise digital representations of physical systems, have evolved from passive simulation tools into intelligent and autonomous entities through the integration of artificial intelligence technologies. This paper presents a unified four-stage framework that systematically characterizes AI integration across the digital twin lifecycle, spanning modeling, mirroring, intervention, and autonomous management. By synthesizing existing technologies and practices, we distill a unified four-stage framework that systematically characterizes how AI methodologies are embedded across the digital twin lifecycle: (1) modeling the physical twin through physics-based and physics-informed AI approaches, (2) mirroring the physical system into a digital twin with real-time synchronization, (3) intervening in the physical twin through predictive modeling, anomaly detection, and optimization strategies, and (4) achieving autonomous management through large language models, foundation models, and intelligent agents. We analyze the synergy between physics-based modeling and data-driven learning, highlighting the shift from traditional numerical solvers to physics-informed and foundation models for physical systems. Furthermore, we examine how generative AI technologies, including large language models and generative world models, transform digital twins into proactive and self-improving cognitive systems capable of reasoning, communication, and creative scenario generation. Through a cross-domain review spanning eleven application domains, including healthcare, aerospace, smart manufacturing, robotics, and smart cities, we identify common challenges related to scalability, explainability, and trustworthiness, and outline directions for responsible AI-driven digital twin systems.

In this chapter, we discuss urban mobility from a complexity science perspective. First, we give an overview of the datasets that enable this approach, such as mobile phone records, location-based social network traces, or GPS trajectories from sensors installed on vehicles. We then review the empirical and theoretical understanding of the properties of human movements, including the distribution of travel distances and times, the entropy of trajectories, and the interplay between exploration and exploitation of locations. Next, we explain generative and predictive models of individual mobility, and their limitations due to intrinsic limits of predictability. Finally, we discuss urban transport from a systemic perspective, including system-wide challenges like ridesharing, multimodality, and sustainable transport.

Urban traffic flow is governed by the complex, nonlinear interaction between land use configuration and spatiotemporally heterogeneous mobility demand. Conventional global regression and time-series models cannot simultaneously capture these multi-scale dynamics across multiple travel modes. This study proposes a GeoAI Hybrid analytical framework that sequentially integrates Multiscale Geographically Weighted Regression (MGWR), Random Forest (RF), and Spatio-Temporal Graph Convolutional Networks (ST-GCN) to model the spatiotemporal heterogeneity of traffic flow patterns and their interaction with land use across three mobility modes: motor vehicle, public transit, and active transport. Applying the framework to an empirically calibrated dataset of 350 traffic analysis zones across six cities spanning two contrasting urban morphologies, four key findings emerge: (i) the GeoAI Hybrid achieves a root mean squared error (RMSE) of 0.119 and an R^2 of 0.891, outperforming all benchmarks by 23-62%; (ii) SHAP analysis identifies land use mix as the strongest predictor for motor vehicle flows and transit stop density as the strongest predictor for public transit; (iii) DBSCAN clustering identifies five functionally distinct urban traffic typologies with a silhouette score of 0.71, and GeoAI Hybrid residuals exhibit Moran's I=0.218 (p<0.001), a 72% reduction relative to OLS baselines; and (iv) cross-city transfer experiments reveal moderate within-cluster transferability (R^2>=0.78) and limited cross-cluster generalisability, underscoring the primacy of urban morphological context. The framework offers planners and transportation engineers an interpretable, scalable toolkit for evidence-based multimodal mobility management and land use policy design.

This paper explores the use of Generative Pre-trained Transformers (GPT) in strategic game experiments, specifically the ultimatum game and the prisoner's dilemma. I designed prompts and architectures to enable GPT to understand the game rules and to generate both its choices and the reasoning behind decisions. The key findings show that GPT exhibits behaviours similar to human responses, such as making positive offers and rejecting unfair ones in the ultimatum game, along with conditional cooperation in the prisoner's dilemma. The study explores how prompting GPT with traits of fairness concern or selfishness influences its decisions. Notably, the "fair" GPT in the ultimatum game tends to make higher offers and reject offers more frequently compared to the "selfish" GPT. In the prisoner's dilemma, high cooperation rates are maintained only when both GPT players are "fair". The reasoning statements GPT produces during gameplay reveal the underlying logic of certain intriguing patterns observed in the games. Overall, this research shows the potential of GPT as a valuable tool in social science research, especially in experimental studies and social simulations.

