End-to-end perception and trajectory prediction from raw sensor data is one of the key capabilities for autonomous driving. Modular pipelines restrict information flow and can amplify upstream errors. Recent query-based, fully differentiable perception-and-prediction (PnP) models mitigate these issues, yet the complementarity of cameras and LiDAR in the query-space has not been sufficiently explored. Models often rely on fusion schemes that introduce heuristic alignment and discrete selection steps which prevent full utilization of available information and can introduce unwanted bias. We propose Li-ViP3D++, a query-based multimodal PnP framework that introduces Query-Gated Deformable Fusion (QGDF) to integrate multi-view RGB and LiDAR in query space. QGDF 1) aggregates image evidence via masked attention across cameras and feature levels, 2) extracts LiDAR context through fully differentiable BEV sampling with learned per-query offsets, and 3) applies query-conditioned gating to adaptively weight visual and geometric cues per agent. The resulting architecture jointly optimizes detection, tracking, and multi-hypothesis trajectory forecasting in a single end-to-end model. On nuScenes, Li-ViP3D++ improves end-to-end behavior and detection quality, achieving higher EPA (0.505) and mAP (0.616) while substantially reducing false positives (FP ratio 0.069), and it is faster than the prior Li-ViP3D variant (139.82 ms vs. 145.91 ms). Additional experiments were performed to evaluate the impact of the reduced resolution of RGB inputs and missing HD maps on the behavior of the model. The results of the experiments indicate that query-space, fully differentiable camera–LiDAR fusion can increase the robustness of end-to-end PnP without sacrificing deployability. © 2026 The Authors.

Autonomous vehicles (AVs) represent a technology with significant social and environmental benefits. By reducing dependence on the human factor, which is responsible for 94% of the 1.35 million annual traffic deaths globally, AVs have the potential to increase road safety and save lives. Complementary technologies, such as Vehicle-to-Everything (V2X) communication, further enhance traffic management, reducing congestion by up to 40% and improving energy efficiency with fuel savings of up to 15%. However, V2X systems are particularly vulnerable to cyber attacks, such as spoofing, which injects false information, disrupting the flow of traffic and compromising the safety of AVs. This paper proposes an innovative framework for detecting and mitigating spoofing attacks in V2X communications. The solution combines Direction of Arrival (DoA) estimation with advanced object detection algorithms, such as YOLOv8, to identify anomalous signals and locate malicious transmitters. By integrating Artificial Intelligence (AI) techniques, the framework makes it possible to accurately classify attackers and select customized countermeasures, ensuring greater network reliability and security. The simulation results demonstrate the framework's effectiveness in various dynamic scenarios using data from antenna arrays and camera-based object detection. In addition, they highlight the importance of sensor data fusion to improve anomaly detection accuracy, optimize decision-making processes in AVs, and enable robust cross-validation of transmitted information. © 2026 The Authors.

Intelligent Transportation Systems (ITS) have significantly improved transportation quality by using applications capable of monitoring, managing, and improving the transportation system. However, the large number of devices required to provide data to ITS applications has become a challenge in recent years, particularly the high installation and maintenance costs made broad deployment impracticable. Despite several advances in smart city research and the internet of things (IoT), research on ITS is still in the early stages. In this sense, to improve data collection and maintenance strategies for ITS systems, this article proposes a virtual infrastructure model based on data reuse, mainly autonomous vehicle (AV) data, to support ITS applications. It presents design choices and challenges for deploying a virtual infrastructure based on Beyond 5G (B5G) communication and data reuse, followed by developing a proof of concept of an AV data acquisition system evaluated through simulation. The results show that the extra data collection module results in a 1.1% increase in total memory usage with direct sensor collection and a 2.6% increase with application performance management (APM) data collection on the reference hardware. This data reuse setup can significantly improve ITS data challenges with minimal impact on current technology stack on the Autonomous vehicles currently in circulation. © 2025 The Authors. This work is licensed under a Creative Commons Attribution 4.0 License. For more information, see https://creativecommons.org/licenses/by/4.0/

As urban mobility increasingly integrates micromobility solutions such as bicycles, innovative road safety solutions have become a priority for these Vulnerable Road Users (VRUs). This work explores the use of Augmented Reality (AR) and Vehicle-to-Everything (V2X) communications to enhance cyclist safety at intersections with obstructed visibility. We propose an AR-based system that enables cyclists to visualize occluded vehicles using an 'X-ray' vision effect, leveraging real-time V2X messages and edge computing for low-latency interaction. The system architecture integrates multiple communication technologies, including 5G and ITS-G5, ensuring reliable transmission of road user data. To evaluate the system's feasibility, we conducted a real-world demonstration in the Aveiro Tech City Living Lab (ATCLL) platform using a Microsoft HoloLens 2 AR headset and an NVIDIA Jetson-based object detection pipeline. The system was tested in a real environment and results show that: (1) response time for total system latency falls within the 300 ms safety threshold defined by ETSI (which assures system safety); (2) the system operates on 3.75 FPS; and (3) increasing the video cameras's frame rate does not significantly affect resource and power consumption. © 2025 IEEE.

