Search for: All records

Accurate vehicle trajectory prediction at signalized intersections is crucial for autonomous driving and traffic management systems, yet remains challenging due to complex multi-modal interactions between vehicles, pedestrians, and infrastructure elements. Existing approaches inadequately integrate the diverse factors influencing intersection dynamics, including vehicle characteristics, environmental context, and human-vehicle interactions. We present Enhanced Knowledge-Informed Generative Adversarial Network (EKI-GAN), a comprehensive framework that addresses these limitations through enhanced multi-modal encoding and advanced attention mechanisms. Our approach introduces a dual-stream architecture with six specialized encoders that comprehensively capture vehicle factors and environmental information for robust trajectory prediction. The framework employs Multi-Dimensional Attention Pooling Network (MAP-Net), which extends conventional attention through multi-dimensional feature extraction, stabilized computation, and adaptive gating to model complex vehicular interactions. Extensive experiments on the SinD dataset demonstrate state-of-the-art performance, achieving ADE of 0.09 and FDE of 0.19 for 9-second predictions, with particularly notable improvements in extended temporal prediction scenarios. 

Accurate detection of Vulnerable Road Users (VRUs) by roadside sensors poses significant challenges due to sparse labeled data and multi-modal sensor fusion complexity. This paper proposes a hierarchical multi-modal fusion framework that integrates traditional 3D object detection methods with deep learning-based approaches. Our method combines the robustness of conventional detection pipelines with neural network adaptability through a novel complementarity strategy. The framework employs DAB-DETR for 2D detection, traditional 3D clustering algorithms, and PointPillars architecture, unified through confidence-based fusion. Experimental evaluation on the Intersection Safety Challenge (ISC) dataset demonstrates the effectiveness of our approach with overall mAP@0.5 of 49.12% compared to 36% for conventional methods and 15.99% for deep learning approaches. Notable improvements include vehicle detection (46.09% vs. 39.12%) and VRU detection, especially for children (30.71% vs. 3.61%). Our hierarchical fusion strategy successfully leverages complementary strengths of traditional classification accuracy and deep learning localization precision. The solution based on this detection method won an award in the ISC Stage 1B competition, validating its effectiveness for data-constrained roadside VRU detection scenarios. 

This paper presents an integrated approach to optimizing microgrid management and electric truck logistics for transportation research. The experiment involves a 100 kW solar photovoltaic system, a 500 kWh battery energy storage system, the electric demand of a commercial building, and a heavy-duty vehicle charging system. The study aims to demonstrate how synchronized optimization of a microgrid control algorithm and a truck route algorithm can reduce overall system costs. The microgrid management system is designed to meet the constraints and requirements of a commercial electric truck charging scheduler. This integrated approach is an improvement over previous systems as it uses the scheduler’s outputs—such as time frames and energy requirements—as constraints for microgrid management. The truck scheduling algorithm iteratively learned to optimize charging times, ensuring that charging occurs during low-cost periods or when renewable energy is available. The electric vehicle scheduler adjusted truck arrival times based on the availability of clean energy sources, creating a feedback loop that continuously improves cost efficiency. Results indicated significant cost savings, with electric utility costs for electric vehicle (EV) charging being only 0% to 20% of the original value while the transportation system is only 23% to 64%, compared with the baseline scenario without the co-optimization framework. These findings suggest that the proposed integrated approach can effectively reduce costs and improve the efficiency of microgrid and electric truck operations. Uncoordinated charging schedules leads to a higher power demand than a well-organized battery electric truck (BET) dispatching strategy. Optimizing truck charging times and energy needs based on microgrid conditions can significantly reduce electricity and transportation costs. 

The accuracy in estimating energy consumption of electric buses is of significant importance for formulating electric bus route planning and charging schedules. Current approaches for estimating energy consumption of electric buses can be categorized into three major types: physics-driven models, statistical models, and deep learning methods. This study develops an Enhanced Bilayer Long Short-Term Memory (EBLSTM) method for energy consumption estimation of electric buses considering real-time passenger load, along with the Powertrain-based Physical Model (PPM) and Scale Tractive Power-based Model (STPM). A linear interpolation model is first implemented to reconstruct the bus trajectory (i.e., position, speed, and acceleration) from 0.1 Hz to 1 Hz for model calibration and verification. A tanh activation function is designed to mitigate fluctuations in the estimation results of the traditional LSTM method. The genetic algorithm, least mean square method and grid search approach were conducted respectively to calibrate the above three different models. Numerical results indicate that the EBLSTM method achieves the best estimation performance, with a verification Root Mean Square Percentage Error (RMSPE) of 0.68 %. In contrast, the RMSPEs of the PPM and STPM models are 0.90 % and 1.13 %, respectively. Furthermore, both qualitative and quantitative analysis were conducted to examine the impacts of initial SOC, travel time, and the heterogeneous characteristic of different bus datasets on the accuracy of the three models. 

