International audience

• This study interpolates urban street-level traffic volume • A graph-based model captures spatio-temporal traffic patterns • Method includes node features and custom loss function • Results are validated on novel Berlin and New York City datasets • Model outperforms baselines, especially under data scarcity Graph Neural Networks have shown strong performance in traffic volume forecasting, particularly on highways and major arterial networks. Applying these models to urban street networks, however, presents unique challenges: urban networks are structurally more diverse, traffic volumes are highly overdispersed with many zeros, spatial dependencies are complex, and sensor coverage is often very sparse. To address these challenges, we introduce the Graph Neural Network for Urban Interpolation (GNNUI), a model designed specifically for citywide traffic volume interpolation. GNNUI employs a designated masking strategy to learn interpolation, integrates node features to capture the different functional roles across the street network, and uses a loss function tailored to zero-inflated traffic distributions. We evaluate GNNUI on two newly constructed, large-scale urban traffic volume benchmarks, covering different transportation modes: Strava cycling data from Berlin and New York City taxi data. Across multiple evaluation metrics, GNNUI outperforms both the state-of-the-art graph-based interpolation model IGNNK and the widely used machine-learning baseline XGBoost, reducing MAE by at least 13% on Strava data and 7% on Taxi data while better capturing the empirical traffic distribution, and improving the identification of zero-traffic streets. Additionally, the model remains robust under the realistic case of extremely scarce ground truth sensor data. When sensor coverage is reduced from 90% to 1%, the MAE increases by approximately 48% on Strava and 76% on the taxi data, despite the near-complete removal of sensor information. We also examine how graph connectivity choices influence model performance, and find that a simple and computationally efficient binary adjacency matrix outperforms distance or similarity based ones.

Urban mobility is on the cusp of transformation with the emergence of shared, connected, and cooperative automated vehicles. Yet, for them to be accepted by customers, trust in their punctuality is vital. Many pilot initiatives operate without a fixed schedule, enhancing the importance of reliable arrival time (AT) predictions. This study presents an AT prediction system for automated shuttles, utilizing separate models for dwell and running time predictions, validated on real-world data from six cities. Alongside established methods such as XGBoost, we explore the benefits of leveraging spatial correlations using graph neural networks (GNN). To accurately handle the case of a shuttle bypassing a stop, we propose a hierarchical model combining a random forest classifier and a GNN. The results for the final AT prediction are promising, showing low errors even when predicting several stops ahead. Yet, no single model emerges as universally superior, and we provide insights into the characteristics of pilot sites that influence the model selection process and prediction performance. Finally, we identify dwell time prediction as the key determinant in overall AT prediction accuracy when automated shuttles are deployed in low-traffic areas or under regulatory speed limits. Our meta-analysis across six pilot sites in different cities provides insights into the current state of autonomous public transport prediction models and paves the way for more data-informed decision-making as the field advances.

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 paper 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.

Hierarchical policies enable strong performance in many sequential decision-making problems, such as those with high-dimensional action spaces, those requiring long-horizon planning, and settings with sparse rewards. However, learning hierarchical policies from static offline datasets presents a significant challenge. Crucially, actions taken by higher-level policies may not be directly observable within hierarchical controllers, and the offline dataset might have been generated using a different policy structure, hindering the use of standard offline learning algorithms. In this work, we propose $\textit{OHIO}$: a framework for offline reinforcement learning (RL) of hierarchical policies. Our framework leverages knowledge of the policy structure to solve the $\textit{inverse problem}$, recovering the unobservable high-level actions that likely generated the observed data under our hierarchical policy. This approach constructs a dataset suitable for off-the-shelf offline training. We demonstrate our framework on robotic and network optimization problems and show that it substantially outperforms end-to-end RL methods and improves robustness. We investigate a variety of instantiations of our framework, both in direct deployment of policies trained offline and when online fine-tuning is performed. Code and data are available at https://ohio-offline-hierarchical-rl.github.io.