The massive digital footprints generated by bike-sharing systems in megacities like Shanghai offer a novel perspective on the urban socio-economic fabric. This study investigates whether these daily mobility patterns can quantitatively map the city's underlying social stratification. To overcome the persistent challenge of acquiring fine-grained socio-economic data, we constructed a multi-layered analytical dataset. We annotated 2,000 raw bike trips with local economic attributes, derived from a novel data enrichment methodology that employs a Large Language Model (LLM), and integrated contextual features of the built environment. A Random Forest model was then utilized as an interpretable framework to determine the key factors governing the relationship between mobility behavior and local economic status. The analysis reveals a compelling and unambiguous finding: a neighborhood's economic level, proxied by housing prices, is the single most dominant predictor of its bike-sharing patterns, substantially outweighing other geographic or temporal factors. This economic determinism manifests in three distinct ways: (1) a spatial clustering of resources, a phenomenon we term the \textit{club effect}, which concentrates mobility infrastructure and usage in affluent areas; (2) a functional dichotomy between necessity-driven, utilitarian usage in lower-income zones and flexible, recreational usage in wealthier ones; and (3) a nuanced inverted U-shaped adoption curve that identifies the urban middle class as the system's primary user base.

Mobility trajectories are essential for understanding urban dynamics and enhancing urban planning, yet access to such data is frequently hindered by privacy concerns. This research introduces a transformative framework for generating large-scale urban mobility trajectories, employing a novel application of a transformer-based model pre-trained and fine-tuned through a two-phase process. Initially, trajectory generation is conceptualized as an offline reinforcement learning (RL) problem, with a significant reduction in vocabulary space achieved during tokenization. The integration of Inverse Reinforcement Learning (IRL) allows for the capture of trajectory-wise reward signals, leveraging historical data to infer individual mobility preferences. Subsequently, the pre-trained model is fine-tuned using the constructed reward model, effectively addressing the challenges inherent in traditional RL-based autoregressive methods, such as long-term credit assignment and handling of sparse reward environments. Comprehensive evaluations on multiple datasets illustrate that our framework markedly surpasses existing models in terms of reliability and diversity. Our findings not only advance the field of urban mobility modeling but also provide a robust methodology for simulating urban data, with significant implications for traffic management and urban development planning. The implementation is publicly available at https://github.com/Wangjw6/TrajGPT_R.

The trend for Urban Air Mobility (UAM) is growing with prospective air taxis, parcel deliverers, and medical and industrial services. Safe and efficient UAM operation relies on timely communication and reliable data exchange. In this paper, we explore Cooperative Perception (CP) for Unmanned Aircraft Systems (UAS), considering the unique communication needs involving high dynamics and a large number of UAS. We propose a hybrid approach combining local broadcast with a central CP service, inspired by centrally managed U-space and broadcast mechanisms from automotive and aviation domains. In a simulation study, we show that our approach significantly enhances the environmental awareness for UAS compared to fully distributed approaches, with an increased communication channel load, which we also evaluate. These findings prompt a discussion on communication strategies for CP in UAM and the potential of a centralized CP service in future research.

Micromobility, which utilizes lightweight mobile machines moving in urban public spaces, such as delivery robots and mobility scooters, emerges as a promising alternative to vehicular mobility. Current micromobility depends mostly on human manual operation (in-person or remote control), which raises safety and efficiency concerns when navigating busy urban environments full of unpredictable obstacles and pedestrians. Assisting humans with AI agents in maneuvering micromobility devices presents a viable solution for enhancing safety and efficiency. In this work, we present a scalable urban simulation solution to advance autonomous micromobility. First, we build URBAN-SIM - a high-performance robot learning platform for large-scale training of embodied agents in interactive urban scenes. URBAN-SIM contains three critical modules: Hierarchical Urban Generation pipeline, Interactive Dynamics Generation strategy, and Asynchronous Scene Sampling scheme, to improve the diversity, realism, and efficiency of robot learning in simulation. Then, we propose URBAN-BENCH - a suite of essential tasks and benchmarks to gauge various capabilities of the AI agents in achieving autonomous micromobility. URBAN-BENCH includes eight tasks based on three core skills of the agents: Urban Locomotion, Urban Navigation, and Urban Traverse. We evaluate four robots with heterogeneous embodiments, such as the wheeled and legged robots, across these tasks. Experiments on diverse terrains and urban structures reveal each robot's strengths and limitations.

