
Urban traffic in developing countries is characterized by heterogeneous vehicle composition, weak lane discipline, and frequent stop-and-go operation, which may not be adequately represented by standardized driving cycles. This study develops and validates a local driving cycle for passenger cars under heterogeneous urban traffic conditions in Jodhpur, India. A total of 61,344 s of second-by-second vehicle activity data were collected along a 12.9 km urban corridor using OBD-II and GPS instrumentation. The processed data were segmented into micro-trips and characterized using nine kinematic parameters. Principal Component Analysis (PCA) and K-means clustering were applied to identify dominant driving patterns, and representative micro-trips nearest to cluster centroids were assembled to construct the Jodhpur Driving Cycle (JDC).The JDC has a duration of 1216 s, covers 6.93 km, and has an average speed of 20.51 km/h. Its modal distribution comprises 12.75% idling, 36.84% acceleration, 29.69% deceleration, and 20.72% cruising, indicating pronounced transient operation. Validation against the complete processed real-world dataset using statistical/kinematic comparison and Speed-Acceleration Frequency Distribution (SAFD) analysis showed relative differences of 7.12% and −2.82% for average speed and speed standard deviation, respectively, with an overall normalized SAFD difference of 0.1237.Comparison with the Modified Indian Driving Cycle (MIDC), international legislative cycles, and Indian city-specific cycles revealed distinct speed characteristics and modal distributions. The JDC provides a representative condensed description of observed driving characteristics, while the proposed data-driven framework supports city-specific emission, energy-consumption, and transport analyses in developing urban areas.
Severe and fatal road traffic crashes represent a small yet disproportionately high-impact share of crash events, posing persistent challenges for predictive modeling due to extreme class imbalance and structural sparsity. Most existing crash severity models treat severe outcomes primarily as minority labels within imbalanced datasets, often overlooking the structural distinctiveness of rare high-impact crashes. This study presents a scoping review and conceptual synthesis of peer-reviewed studies on crash severity modelling, rare-event learning, class-imbalance mitigation, and anomaly-aware transportation safety approaches. The review searched Scopus and Web of Science for journal articles published between January 2010 and February 2026. After screening 270 records, 74 studies were included and analyzed. The review critically evaluates how the existing methods handle rare-event conditions, minority class metrics, and structural model capacity. Findings indicate that while imbalance mitigation techniques improve minority recall, they do not fully capture the structural rarity of severe crashes. Anomaly-aware and hybrid extreme value frameworks provide a complementary perspective by identifying high-risk configurations embedded within complex spatiotemporal systems, although only three of the 74 studies applied anomaly detection to a severity-related outcome. Accordingly, this review proposes an anomaly-guided rare-event crash severity modeling framework integrating spatiotemporal architectures with structural anomaly detection and rare-event-aware optimization intended to improve predictive reliability and policy relevance, offered as a roadmap for empirical testing.
Origin–Destination (OD) matrices are essential for understanding travel behavior and supporting short- and long-term planning in complex transit systems. This study examines the New York City subway network using two complementary data sources: Metropolitan Transportation Authority (MTA) turnstile counts and Subway Wi-Fi based OD observations. Turnstile data provide complete coverage of entries and exits but no OD detail, whereas Wi-Fi data capture OD patterns but are sparse and biased. To address these limitations, we apply a Generative Adversarial Network (GAN) framework to complete sparse Wi-Fi OD flows using MTA ridership estimation data as structural guidance, producing fused OD matrices for model training. We assess source compatibility at station and OD-pair levels and train FuseODNet, a graph-based predictor that learns spatial demand structure and reconstructs four hourly OD matrices from each 4-hour turnstile input block. In the routine evaluation, FuseODNet attains the lowest normalized RMSE among the external comparators. Its template ensemble prior accounts for most of the routine RMSE performance, while the residual component reduces absolute and high-flow error. A weekday-hour average attains the lowest mean absolute error. For the Hurricane Ida case, FuseODNet has the lowest OD-share RMSE point estimate among all evaluated methods against a September 2021 MTA weekday-hour average benchmark. Yankees and synthetic evaluations examine performance under a localized surge and controlled stress scenarios. The enhanced OD forecasts can support applications ranging from service scheduling to disruption management and emergency evacuation planning, demonstrating the value of combining generative data fusion with graph based learning for urban transit systems.
The transition from bounded robotaxi pilots to broader service in complex cities depends on a credible public authorization pathway that can review evidence, assign accountability, and determine when expansion is justified. Existing research identifies major governance, safety, accessibility, and legitimacy concerns, while offering limited guidance on authorization sequencing, barrier-instrument translation, and the revision of gate conditions after affected stakeholders identify operational failure modes. This paper develops a stage-gated framework for robotaxi authorization in New York City using fuzzy DEMATEL, the Best–Worst Method (BWM), and a Policy-QFD mapping, followed by an independent qualitative review based on 24 semi-structured stakeholder interviews across seven groups. The institutional analysis identifies two principal upstream domains: Policy & Governance and Safety, Assurance & Service Reliability. The interview layer produced 92 participant-level evidence units and showed that some safeguards with modest institutional or QFD weights nevertheless become non-deferrable gate conditions, especially end-to-end accessibility and privacy/security controls. Stakeholder evidence also refined accountability and redress, labor-chain disclosure and transition planning, site- and time-specific curb controls, public-value and neighborhood-burden evaluation, and the granularity of gate evidence. The paper contributes a staged public decision framework and demonstrates how stakeholder-identified failure modes can tighten gate conditions without overwriting the institutional dependency structure.
Background: Intelligent Transportation Systems (ITS) increasingly draw on artificial intelligence for prediction, classification, and control tasks. Quantum computing and quantum machine learning (QML) have emerged as candidate tools, though current hardware remains in the noisy intermediate-scale quantum (NISQ) era. Objective: This review maps the literature applying quantum or quantum-inspired AI to ITS, with attention to road-safety applications. Methods: A PRISMA-ScR-adapted scoping review searched five databases (Web of Science, ScienceDirect, arXiv, Scopus, IEEE Xplore). Screening was conducted jointly by the review author and a co-author against structured eligibility criteria; an independent, criteria-blind AI-assisted second pass assessed reliability (substantial agreement, kappa = 0.77). From 271 records identified, 50 were synthesized: 41 primary studies and 9 background reviews. A five-item quality checklist was applied to the 15 primary studies with confirmed full-text access. Results: The 41 primary studies span five application themes, led by collision/accident detection and avoidance (22 studies), ahead of traffic-state forecasting, ITS operations, electric-vehicle systems, and network security. Reported methods include quantum neural networks, support vector machines, variational circuits, game theory, annealing/QAOA, and federated learning. Quantum and hybrid approaches outperformed classical baselines on structured, interaction-heavy tasks using markedly fewer parameters, though fixed quantum-kernel methods showed recurring overfitting; most studies ran on simulators rather than physical hardware. Conclusion: Quantum AI for ITS is a young, rapidly growing field (mostly 2024–2026) in which road safety is the largest application cluster. Reported advantage centers on parameter efficiency, not raw accuracy. Key gaps include near-absent physical-hardware validation, standardized benchmarks, and fully independent dual screening.