Inferring the true demand for a product or a service from aggregate data is often challenging due to the limited available supply, thus resulting in observations that are censored and correspond to the realized demand, thereby not accounting for the unsatisfied demand. Censored regression models are able to account for the effect of censoring due to the limited supply, but they don't consider the effect of substitutions, which may cause the demand for similar alternative products or services to increase. This paper proposes Diffusion-aware Censored Demand Models, which combine a Tobit likelihood with a graph-based diffusion process in order to model the latent process of transfer of unsatisfied demand between similar products or services. We instantiate this new class of models under the framework of GPs and, based on both simulated and real-world data for modeling sales, bike-sharing demand, and EV charging demand, demonstrate its ability to better recover the true demand and produce more accurate out-of-sample predictions.

Rapid urbanization in the past decades has significantly escalated mobility demand and imposed higher service quality standards. As a promising solution to address this challenge, Autonomous Mobility-on-Demand (AMoD) systems offer tailored mobility services while facilitating centralized control. In this paper, we formulate the joint rebalancing and dynamic pricing problem in AMoD systems as a reinforcement learning problem over graph elements. By proposing a hierarchical policy, we exploit the benefits of both optimization and reinforcement learning methods. Through experiments conducted with real-world data from New York City and San Francisco, we demonstrate that, by leveraging the joint policy, the two control mechanisms can work in tandem to effectively address the limitations inherent to independent dynamic pricing or rebalancing policies under different scenarios. Crucially, we show that our approach is effective when learning from both (1) online interaction with the transportation system and (2) offline from static historical data, thus avoiding the potentially expensive interaction needed for training. The success of our proposed framework in improving system performance under different problem settings underscores its potential as a viable solution for controlling real-world transportation systems.

Optimization problems over dynamic networks have been extensively studied and widely used in the past decades to formulate numerous real-world problems. However, (1) traditional optimization-based approaches do not scale to large networks, and (2) the design of good heuristics or approximation algorithms often requires significant manual trial-and-error. In this work, we argue that data-driven strategies can automate this process and learn efficient algorithms without compromising optimality. To do so, we present network control problems through the lens of reinforcement learning and propose a graph network-based framework to handle a broad class of problems. Instead of naively computing actions over high-dimensional graph elements, e.g., edges, we propose a bi-level formulation where we (1) specify a desired next state via RL, and (2) solve a convex program to best achieve it, leading to drastically improved scalability and performance. We further highlight a collection of desirable features to system designers, investigate design decisions, and present experiments on real-world control problems showing the utility, scalability, and flexibility of our framework.

This paper presents two novel approaches for uncertainty estimation adapted and extended for the multi-link bus travel time problem. The uncertainty is modeled directly as part of recurrent artificial neural networks, but using two fundamentally different approaches: one based on Deep Quantile Regression and the other on Bayesian neural network. Both approaches use a recurrent neural network to predict multiple time steps into the future, but handle the time-dependent uncertainty estimation differently. We present a novel sampling technique in order to aggregate quantile estimates for link level travel time to yield the multi-link travel time distribution needed for a vehicle to travel from its current position to a specific downstream stop point or transfer site. To motivate the relevance of uncertainty-aware models in the domain, we focus on the connection protection application as a case study: An expert system to determine whether a bus driver should hold and wait for a connecting service, thus ensuring the connection, or break the connection and reduce its own delay. Our results show that the proposed quantile sampling method performs overall best for the 80%, 90% and 95% prediction intervals, both for a 15 min time horizon into the future (t+1), but also for the 30 and 45 min time horizon (t+2 and t+3), with a constant, but very small underestimation of the uncertainty interval (1–4 pp.). However, we also show, that the Bayesian model still can outperform the DQR for specific cases. Lastly, we demonstrate how a simple decision support system can take advantage of our uncertainty-aware travel time models to prioritize the difference in travel time uncertainty for bus holding at strategic points, thus reducing the introduced delay for the connection protection application.