Vehicle-to-everything (V2X) communication plays a crucial role in enabling collaborative intelligence among autonomous vehicles, by utilizing paradigms such as Federated Learning (FL). However, the dynamic and decentralized nature of vehicular networks poses challenges, particularly in maintaining model convergence and robustness without centralized coordination. In this paper, we propose a dynamic Practical Byzantine Fault Tolerant consensus mechanism tailored for decentralized FL in vehicular environments. Our method optimizes federated averaging by addressing the high mobility and intermittent connectivity of vehicles. Through simulations, we evaluate its performance against baseline method, demonstrating improved resilience, efficiency, and adaptability in the presence of adversarial conditions. © 2025 IEEE.

Urban traffic environments are rapidly evolving with the adoption of connected and automated vehicles (CAVs), yet challenges remain regarding interactions between these vehicles and vulnerable road users (VRUs). This paper introduces the concept of an Ethical Social Robot Moderator (ESRM) that facilitates coordination and communication in mixed-traffic contexts. By consolidating insights from research on robot trust and tele-operation, vehicular communication systems, and ethical frameworks for autonomous driving, the proposed ESRM aims to reduce collision risks, enhances cooperation among heterogeneous road users, and provides transparent decision-making based on well-defined moral principles. The proposal is a structured design for the ESRM, extending simulation strategies in the CARLA environment with detailed metrics, and integrating references from prior literature to illustrate how these social robots can bridge multiple technological domains. This work serves as a unifying contribution to a broader field that spans robotics, communication networks, and ethical AI in urban mobility. © 2025 IEEE.

Augmented Reality (AR) and edge computing can be used to enhance Vulnerable Road User (VRU) safety, leveraging a greater degree of interaction with the user and presenting real-time warning notifications, a use case that demands low latency for critical warnings. However, a new Multi-Access Edge Computing (MEC) system that distributes computational needs across edge nodes within the infrastructure is needed to meet these low-latency requirements. This paper proposes a microservice architecture designed to process data from various sources, including awareness and perception messages from road users and smart city sensors, through multiple communications technologies such as ITS-GS and 5G. This work employs peer-to-peer decentralized communications to perform seamless inte-gration and real-time processing, ensuring timely and relevant warnings for VRUs. Real-world road tests conducted in the Aveiro Tech City Living Lab in Aveiro, Portugal, and in commercial V2X equipment in the Autonomous Driving Test Field Baden-Wurttemberg in Karlsruhe, Germany, showed a maximum end-to-end latency of 110 ms, showcasing the system's capabilities in real-life demanding use cases and the system's interoperability. © 2025 IEEE.

This paper investigates the effect of varying Vulnerable Road User Awareness Message (VAM) sampling rates on the network performance in Vehicular Ad-hoc Networks (VANETs), specifically focusing on e-bikes. Basic traffic scenarios were analyzed using the Artery simulation framework, which integrates SUMO for microscopic traffic modeling, OMNeT++ for network simulation, and the ETSI ITS-G5 protocol stack via Vanetza. To enhance realism, recorded cycling speeds from 15 e-bike riders were integrated into the simulation model. The results show that strategic adjustments of VAM transmission intervals improve channel efficiency to maintain high levels of timely attention with reduced channel load. We provide insights for optimizing the trade-off between communication frequency and performance for future deployments of Cooperative Intelligent Transportation Systems (C-ITS) involving Vulnerable Road Users. © 2025 IEEE.

Vehicle-to-Everything (V2X) communication enables vehicles to exchange information with other road users and infrastructure to improve road safety and traffic efficiency. However, how Vulnerable Road Users (VRUs), can effectively benefit from this technology remains an open question. One emerging concept is the use of social robots equipped with V2X capabilities to support traffic orchestration and interaction with VRUs. While technically promising, such systems raise important human-centered questions, particularly regarding public acceptance. In this work, we present a first step toward understanding societal attitudes by conducting a qualitative preliminary study. Through interviews with pedestrians, we explore their perceptions of using a V2X-enabled social robot in traffic context. © 2025 IEEE.

The development of autonomous vehicles and vehicular communications (V2X) promises to significantly enhance traffic safety and efficiency. However, challenges remain in ensuring safe interactions between autonomous vehicles and vulnerable road users (VRUs). We promote the idea to use a social robot as interface between social interaction and V2X. We propose to use such a robot for crash prediction: the social robot, equipped with an RGB-D camera and V2X-communication capabilities, gathers data on pedestrians' trajectories and vehicles' movement within a shared environment. The data can then be used to predict possible crashes. In this ongoing work, we present a framework that integrates the basic functionality to implement such an approach. To test the approach a data set consisting of videos of crossing pedestrians and V2X data of an eBike was collected. The system effectively converts the trajectories of pedestrians and vehicles into a shared coordinate frame, enabling precise detection of potential collisions. The preliminary findings show potential for a novel method for crash prediction. © 2025 IEEE.