TrialsNet Newsletter #6 - Trials

TrialsNet: TRials supported by Smart Networks beyond 5G Six Months Digest #6 October 2025 (Trials) www.trialsnet.eu TRIALSNET TRIALS TrialsNet successfully reached its final Milestone (M6 - Final Trials Execution), by satisfactorily terminating its trial activity, and by timely submitting deliverables D3.3, D4.3, D5.3 - Use Cases final implementation and trials results for ITSS/ eHE /CTE domain. These deliverables describe a comprehensive account of each project use case, including its final implementation, pre-trial/s activities, trial execution, KPIs and KVIs collection and analysis, as well as a thorough overview of sub-projects activities and outcome, fully integrated into the project as a whole. 22 trials of the 13 project use cases and 56 of the Open Call sub-projects were conducted, involving thousands of end-users. A truly outstanding result! In this newsletter, we showcase the variety of project trials performed during 2025, highlighting their key results and major achievements. UC1 – SMART CROWD MONITORING Madrid: This use case aims to enhance safety and situational awareness in crowded public venues through the use of AI-based surveillance and robotics over a 5G/B5G network. The objective was to validate the integration of intelligent sensing, analytics, and communication technologies to improve real-time security management during large events. The trial took place on March 16th, 2025, at the Movistar Arena in Madrid, during a basketball game with around 2,000 attendees. The setup included highresolution PTZ and fixed cameras, a LiDAR sensor, and three autonomous ground robots equipped with AI-based analytics for event detection such as crowding, vandalism, and abnormal behavior. All devices were connected to a 5G Rel-17 standalone Non-Public Network provided by Ericsson and remotely supervised from a mobile intelligent Security Operations Centre (iSOC). The system architecture integrated two 5G Customer Premises Equipment (CPEs) for stable uplink and downlink performance and Genetec software at the iSOC for alarm handling, device control, and event management. The field trial followed a series of pre-lab tests focusing on communication stability, AI event detection accuracy, and system responsiveness. The trial was executed in a joint setup including two sub-trials: i) a first trial focused on crowd monitoring through static infrastructure using AI-enabled cameras and LiDAR to detect security incidents such as abandoned objects, intrusions, or aggressive behavior; ii) a second trial including the operation of three robots performing programmed patrol routes and tele-operated missions for security and public assistance. Throughout the event, service availability reached 100%, ensuring continuous video streaming and remote-control capability. The network latency requirement was defined at 12 ms, meeting operational needs for real-time video and control communication. Key Value Indicators (KVIs) were collected through surveys with five security personnel involved in the trial. Results showed positive feedback across categories such as trustworthiness, resilience, security, and safety, all with mean values above 3 points out of 5. The system was perceived as secure and useful for supporting field operations, with minor concerns on AI accuracy and early-stage connectivity stability. The trial successfully validated the feasibility of using 5G-enabled AI robotics and sensing technologies for real-time crowd monitoring and event safety management in large-scale public venues. Iași: This use case focuses on real-time crowd monitoring during public events in Iași, Romania, aiming to enhance safety and situational awareness in urban environments. The trials took place on Ștefan cel Mare Boulevard and in front of the Palace of Culture, two of the city’s busiest pedestrian areas, often used for large gatherings and public celebrations. The system uses high-resolution cameras mounted on lighting poles and connected through a 5G network provided by Orange Romania (ORO). The solution enables detection of crowd density, people movement, and potential anomalies such as unauthorized vehicles entering pedestrian zones. Video feeds are transmitted via Nokia Fastmile 5G routers and Cisco switches, securely connected to ORO’s 5G Lab in Bucharest through a Virtual Private Network (VPN). The data are processed in real time on servers hosted by the Technical University Gheorghe Asachi (TUIASI), using AIbased analytics implemented within the Nvidia DeepStream framework. The analytics provide dynamic people counting, heatmaps of density, and automatic alerts in case of restricted area breaches, all displayed on an operational dashboard developed under the Savant software framework. The first field test activities were organized in December 2024 under a 5G Non-Standalone configuration. Following network upgrades and system refinements, the final trial took place in May 2025 under a full 5G Standalone setup with local edge computing and URLLC slicing support. This setup enabled lowlatency transmission and stable processing, confirming the system’s robustness in real urban environments. Pre-trial stages included tests with single, multi, and