arXiv subjects

Tareq Alsaleh

Publications and source records attributed to Tareq Alsaleh.

6 recordsLinked to original sources

arXiv2026

XARA: A Multimodal Agentic System for eXtended Reality Experiment Generation

Controlled experiments are essential for advancing transportation research, yet designing eXtended Reality (XR) environments for behavioural studies remains technically demanding and inaccessible to many domain researchers. Current workflows require proficiency in game engines such as Unity or Unreal Engine, imposing significant barriers to rapid scenario prototyping and iterative experiment design. This paper presents XARA, a novel system that integrates a dual large language model (LLM) architecture with an immersive XR environment to enable voice driven experiment generation for transportation research. XARA employs a pipeline in which spoken natural language commands are first transcribed and then interpreted into structured commands that manipulate a Unity based XR scene in real time. The system supports object creation, spatial manipulation, animation control, and scene persistence through a modular architecture. By lowering the technical threshold for XR experiment creation, XARA enables transportation researchers to rapidly prototype driving, pedestrian, and urban design scenarios using intuitive voice commands and hand gestures within an immersive 3D environment. This paper describes the system architecture, the voice to command pipeline, and discusses potential applications across behavioural travel experiments, safety studies, human factors research, and policy testing.

cs.HC↗

arXiv2026

Mobility Behaviour of Immigrants in Canada: Analyzing Mode Choice Using GPS Panel Data and Mixed Logit Models

We examine these relationships using a panel dataset of more than 80,000 trip observations from 100 participants through a custom-built mobile application. A joint revealed preference (RP) and stated preference (SP) framework is used to estimate multinomial logit (MNL) and mixed logit (MXL) models. The level of integration is represented through a composite index capturing economic, social, civic, and health dimensions of integration. Results indicate two distinct patterns. First, the estimated models suggest that new immigrants in the sample exhibit lower sensitivity to in-vehicle travel time than Canadian-born respondents. The mixed logit specification suggests that the value of travel time for the sampled immigrants is approximately 66% lower than that of Canadian-born residents, with a immigrant-to-Canadian-born ratio of 0.34 that is consistent across both MXL specifications. Second, higher levels of integration are associated with reduced transit use and greater car reliance. A one standard deviation increase in the integration index decreases the probability of choosing public transit by approximately five percentage points. The joint RP-SP specification allows the inclusion of emerging e-mobility alternatives not yet observed in revealed behaviour; these face no inherent preference penalty, competing purely on their level-of-service attributes. Out-of-sample validation using five-fold cross-validation produces a mean prediction accuracy between 80% and 82% across model specifications. The findings suggest that transit policies in immigrant-receiving cities could prioritize service quality improvements, particularly reductions in access time, which are approximately three times more effective than fare reductions in shifting immigrants toward transit use.

econ.EM↗

arXiv2026

Exclusive and Shared Electric Flying Taxis: Evidence on Modal Shares, Stated Reasons, and Modal Shifts

This study examines travelers' preferences for electric flying taxi services in the United Arab Emirates (UAE) under varying travel conditions and service configurations. A stated preference (SP) survey of 213 respondents was conducted to analyze behavior across multiple transport alternatives, including private vehicles, public transport, ground taxis, and both shared and exclusive flying taxi services. The analysis considered key attributes such as travel time and cost, along with contextual factors including travel distance, congestion conditions, day of travel, and trip purpose. In addition, follow-up questions were used to capture the underlying reasons for mode choice and to assess potential modal shifts under changes in travel conditions. The results show that flying taxi services account for 22.6% of total responses, with higher shares under congested conditions and declining shares as travel distance increases. Clear differences are observed between shared and exclusive services. Shared flying taxis achieve higher modal shares and exhibit greater responsiveness to travel conditions, particularly at moderate distances, during weekdays, and for leisure trips. In contrast, exclusive flying taxis maintain lower modal shares, decline with increasing travel distance, and are more associated with business and weekend travel. The modal shift analysis further indicates that ground taxi users exhibit the highest propensity to switch to shared flying taxi services, particularly under cost increases. These findings highlight the importance of pricing and service design in promoting the adoption of shared flying taxi services as a more sustainable mobility option. In particular, maintaining affordable shared services, ensuring clear price differentiation from exclusive services, and prioritizing deployment in congested corridors and medium-distance travel markets can enhance adoption.