Carsharing is a model of renting vehicles for short periods of time, where the payment is made according to the time and distance effectively traveled. Carsharing offers a simple, economical and smart alternative to urban mobility, that is already being adopted in the major cities in the world. The proposed methodology consisted in the development of a decision support system that simplifies the process of choosing carsharing services. Adopting the AHP method, the user can indicate their preferences in the choice of vehicles, and the system returns an ordered list of the most suitable available vehicles based on their geographic location. The findings of the project indicate that the use of this system encourage and simplify the use of carsharing services, which will allow to enhance the financial, mobility and environment advantages inherent to their use.

Urban Artificial Intelligence (Urban AI) has advanced human-centered urban tasks such as perception prediction and human dynamics. Large Language Models (LLMs) can integrate multimodal inputs to address heterogeneous data in complex urban systems but often underperform on domain-specific tasks. Urban-MAS, an LLM-based Multi-Agent System (MAS) framework, is introduced for human-centered urban prediction under zero-shot settings. It includes three agent types: Predictive Factor Guidance Agents, which prioritize key predictive factors to guide knowledge extraction and enhance the effectiveness of compressed urban knowledge in LLMs; Reliable UrbanInfo Extraction Agents, which improve robustness by comparing multiple outputs, validating consistency, and re-extracting when conflicts occur; and Multi-UrbanInfo Inference Agents, which integrate extracted multi-source information across dimensions for prediction. Experiments on running-amount prediction and urban perception across Tokyo, Milan, and Seattle demonstrate that Urban-MAS substantially reduces errors compared to single-LLM baselines. Ablation studies indicate that Predictive Factor Guidance Agents are most critical for enhancing predictive performance, positioning Urban-MAS as a scalable paradigm for human-centered urban AI prediction. Code is available on the project website:https://github.com/THETUREHOOHA/UrbanMAS

Simulations play a vital role in implementing, testing and validating proposed algorithms and protocols in VANET. Mobility model, defined as the movement pattern of vehicles, is one of the main factors that contribute towards the efficient implementation of VANET algorithms and protocols. Using near reality mobility models ensure that accurate results are obtained from simulations. Mobility models that have been proposed and used to implement and test VANET protocols and algorithms are either the urban mobility model or highway mobility model. Algorithms and protocols implemented using urban or highway mobility models may not produce accurate results in hybrid mobility models without enhancement due to the vast differences in mobility patterns. It is on this score the Hybrist, a novel hybrid mobility model is proposed. The realistic mobility pattern trace file of the proposed Hybrist hybrid mobility model can be imported to VANET simulators such as Veins and network simulators such as ns2 and Qualnet to simulate VANET algorithms and protocols.

Shared Mobility Services (SMSs) are transforming urban transportation systems by offering flexible travel options. These services, which help reduce the number of cars on the roads, have the potential to enhance the transportation system's performance, leading to improvements in travel times and emissions. This emphasizes the importance of assessing their impact on the system and users' choices, particularly when integrated into complex multi-modal systems that include public transport (PT). However, many studies overlook the synergies between SMSs and PT, leading to inaccurate traffic estimations and planning. This research presents an extensive review of multi-modal transportation system models incorporating SMSs. It then introduces a multimodal traffic assignment model including almost all mobility options in urban transportation systems applicable in both continuous and integer settings, leading to a Mixed-Integer Bilinear Programming (MIBLP) formulation. The model comprises diverse travel options, including SMSs, and accounts for intermodality by allowing commuters to combine modes to optimize time and monetary expense. An in-depth examination of commuters' mode and path choices on two test cases and an analysis of the price of anarchy reveals the disparities between user equilibrium and system optimum in such complex networks.