large-scale camera configurations, culminating in a full deployment with up to 13 cameras streaming simultaneously. The results demonstrated that the infrastructure can handle multiple highresolution video streams efficiently, with throughput and latency values meeting the expected performance requirements. The system proved scalable and resilient, maintaining stable operation even under increased data loads. The overall evaluation showed strong performance in terms of trust, security, and operational efficiency, with positive feedback from participants representing local authorities, safety services, academia, and SMEs. The trial confirmed the feasibility and reliability of 5G-enabled smart crowd monitoring in public safety contexts, providing valuable insights for future deployments in smart city environments. UC2 – PROACTIVE PUBLIC INFRASTRUCTURE ASSETS MANAGEMENT This use case demonstrates how 5G can support proactive maintenance and condition monitoring of public infrastructure assets using autonomous systems and AI-based analytics. The goal was to validate a complete workflow where mobile robots, drones, and fixed cameras capture visual and sensor data for real-time analysis, enabling early detection of infrastructure degradation such as surface cracks or pavement damage. The solution integrated advanced AI models with a unified visualization platform to support remote monitoring, inspection, and preventive maintenance planning. In Athens, two field trials were conducted to evaluate the solution in both airport and urban municipal contexts. The first trial took place at Athens International Airport (AIA), where an autonomous mobile robot equipped with cameras and sensors performed inspection tasks in a controlled outdoor environment. The robot transmitted live video and sensor data via a 5G connection to the wi.MOVE platform, where AI algorithms analyzed the footage and provided immediate visualization of detected surface defects. The trial confirmed the capability of the system to operate in a real airport setting, validating the integration between the video analytics pipeline, data transmission modules, and the visualization dashboard. The second trial was held in the Technopolis area of the City of Athens, extending the system validation to an urban setting. The robotic platform, connected over 5G, was remotely teleoperated to inspect pedestrian areas and road surfaces, while the wi.MOVE dashboard provided live views, interactive mapping, and automated detection of infrastructure issues. This setup demonstrated the scalability and flexibility of the system, showing that it can adapt to different operational environments, from airport aprons to city streets. Both trials were coordinated by WINGS in close collaboration with AIA and the City of Athens IT company (DAEM). They jointly contributed to validating the technical maturity of the platform and the benefits of 5G connectivity for high-bandwidth video streaming, lowlatency remote control, and AI-assisted infrastructure inspection. The trials also highlighted the current limitations of commercial 5G networks when handling large data volumes, reinforcing the need for future evolution toward 5G Standalone and 6G systems that can better support ultra-low-latency, high-throughput use cases. Through these demonstrations, the Public Infrastructure Assets Management use case proved the feasibility of a connected, data-driven maintenance approach that combines robotics, AI, and 5G to improve the efficiency and sustainability of infrastructure management in both airport and urban environments. UC3 – AUTONOMOUS APRON This use case focuses on demonstrating autonomous ground handling operations at an airport, showcasing how robotic systems, digital twins, and B5G communication can enhance safety, coordination, and efficiency. The main goal was to automate baggage transport using an Automated Guided Vehicle (AGV) connected through a low-latency communication link and monitored remotely via a digital twin interface. The trial took place at Athens International Airport (AIA) in a secured area of the apron dedicated to testing autonomous mobility. The AGV operated under real airport conditions, performing luggage transport tasks along predefined routes. The system combined realtime localization, obstacle detection, and video streaming, ensuring that the vehicle could move safely and efficiently across the operational zone while being continuously supervised through the digital platform. The solution was built around the Wi.SUPPLY platform, developed by WINGS, which enabled operators to remotely control and monitor the AGV through a visualization dashboard that integrated sensor data, camera feeds, and navigation information. The robot’s perception system included cameras, LiDAR, and localization sensors, supporting precise movement and safe operation. Communication between the robotic unit and the control platform relied on 5G connectivity, which provided stable and low-latency links for control commands, data transmission, and live video feeds. During the trial, the AGV successfully performed autonomous transport operations between predefined waypoints, using real-time data to adjust its trajectory and