Privacy-preserving transportation mode identification (TMI) is among the key challenges toward future intelligent transportation systems. With recent developments in federated learning (FL), crowdsourcing has emerged as a promising cost-effective data source for training powerful TMI classifiers without compromising users’ data privacy. However, existing TMI approaches have relied heavily on the availability of transportation mode labels, which is often limited in real-world applications. While recent semisupervised studies have partially addressed this issue by assigning pseudolabels to unlabeled data, such practice often degrades classification performance as more unlabeled data are incorporated. In response to this issue, we present a semisupervised FL scheme for TMI termed mean teacher semisupervised FL (MTSSFL). MTSSFL trains a deep neural network ensemble under a novel semisupervised FL framework, achieving highly accurate and privacy-protected crowdsourced TMI without depending on the availability of massive labeled data. MTSSFL introducesconsistency updatingto insert the global model in the gradient updates of the local models that only have unlabeled data to improve their training. We also devisemean-teacher-averaging, a secure parameter aggregation mechanism that further boosts the global model’s TMI performance without requiring additional training. Our extensive case studies on a real-world data set demonstrate that MTSSFL’s classification accuracy is merely 1.1% lower than the state-of-the-art semisupervised TMI approach while being the only one to satisfy FL’s privacy-preserving constraints. In addition, MTSSFL can achieve high accuracy with less training overhead due to the proposed semisupervised learning design.

Estimated time of arrival (ETA) is one of the critical services offered by navigation and hailing providers. The majority of existing solutions approach ETA as a regression problem and leverage GPS trajectories for estimation. However, the travel time fluctuates greatly between different trips, making simple regression methods skewed. Additionally, these methods are incapable of conducting estimation in practice because the trajectories of future trips are unknown. To jointly tackle these problems, we propose a novel Categorical approximate method to Estimate Time of Arrival (CatETA). Specifically, we formulate the ETA problem as a classification problem and label it with the average time of each category. To eliminate bias in categorical labeling, we approximate travel time using the weighted average of different classes in the testing stage. Then, we design a network structure that extracts the spatio-temporal features of link sequences and integrates a set of global information. Furthermore, we merge link sequences according to network topology and graph embedding to alleviate the computational burden associated with large-scale link networks. Comprehensive experiments on real-world datasets demonstrate that CatETA considerably improves the estimation performance and significantly reduces computational effort.

Traffic speed prediction is among the foundations of advanced traffic management and the gradual deployment of internet of things sensors is empowering data-driven approaches for the prediction. Nonetheless, existing research studies mainly focus on short-term traffic prediction that covers up to one hour forecast into the future. Previous long-term prediction approaches experience error accumulation, exposure bias, or generate future data of low granularity. In this paper, a novel data-driven, long-term, high-granularity traffic speed prediction approach is proposed based on recent development of graph deep learning techniques. The proposed model utilizes a predictor-regularizer architecture to embed the spatial-temporal data correlation of traffic dynamics in the prediction process. Graph convolutions are widely adopted in both sub-networks for geometrical latent information extraction and reconstruction. To assess the performance of the proposed approach, comprehensive case studies are conducted on real-world datasets and consistent improvements can be observed over baselines. This work is among the pioneering efforts on network-wide long-term traffic speed prediction. The design principles of the proposed approach can serve as a reference point for future transportation research leveraging deep learning.

Autonomous vehicle (AV) integration poses a significant challenge for intelligent transportation systems (ITSs). The ability to automatically coordinate complex AV operations at scale is crucial for advancing the quality of core transportation services, such as ride-sharing and parcel delivery. However, existing studies have only considered either of these two services independently from the other, disregarding the potential benefits of their combined optimization. To address this open problem, we design an autonomous vehicle intelligent system (AVIS) providing joint ride-sharing and parcel delivery services under realistic ride and route constraints. We formulate the joint optimization problem through the scope of mixed-integer linear programming and solve it using the Lagrangian dual decomposition method to ensure scalability. We conduct extensive case studies to evaluate the performance of the proposed AVIS and its constituting components. Our experimental results demonstrate that AVIS can effectively provide both ride-sharing and parcel delivery services while satisfying service requests in transportation networks of various scales. In addition, the distributed method is shown to generate near-optimal solutions in reduced computation time.

Intelligent transportation management requires not only statistical information on users' mobility patterns, but also knowledge of their corresponding transportation modes. While GPS trajectories can be readily obtained from GPS sensors found in modern smartphones and vehicles, these massive geospatial data are neither automatically annotated nor segmented by transportation mode, subsequently complicating transportation mode identification. In addition, predictive uncertainty caused by the learned model parameters or variable noise in GPS sensor readings typically remains unaccounted for. To jointly address the above issues, we propose a Bayesian deep learning framework for unsupervised GPS trajectory segmentation. After unlabeled GPS trajectories are preprocessed into sequences of motion features, they are used in unsupervised training of a channel-calibrated temporal convolutional neural network for timestep-level transportation mode identification. At test time, we approximate variational inference via Monte Carlo dropout sampling, leveraging the mean and variance of the predicted distributions to classify each input timestep and estimate its predictive uncertainty, respectively. The proposed approach outperforms both its non-Bayesian variant and established GPS trajectory segmentation baselines on Microsoft's Geolife dataset without using any labels.

The deployment of Global Positioning System (GPS) sensors in modern smartphones and wearable devices has enabled the acquisition of high-coverage urban trajectories. Extracting knowledge from such diverse spatiotemporal data is essential for optimizing intelligent transportation system operations. Yet a deeper understanding of users’ mobility patterns also requires identifying their associated transportation modes. Combined with growing privacy concerns, the considerable effort involved in manual data annotation means that GPS trajectories are in reality not labeled by transportation mode. This poses a significant challenge for machine learning classifiers, which often perform best when trained on large amounts of labeled data. As such, this paper investigates a wide range of time series augmentation methods aiming to improve the real-world applicability of transportation mode identification. In our extensive experiments on Microsoft’s Geolife dataset, both discrete wavelet transform and flip augmentations pushed the transportation mode identification accuracy of a convolutional neural network from 85.1% to 87.3% and 87.2%, respectively.

Federated Learning (FL) is a privacy-oriented framework that allows distributed edge devices to jointly train a shared global model without transmitting their sensed data to centralized servers. FL aims to balance the naturally conflicting objectives of obtaining massive amounts of data while protecting sensitive information. However, the data stored locally on each edge device are typically not independent and identically distributed (non-IID). Such data heterogeneity poses a severe statistical challenge for the optimization and convergence of the global model. In response to this issue, we propose Federated One-vs-All (FedOVA), an efficient FL algorithm that first decomposes a multi-class classification problem into more straightforward binary classification problems and then combines their respective outputs using ensemble learning. Experiments on several public datasets show that FedOVA achieves higher accuracy and faster convergence than federated averaging and data sharing. Furthermore, our approach can support practical settings with a large number of clients (up to 1000 clients) in FL.