Robust and accurate perception of the environment is crucial for autonomous driving systems. Radar sensors offer unique advantages over LiDAR and cameras, such as robustness to adverse weather, long detection range, and instantaneous velocity measurement via the Doppler effect, while being cost-effective. However, radar point clouds present challenges, including sparsity, noise, and ambiguous measurements. In this paper, we propose a graph-based framework for clustering and tracking unlabeled radar point clouds using velocity and spatial information through connected-component analysis. Our approach aggregates multi-sweep radar data into a unified global coordinate frame, enhancing spatial density and temporal continuity. We introduce a centroid-based temporal tracking algorithm that maintains consistent cluster identities across frames and multiple camera views. To objectively evaluate tracking performance, we develop the Multi-Attribute Temporal Consistency Precision (MATC-Precision) metric, which independently assesses temporal consistency using radar cross section and dynamic properties, features orthogonal to the clustering process. Experiments on the nuScenes dataset demonstrated that our method achieved over 80% tracking precision for clusters tracked across three consecutive frames. Qualitative results validated the robustness of our approach in complex urban environments, effectively handling occlusions, multi-camera views, and static and dynamic objects. This framework provides a scalable solution for reliable radar-based perception in autonomous driving.
This study investigated the effects of centerline and edgeline presence on driver speed selection and lateral positioning behavior on a low-speed urban collector street. The study was conducted along a two-lane two-way street segment that had been recently reconstructed, which allowed for the centerline and edgeline markings to be installed in separate phases. Cameras were used to record vehicle speed and lateral position data at five locations along the corridor during each striping phase, which included: (1) no striping, (2) centerline only striping, and (3) centerline plus edgeline striping. Linear regression models were developed to assess the impacts of striping configuration and roadway cross-sectional characteristics on speed and lateral position. The absence of any longitudinal striping resulted in speeds that were 0.8 mph lower than when a centerline was present and 2.3 mph lower than when both a centerline and edgeline were present. In terms of lateral position, drivers were positioned closest to the center of the roadway when both the centerline and edgeline were present and shifted approximately 1-ft further to the right when no edgeline was present (with or without a centerline). Furthermore, the combined use of centerlines and edgelines resulted in a more uniform vehicle lane position along the corridor, with reduced lateral variability during roadway width transitions or when vehicles were parked along the curb. Collectively, the findings provide preliminary evidence that the exclusion of longitudinal pavement markings contributes to lower speeds and generally more cautious travel along two-lane, two-way urban collector streets.
This study investigates the chemical composition and field-aging behavior of 12 spray-on rejuvenators applied to asphalt pavement test sections at the Minnesota Road Research facility. The products (seven bio-based, one hybrid, and four petroleum-based) were classified by gas chromatography–mass into eight chemical groups and applied to pavements constructed with PG 58S-28 (unmodified) and PG 58H-34 (polymer-modified) binders. Over 24 months, binder samples from control and treated sections were evaluated using saturates, aromatics, resins, and asphaltenes fractionation, Colloidal Instability Index, and Fourier transform infrared (FTIR) spectroscopy. The CII for bio-based products was highly variable, likely owing to polar carboxylate functionalities from the fatty acids bonding to and precipitating from the heptane solvent as asphaltenes. Petroleum-based products dissolved in heptane and were later separated into maltene fractions. For bio-based formulations, strong ester absorbances overlapped the carbonyl (C = O) region in FTIR spectra, requiring a modified FTIR approach to accurately distinguish ester content from oxidation products. Accurate evaluation of oxidative aging using typical FTIR analyses without such differentiation is unreliable when bio-oils are present. Using the supplier-recommended application rate resulted in very different residual oil masses for each product. Chemical properties of aged residues were generally improved with higher oil dosages. PG 58S-28 exhibited greater treatment differentiation owing to higher oxidative susceptibility. Overall, results highlight the importance of product formulation, binder type, and tailored analysis methods in selecting effective spray-on rejuvenators for pavement preservation.
Activity-based models (ABMs) are essential for travel demand forecasting, but traditional rule-based and econometric approaches for activity scheduling are often complex, rigid and difficult to scale. This paper presents a data-driven framework to predict 24-h human activity schedules using deep learning and behavioural segmentation. A novel, cluster-informed bidirectional long short-term memory network is proposed, which learns directly from daily activity sequences represented as 96 discrete 15-min time steps. The model’s key innovations include the integration of unsupervised k-prototypes clustering to create a latent behavioural signature for each user and a progressive minority training strategy with focal loss to address severe class imbalance. Trained on data from the Halifax Travel Activity Survey, the model’s effectiveness is validated through extensive ablation studies. Results show that the cluster-informed, bidirectional architecture significantly outperforms alternatives. While achieving high time step-level accuracy (macro F1: 59.9%), the model’s primary strength lies in generating functionally plausible schedules. A novel grouped sequence alignment method distance metric reveals that many prediction errors are semantically minor substitutions within functionally similar activity categories (e.g., ‘home chore’ for ‘staying at home’). The framework successfully replicates aggregate temporal patterns for critical activities such as work and school, making it a robust and practical tool for transportation planners to simulate policy impacts with greater behavioural realism.
Using six US hurricane-strength tropical-cyclone events as the independent transfer units, this study evaluates advisory-time warnings of port service disruption from public automatic identification system (AIS) activity, National Hurricane Center advisories, and port-network data. The 19-port panel contains 3312 repeated advisory-lead-port rows, but leave-one-event-out validation is limited by the six-event ceiling. For a 10% AIS-based service-loss label, the full calibrated logistic model has a mean area under the receiver operating characteristic curve (AUROC; where 0.5 denotes random ranking) of 0.702 (95% event-bootstrap interval: [0.487, 0.864]) and an F1 score (the harmonic mean of precision and recall) of 0.540 (95% interval: [0.370, 0.752]). Thus, absolute performance is uncertain. The more robust comparative result is that forecast exposure alone has AUROC = 0.427, below random ranking, whereas adding a recent port state and network substitutability improves the ranking. Several simple rules match or exceed the model’s F1 score only by warning almost every sample; calibrated probabilities instead support an explicit precision-recall-warning-burden tradeoff. Sally 2020 is nearly separable because static network variables distinguish four positive ports from two negative ones, whereas Ida 2021 fails through a corridor-level ranking reversal. The framework is, therefore, a promising, interpretable decision-support prototype, not evidence of universal predictive performance.