Autonomous Mobility-on-Demand (AMoD) systems represent an attractive alternative to existing transportation paradigms, currently challenged by urbanization and increasing travel needs. By centrally controlling a fleet of self-driving vehicles, these systems provide mobility service to customers and are currently starting to be deployed in a number of cities around the world. Current learning-based approaches for controlling AMoD systems are limited to the single-city scenario, whereby the service operator is allowed to take an unlimited amount of operational decisions within the same transportation system. However, real-world system operators can hardly afford to fully re-train AMoD controllers for every city they operate in, as this could result in a high number of poor-quality decisions during training, making the single-city strategy a potentially impractical solution. To address these limitations, we propose to formalize the multi-city AMoD problem through the lens of meta-reinforcement learning (meta-RL) and devise an actor-critic algorithm based on recurrent graph neural networks. In our approach, AMoD controllers are explicitly trained such that a small amount of experience within a new city will produce good system performance. Empirically, we show how control policies learned through meta-RL are able to achieve near-optimal performance on unseen cities by learning rapidly adaptable policies, thus making them more robust not only to novel environments, but also to distribution shifts common in real-world operations, such as special events, unexpected congestion, and dynamic pricing schemes.

Mobility-on-demand (MoD) systems represent a rapidly developing mode of transportation wherein travel requests are dynamically handled by a coordinated fleet of vehicles. Crucially, the efficiency of an MoD system highly depends on how well supply and demand distributions are aligned in spatio-temporal space (i.e., to satisfy user demand, cars have to be available in the correct place and at the desired time). To do so, we argue that predictive models should aim to explicitly disentangle between temporal and spatial variability in the evolution of urban mobility demand. However, current approaches typically ignore this distinction by either treating both sources of variability jointly, or completely ignoring their presence in the first place. In this paper, we propose recurrent flow networks 1 (RFN), where we explore the inclusion of (i) latent random variables in the hidden state of recurrent neural networks to model temporal variability, and (ii) normalizing flows to model the spatial distribution of mobility demand. We demonstrate how predictive models explicitly disentangling between spatial and temporal variability exhibit several desirable properties, and empirically show how this enables the generation of distributions matching potentially complex urban topologies.

Shared mobility services require accurate demand models for effective service planning. On the one hand, modeling the full probability distribution of demand is advantageous because the entire uncertainty structure preserves valuable information for decision-making. On the other hand, demand is often observed through the usage of the service itself, so that the observations are censored, as they are inherently limited by available supply. Since the 1980s, various works on Censored Quantile Regression models have performed well under such conditions. Further, in the last two decades, several papers have proposed to implement these models flexibly through Neural Networks. However, the models in current works estimate the quantiles individually, thus incurring a computational overhead and ignoring valuable relationships between the quantiles. We address this gap by extending current Censored Quantile Regression models to learn multiple quantiles at once and apply these to synthetic baseline datasets and datasets from two shared mobility providers in the Copenhagen metropolitan area in Denmark. The results show that our extended models yield fewer quantile crossings and less computational overhead without compromising model performance.

Specifying utility functions is a key step towards applying the discrete choice framework for understanding the behaviour processes that govern user choices. However, identifying the utility function specifications that best model and explain the observed choices can be a very challenging and time-consuming task. This paper seeks to help modellers by leveraging the Bayesian framework and the concept of automatic relevance determination (ARD), in order to automatically determine an optimal utility function specification from an exponentially large set of possible specifications in a purely data-driven manner. Based on recent advances in approximate Bayesian inference, a doubly stochastic variational inference is developed, which allows the proposed MNL-ARD model to scale to very large and high-dimensional datasets. Using semi-artificial choice data, the proposed approach is shown to be able to accurately recover the true utility function specifications that govern the observed choices. Moreover, when applied to real choice data, MNL-ARD is able discover high quality specifications that can outperform previous ones from the literature according to multiple criteria, thereby demonstrating its practical applicability.