maintain situational awareness. The digital twin offered an accurate visualization of the robot’s position and status, while the operators could monitor telemetry and video in real time. The tests demonstrated how advanced connectivity and automation can improve airport logistics by minimizing manual handling and optimizing the coordination between human operators and robotic systems. The trial was organized by WINGS in collaboration with AIA. Through this cooperation, the demonstration confirmed the technical feasibility of deploying autonomous robotic systems in active airport environments. The system proved capable of integrating artificial intelligence, robotic automation, and 5G communication into a unified operational workflow. Overall, the Athens trial validated that autonomous ground vehicles supported by real-time digital twins and high-speed connectivity can significantly improve the safety and efficiency of airport apron operations. The experiment provided a realistic example of how next-generation networks and intelligent control platforms can support future airport automation and ground service innovation. UC4 – SMART TRAFFIC MANAGEMENT This use case focused on improving urban mobility and road safety at the Podu Roș intersection in Iași through the use of AI-driven traffic monitoring over a 5G network. The goal was to enhance real-time visibility of road conditions, identify potential hazards involving pedestrians and vehicles, and support faster responses in complex traffic environments. The trial was carried out by TUIASI and ORO and implemented a distributed architecture combining high-resolution video cameras, edge computing, and a centralized dashboard. Six cameras installed at key points of the intersection provided continuous video streams that were processed locally on the edge infrastructure using an AI model based on the YOLOv7 framework. The system detected vehicles, pedestrians, and other road users, analyzed their trajectories, and generated alerts for abnormal or unsafe behaviors, such as cars entering restricted zones or close interactions with pedestrians. The application was built on a software stack similar to that used in UC1, with dedicated modules for video processing, metadata handling, and visualization. All processed data were published to a Kafka-based backend and displayed through a web dashboard that allowed operators to monitor the intersection in real time. The architecture integrated IMEC’s zero-touch orchestration and dual connectivity through Multipath TCP, which ensured stable transmission, load balancing, and high throughput across both NSA and SA network setups. Field test/trial campaigns were organized to validate the solution under different network conditions. As first, field test took place in December 2024 using a 5G NSA setup and confirmed the correct operation of the AI pipeline and data flow between the cameras, edge servers, and dashboard. Then, the trial was performed in May 2025, leveraged the upgraded 5G SA network and tested the system under full operational conditions with multiple camera inputs and increased traffic activity. During the trial, the solution demonstrated stable real-time video analytics and consistent network performance. The system successfully identified vehicles and pedestrians, estimated trajectories, and detected traffic violations with minimal delay between capture and visualization. The 5G SA configuration allowed efficient processing of large video streams while maintaining reliable connectivity between the field devices and the edge infrastructure. Feedback from the operators confirmed that the system improved their ability to observe, interpret, and react to evolving traffic situations. The combination of AI-based analytics and low-latency connectivity provided a more detailed overview of the intersection, supporting better decision-making for traffic management and safety. The evaluation also included an assessment of user perception regarding trust, security, resilience, societal value, and operational efficiency. Overall, participants expressed a high level of satisfaction with the solution’s performance, reliability, and usefulness for public safety applications. The results indicate strong confidence in the system’s capacity to operate continuously in real-world conditions and to contribute meaningfully to the development of intelligent urban mobility solutions. Through this trial, the UC4 demonstration confirmed that 5G-enabled AI video analytics can effectively support smart city applications. The Iași deployment showed that nextgeneration connectivity and distributed intelligence can be combined to improve road safety and traffic efficiency, laying the groundwork for scalable urban mobility management systems in future 5G and B5G networks. UC6 – SMART CROWD MONITORING AND EMERGENCY RESCUE IN POPULATED AREA This use case has the ambition to demonstrate the viability of a coordinated response in a densely populated area as well as more effective and digitally traceable pre-hospital care by first responders in the event of MCI. Through the use of cutting edge technologies, this UC aims at achieving more effective first-responder communication, quicker and more