cs.CY↗

arXiv2026

Traffic Collisions: Temporal Patterns and Severity-Weighted Hotspot Analysis

Understanding traffic collision patterns is of high importance for effective road safety planning in fast-growing urban environments. This study examines the temporal and spatial patterns of traffic collisions in Dubai, UAE, with a particular focus on collision severity. To this end, traffic collision records from November 2024 to June 2025 were analyzed to examine hourly, daily, and monthly variations in collision frequency and severity for both overall traffic collisions and pedestrian-related accidents. Temporal associations with severity were evaluated using chi-square tests and Cramer's V, while spatial patterns were analyzed using severity-weighted hotspot analysis based on the Getis-Ord Gi* statistic, complemented by inverse distance weighting (IDW) interpolation. The results show a clear temporal variation in overall collision frequency and severity, with higher collision frequencies during evening and nighttime periods with 44% higher probability of high-severity outcomes at night compared to the afternoon. On the other hand, pedestrian-related accidents showed a distinct temporal profile, characterized by higher occurrence during late-evening hours and relatively limited variation across days of the week and months. Spatial analysis identified statistically significant severity hotspots for overall collisions in the northern and northwestern parts of Dubai and along the Al Ain-Dubai Highway, while pedestrian severity hotspots were concentrated near industrial areas in the southwestern region. Several policy measures are proposed based on the findings including, reducing nighttime speed limits, enhancing automated enforcement, improving roadway lighting, and implementing pedestrian-focused treatments in statistically significant hotspots.

cs.CE↗

arXiv2025

Towards Locally Deployable Fine-Tuned Causal Large Language Models for Mode Choice Behaviour

This study investigates the adoption of open-access, locally deployable causal large language models (LLMs) for travel mode choice prediction and introduces LiTransMC, the first fine-tuned causal LLM developed for this task. We systematically benchmark eleven open-access LLMs (1-12B parameters) across three stated and revealed preference datasets, testing 396 configurations and generating over 79,000 mode choice decisions. Beyond predictive accuracy, we evaluate models generated reasoning using BERTopic for topic modelling and a novel Explanation Strength Index, providing the first structured analysis of how LLMs articulate decision factors in alignment with behavioural theory. LiTransMC, fine-tuned using parameter efficient and loss masking strategy, achieved a weighted F1 score of 0.6845 and a Jensen-Shannon Divergence of 0.000245, surpassing both untuned local models and larger proprietary systems, including GPT-4o with advanced persona inference and embedding-based loading, while also outperforming classical mode choice methods such as discrete choice models and machine learning classifiers for the same dataset. This dual improvement, i.e., high instant-level accuracy and near-perfect distributional calibration, demonstrates the feasibility of creating specialist, locally deployable LLMs that integrate prediction and interpretability. Through combining structured behavioural prediction with natural language reasoning, this work unlocks the potential for conversational, multi-task transport models capable of supporting agent-based simulations, policy testing, and behavioural insight generation. These findings establish a pathway for transforming general purpose LLMs into specialized and explainable tools for transportation research and policy formulation, while maintaining privacy, reducing cost, and broadening access through local deployment.

cs.CL↗

arXiv2024

Simulation Models for Sustainable, Resilient, and Optimized Global Electric Vehicles Supply Chain

While the transition to electric vehicles (EVs) is essential for decarbonizing the transportation system, the production and distribution of EVs entail substantial carbon costs. To ensure these emissions are accurately accounted for and effectively mitigated, this research introduces a digital twin of the EV's supply chain, addressing a critical gap in current EV life cycle analyses and providing the first comprehensive quantification of its environmental sustainability and resilience. This simulation model replicates global market dynamics and captures the complexity and uncertainty of the EV supply chain, enabling a thorough evaluation of its carbon footprint, sustainability, resilience, and what-if counterfactual scenarios for alternative market structures. The results reveal that average supply chain emissions range from 6.42 to 6.94 Kg e-CO2/KWh across different battery technologies. Additionally, the mass flow analysis shows unbalanced dependencies at all supply phases, with one geographical region significantly dominating the supply chain structure, highlighting the current supply chain architecture's low resilience and high vulnerability. In light of these findings, the study introduces an optimization model for hub and resource allocation configuration, effectively reducing vulnerability levels and supply chain emissions by up to 80%.

math.OC↗