In transportation, nature, economy, environment, and many other settings, there are multiple simultaneous phenomena happening that are of interest to model and predict. Over the last few years, the traffic data that we have at our disposal have significantly increased, and we have truly entered the era of big data for transportation. Most existing travel demand prediction methods mainly focus on capturing recurrent mobility trends that relate to habitual/routine behavior, and on exploiting short-term correlations with recent observation patterns. However, valuable information that is often available in the form of unstructured data is neglected when attempting to improve forecasting results. Particularly, under non-recurrent conditions, such as large events, or incidents, we need much better models. In this paper, we explore time-series data and semantic information combinations using machine learning and deep learning techniques in the context of creating a prediction model that is able to capture in real-time future stressful situations of the studied transportation system. We apply the proposed approaches in event areas in New York using publicly available taxi data. We empirically show that the proposed models are able to significantly reduce the error in the forecasts. The importance of semantic information is highlighted in all presented methods and the final mean absolute error of our prediction is decreased by 23.8% for a three months testing period.

Spatiotemporal problems are ubiquitous and of vital importance in many research fields. Despite the potential already demonstrated by deep learning methods in modeling spatiotemporal data, typical approaches tend to focus solely on conditional expectations of the output variables being modeled. In this article, we propose a multioutput multiquantile deep learning approach for jointly modeling several conditional quantiles together with the conditional expectation as a way to provide a more complete "picture" of the predictive density in spatiotemporal problems. Using two large-scale data sets from the transportation domain, we empirically demonstrate that, by approaching the quantile regression problem from a multitask learning perspective, it is possible to solve the embarrassing quantile crossings problem while simultaneously significantly outperforming state-of-the-art quantile regression methods. Moreover, we show that jointly modeling the mean and several conditional quantiles not only provides a rich description about the predictive density that can capture heteroscedastic properties at a neglectable computational overhead but also leads to improved predictions of the conditional expectation due to the extra information and the regularization effect induced by the added quantiles.

Traffic speed data imputation is a fundamental challenge for data-driven transport analysis. In recent years, with the ubiquity of GPS-enabled devices and the widespread use of crowdsourcing alternatives for the collection of traffic data, transportation professionals increasingly look to such user-generated data for a good deal of analysis, planning, and decision support applications. However, due to the mechanics of the data collection process, crowdsourced traffic data such as probe-vehicle data is highly prone to missing observations, making accurate imputation crucial for the success of any application that makes use of that type of data. In this paper, we propose the use of multi-output Gaussian processes (GPs) to model the complex spatial and temporal patterns in crowdsourced traffic data. While the Bayesian nonparametric formalism of GPs allows us to model observation uncertainty, the multi-output extension based on convolution processes effectively enables us to capture complex spatial dependencies between nearby road segments. Using six months of crowdsourced traffic speed data or “probe vehicle data” for several locations in Copenhagen, the proposed approach is empirically shown to significantly outperform popular state-of-the-art imputation methods.

Over the last few years, deep learning has revolutionized the field of machine learning by dramatically improving the state-of-the-art in various domains. However, as the size of supervised artificial neural networks grows, typically so does the need for larger labeled datasets. Recently, crowdsourcing has established itself as an efficient and cost-effective solution for labeling large sets of data in a scalable manner, but it often requires aggregating labels from multiple noisy contributors with different levels of expertise. In this paper, we address the problem of learning deep neural networks from crowds. We begin by describing an EM algorithm for jointly learning the parameters of the network and the reliabilities of the annotators. Then, a novel general-purpose crowd layer is proposed, which allows us to train deep neural networks end-to-end, directly from the noisy labels of multiple annotators, using only backpropagation. We empirically show that the proposed approach is able to internally capture the reliability and biases of different annotators and achieve new state-of-the-art results for various crowdsourced datasets across different settings, namely classification, regression and sequence labeling.