Most existing signal control systems utilize fixed-location sensors (e.g., loop detectors or roadside cameras), which are constrained by limited spatial coverage and have relatively high installation and maintenance costs. As an alternative method for vehicle detection and traffic management at signalized intersections, it is possible to use vehicle-probe data, collected from smartphones, navigational aids, GNSS receivers, and other types of mobile devices. In this paper, we propose a cost-effective vehicle-probe-based signal management technology to leverage the widely available probe data for traffic signal control. To demonstrate its effectiveness, we then develop a microscopic simulation to compare its performance to a state-of-the-practice system. A candidate “unbalanced” intersection is identified in the City of Riverside and replicated in a microscopic simulation. The performance of the vehicle-probe-based optimized signal control plan is evaluated under the calibrated traffic demand. The simulation results demonstrated that it significantly reduces the length of the eastbound peak-hour queue from 800 m to under 50 m. The average travel time was reduced by 69% for the eastbound and 40% for the westbound, while maintaining a similar level in the northbound and southbound. The vehicles’ emissions were decreased by 32%–52% due to the mitigation of congestion in the major westbound-eastbound direction, resulting in fewer vehicles being halted in queues. Overall, the proposed vehicle-probe-based signal management technology demonstrated significant potential in enhancing traffic efficiency and reducing environmental impact. It provides a low-cost, scalable, and sustainable solution that can be potentially applied to both fixed-time and adaptive signalized intersections. 

Positioning integrity refers to the trust in the performance of a navigation system. Accurate and reliable position information is needed to meet the requirements of connected and automated vehicle applications, particularly in safety-critical scenarios. Receiver autonomous integrity monitoring (IM) and its variants have been widely studied for global navigation satellite system-based vehicle positioning, often fused with kinematic (e.g., odometry) and perception sensors (e.g., cameras). However, IM for cooperative positioning solutions that leverage vehicle-to-everything (V2X) communication has received comparatively limited attention. This article reviews existing research in the field of positioning IM and identifies various research gaps. Particular attention has been placed on identifying research that highlights cooperative-IM methods. It also examines key automotive safety standards and public V2X datasets to map current research priorities and uncover critical gaps. Finally, the article outlines promising future directions, highlighting research topics aimed at advancing and benchmarking positioning integrity. 

Many cities across the world are looking to use technology and innovation to improve the overall efficiency and safety for their residents. At the heart of these smart-city plans, a variety of intelligent transportation system technologies can be used to improve safety, enhance mobility measures (e.g., traffic flow), and minimize environmental impacts of a city’s mobility ecosystem. Early implementations of these ITS technologies often take place in affluent cities, where there are many funding opportunities and suitable areas for deployment. However, it is critical that we also develop smart city solutions that are focused on improving conditions of disadvantaged and environmental justice communities, whose residents have suffered the most from unmitigated urban sprawl and its environmental and health impacts. As a leading example, Inland Southern California has grown to be one of the largest hubs of goods movement in the world. Numerous logistics facilities such as warehouses, rail facilities, and truck depots have rapidly spread throughout these communities, with the local residents bearing a disproportionate burden of truck traffic, poor air quality, and adverse health effects. Further, the majority of residents have lower-wage jobs and very few mobility options, other than low-end personal car ownership. To improve this situation, UC Riverside researchers have focused their smart city research on these impacted communities, finding innovative solutions to eco-friendly traffic management, developing better-shared (electric) mobility solutions for the community, improving freight movements, and enhancing the transition to vehicle electrification. Numerous research and development projects are currently underway in Inland Southern California, spanning advanced smart city modeling and impact analysis, community outreach events, and real-world technology demonstrations. This chapter describes several of these ITS solutions and their potential for improving many cities around the world. 

The emergence of battery electric trucks (BETs) in recent years has shown great promise in reducing greenhouse gas (GHG) emissions in urban freight logistics. However, designing a customer-oriented dispatching strategy for a BET fleet is more complex than traditional vehicle routing problems (VRP) due to several constraints, such as limited driving range, potential need for en route recharging, and long recharging times. Also, in practice, the uncertain travel times in urban transportation network may lead to the violation of scheduled customer time windows and impact overall energy consumption. To better utilize the BET fleet, this paper introduces a robust BET dispatching problem with backhauls and time windows under travel time uncertainty, which aims to minimize the overall fleet energy consumption while also minimizing the risk of violating customer time window. A mathematical optimization model based on novel route-related sets is developed, and an adaptive large neighborhood search (ALNS) metaheuristic algorithm is used to find robust dispatching solutions. Based on real-world data from a truck fleet in San Bernardino County, California, a simulation study is conducted to demonstrate the robustness of the solutions obtained by the proposed method. Moreover, a sensitivity analysis with respect to uncertainty parameters is performed to assess the trade-off between the overall fleet energy consumption and the robustness of the solutions.