efficient triage, and pre-hospital treatment. The Emergency Rescue in Populated Area trial took place in Madrid on March 16th, 2025, during two basketball games at the Movistar Arena, in an operational environment involving real users. The venue was hosting more than 10.000 people, among spectators and workers. In such a setting an efficient use of rescue resources is key, in the case of an emergency. The trial execution was possible thanks to the collaboration between UC3M (managing the trial), ERC (providing the infrastructure), TID (allowing the bandwidth usage), and Movistar Arena (giving access to the venue and helping with the operative logistics), as well as WINGS (providing the smartwatches and the STARLIT++ platform) and CERTH (providing the evacuation path computation). The trial was executed in the context of a trial event, including trialing of UC1 - Smart Crowd Monitoring and UC10 - Immersive fan engagement. Through this trial, the solution has been validated, leveraging a Stand Alone 5G mmWave deployment over the 26GHz band. Key Values generated by the solution have been measured through a questionnaire specifically designed for the scope, resulting in positive evaluation. Similarly KPIs have been measured, assessing the suitability of the network infrastructure to the solution. The Athens trial of UC6 - Mass Casualty Incident (MCI) and Emergency Rescue in Populated Area demonstrated how 5G/B5G-enabled technologies can transform emergency response operations in densely populated areas. Conducted in June 2025 at the Technopolis area, the trial simulated a mass casualty incident (MCI) caused by an earthquake and building collapse during a public event, testing advanced coordination, triage, and rescue capabilities. Using the WINGS STARLIT++ platform, robots equipped with cameras scanned the area and transmitted real-time video over the 5G network. AIpowered analytics detected victims, estimated movement and respiration, and enabled semi-automated triaging. Meanwhile, first responders used smartwatches to monitor vital signs—ECG, SpO₂, temperature, and blood pressure—transmitted directly to the command platform. The trial achieved excellent network performance and responsiveness, with latency below 100 ms and high accuracy and reliability across all measured KPIs, including throughput, precision, and location accuracy (<2 m). Moreover, user satisfaction and trust indicators (KVIs) were rated highly (above 4/5), confirming the solution’s effectiveness and usability in real operational conditions. This successful demonstration validated how 5G/B5G networks and AI-driven tools can significantly improve situational awareness, triage efficiency, and coordination during mass casualty emergencies, supporting faster and more informed decision-making for first responders in crowded urban environments. UC7 – REMOTE PROCTORING The UC7 trial successfully demonstrated the feasibility of remote surgical guidance using 5G-enabled telepresence systems. Conducted between Pisa and Massa, the trial connected expert cardiologists with surgical trainees via VR headsets and smart glasses, enabling real-time collaboration during cardiac procedures. The application achieved ultra-low latency, with round-trip times measured at 10 milliseconds—well below the 100 ms threshold. Network throughput was also impressive, reaching 800 Mbps for downlink and 200 Mbps for uplink. Connectivity remained stable throughout the trial, with 100% service availability ensuring uninterrupted communication. Pre-trial activities revealed that XRinterfaces significantly improved task execution time compared to traditional screens, particularly under low-latency conditions. Surgeons reported a notably enhanced user experience and greater realism in telepresence. Satisfaction scores increased from 3.4 to 4.6 on a 5-point scale when switching from Wi-Fi to 5G. Overall, the trial validated the system’s usability and scalability, underscoring the transformative potential of 5G in remote surgical training and support. Looking ahead, future 6G networks are expected to further elevate immersive experiences and enable broader adoption. UC8 – SMART AMBULANCE UC8 focused on enabling real-time, bi-directional communication between mobile emergency units and hospital hubs. The smart ambulance was equipped with AR headsets, portable echocardiography systems, and a 5G hub to support remote diagnostics and expert guidance during patient transport. During the trial held on June 27 in Pisa, the system delivered robust performance, achieving 900 Mbps downlink, 200 Mbps uplink, and 10 ms latency—meeting all key performance indicators. Communication remained stable even at speeds up to 20 km/h, with 100% service availability throughout the test. Interaction between emergency operators and remote experts was enhanced by a gesturebased system, which allowed for real-time guidance and improved diagnostic accuracy. User feedback was overwhelmingly positive, with all participants rating the system highly for both usability and communication stability. The trial confirmed the viability of 5G in supporting advanced mobile healthcare scenarios and laid the groundwork for scalable, real-time emergency response systems.