Public special events, like sports games, concerts and festivals are well known to create disruptions in transportation systems, often catching the operators by surprise. Although these are usually planned well in advance, their impact is difficult to predict, even when organisers and transportation operators coordinate. The problem highly increases when several events happen concurrently. To solve these problems, costly processes, heavily reliant on manual search and personal experience, are usual practice in large cities like Singapore, London or Tokyo. This paper presents a Bayesian additive model with Gaussian process components that combines smart card records from public transport with context information about events that is continuously mined from the Web. We develop an efficient approximate inference algorithm using expectation propagation, which allows us to predict the total number of public transportation trips to the special event areas, thereby contributing to a more adaptive transportation system. Furthermore, for multiple concurrent event scenarios, the proposed algorithm is able to disaggregate gross trip counts into their most likely components related to specific events and routine behavior. Using real data from Singapore, we show that the presented model outperforms the best baseline model by up to 26 percent in R2 and also has explanatory power for its individual components.

The growing need to analyze large collections of documents has led to great developments in topic modeling. Since documents are frequently associated with other related variables, such as labels or ratings, much interest has been placed on supervised topic models. However, the nature of most annotation tasks, prone to ambiguity and noise, often with high volumes of documents, deem learning under a single-annotator assumption unrealistic or unpractical for most real-world applications. In this article, we propose two supervised topic models, one for classification and another for regression problems, which account for the heterogeneity and biases among different annotators that are encountered in practice when learning from crowds. We develop an efficient stochastic variational inference algorithm that is able to scale to very large datasets, and we empirically demonstrate the advantages of the proposed model over state-of-the-art approaches.

The growing need to analyze large collections of documents has led to great developments in topic modeling. Since documents are frequently associated with other related variables, such as labels or ratings, much interest has been placed on supervised topic models. However, the nature of most annotation tasks, prone to ambiguity and noise, often with high volumes of documents, deem learning under a single-annotator assumption unrealistic or unpractical for most real-world applications. In this paper, we propose a supervised topic model that accounts for the heterogeneity and biases among different annotators that are encountered in practice when learning from crowds. We develop an efficient stochastic variational inference algorithm that is able to scale to very large datasets, and we empirically demonstrate the advantages of the proposed model over state of the art approaches.

Public transport smartcard data can be used for detection of large crowds. By comparing statistics on habitual behavior (e.g., average by time of day), one can specifically identify nonhabitual crowds, which are often very problematic for transport systems. While habitual overcrowding (e.g., peak hour) is well understood both by traffic managers and travelers, nonhabitual overcrowding hotspots can become even more disruptive and unpleasant because they are generally unexpected. By quickly understanding such cases, a transport manager can react and mitigate transport system disruptions. We propose a probabilistic data analysis model that breaks each nonhabitual overcrowding hotspot into a set of explanatory components. The potential explanatory components are initially retrieved from social networks and special events websites and then processed through text-analysis techniques. Finally, for each such component, the probabilistic model estimates a specific share in the total overcrowding counts. We first validate with synthetic data and then test our model with real data from the public transport system (EZLink) of Singapore, focused on three case study areas. We demonstrate that it is able to generate explanations that are intuitively plausible and consistent both locally (correlation coefficient, i.e., CC, from 85% to 99% for the three areas) and globally (CC from 41.2% to 83.9%). This model is directly applicable to any other domain sensitive to crowd formation due to large social events (e.g., communications, water, energy, waste).

Learning from multiple annotators took a valuable step towards modelling data that does not fit the usual single annotator setting. However, multiple annotators sometimes offer varying degrees of expertise. When disagreements arise, the establishment of the correct label through trivial solutions such as majority voting may not be adequate, since without considering heterogeneity in the annotators, we risk generating a flawed model. In this paper, we extend GP classification in order to account for multiple annotators with different levels expertise. By explicitly handling uncertainty, Gaussian processes (GPs) provide a natural framework to build proper multiple-annotator models. We empirically show that our model significantly outperforms other commonly used approaches, such as majority voting, without a significant increase in the computational cost of approximate Bayesian inference. Furthermore, an active learning methodology is proposed, which is able to reduce annotation cost even further.