VITAL 5G

VITAL-5G Platform and Assets

Galati Riverport Available Assets

The Romanian VITAL-5G testbed is built on an end-to-end (E2E) 5G SA network, 3GPP Release 16 compliant, which provides 5G coverage in Galati with a bi-sector gNodeB that covers Navrom ships located in the port and an additional limited area near the port, on the Danube River. Moreover, the testbed also includes the Orange 5G Lab in Bucharest, providing 3rd parties with the opportunity to perform tests and experiments in both a T&L operational facility and an indoor/outdoor lab environment.
The 5G SA network in the Romanian testbed efficiently supports both eMBB and URLLC slices. The network KPIs measured on field are in the order of 600 Mbps for downlink (DL) throughput and 100 Mbps for uplink (UL) throughput for the eMBB slice and < 3.5 ms one way delay for the latency in the URLLC slice, using a subscriber connected to one of the slices.
All the Romanian 5G testbed’s components are fully deployed and integrated covering the needs to support the main capabilities and network features keys to respond to the 5G Network Applications deployment, starting with 5G RAN, Core, Transport, Virtualization and Orchestration and extending the infrastructure to support also VITAL-5G Platform. The Romanian testbed orchestrator, based on OSMv12, is integrated with the testbed VIMs, by using an Openstack platform deployed in High-Availability (HA) mode and a Kubernetes cluster.

The Romanian VITAL-5G testbed components are represented in Figure below, which also highlights the high availability deployment of transport network and virtual computing infrastructure, which guarantees seamless resiliency and application migration. The high availability and reliability tests conducted by ORO registered no 5G SA network issues for several months (more than 5 months, with at least one user permanently connected to the 5G SA network and generating traffic).

Available assets from Galati trial site

Available T&L Services and Network Applications in RO testbed
The design and implementation of new services from 3rd parties can be accelerated and facilitated by the use of Network Applications designed for UC2, which can be reused and combined across numerous vertical services and even across Use Cases. With the aid of assisted operations and navigation that utilise the IoT detection system and video cameras installed in Galati port and also on ships and cargoes, the vertical services that are used in UC2 as well as the corresponding Network Applications are used to ensure a safer operation in the port and to bring more security to the navigation of ships in the port of Galati. The 5G network will encompass all of this infrastructure to ensure quick transfer of data from the devices. Below a brief description of three types of vertical services that already use Network Applications developed in UC2, which can be used as examples for new 3rd party applications.

Service 1: Data-enabled assisted navigation

The service makes use of the Internet of Things sensing technology and NAVROM vessel’s and Galati port’s video cameras. For a specific data collection from the NAVROM vessel, VS9 is used. VA6 is used to classify the data stream, assign the appropriate slice (URLLC or mMTC) in accordance with the data supplied from the vessel, and provide interfaces for sending warnings and classifying events. Each incident can be quickly, clearly, and precisely categorized as either dangerous or non-threatening, as well as according to the type of dangerous occurrence. Instantaneous warnings are given to improve the vessel crew’s response time.


Service 2: Accurate electronic navigation maps creation


The base of the service is determining the ideal safe distance for a ship. The accuracy of electronic navigational charts is improved by fusing velocity, heading, water, and wind speed data utilizing the distributed sensor data ingestion, fusion, and post-processing services provided by VA2.
Service 3: Predictive maintenance and sanity checks applied on the sensor data for ship safety purposes
To obtain data for on-board diagnosis and monitoring, the service utilizes VS9 On Board Data Collection & Interfacing for Vessels. In order to reduce the influence of the human factor in risk assessment and the possibility of making poor decisions, VA3 uses the data for remote inspection and risk assessment.


Upgrades to RO testbed
UC2 revolves around the use of sensors and ML for improving business operations in the port. A network of IP cameras, interconnected with the 5G network using a 5G gateway, has been deployed and the related video streams are available for 3rd parties. Locally the connection to the gateway is done using 5G, Wi-Fi or by a LAN Cat 6 cable, depending on the location of the camera. Some areas experience a lot of shielding due to the large metal walls and the Wi-Fi signal is unreliable; for this reason, some cameras have been connected to the 5G gateway using cables. On the video stream received from the camera located on the vessel, we implement object detection. The video stream on which the detection is applied along with all other data that is captured is sent to the 5G network using the gateway and is gathered in the testbed servers where it is logged in an InfluxDB database or streamed to a logged in user (camera video), so that they can be easily consumed by new Network Applications developed by 3rd parties. Moreover, 3rd parties can also access real-time or historical data collected from a monitoring station that measures the air quality in its proximity, as well as a wide set of engine data. The installation of additional equipment, as needed for the execution of additional 3rd party experiments, can be discussed, evaluated and negotiated on a per-experiment basis.

Asset Type of asset Availability (from) Description
Distributed sensor data ingestion fusion & post-processing (VA2) NetApp September 2022 In charge of data ingestion, fusion and post-processing functionality from distributed data sources.
Predictive maintenance (VA3) NetApp 2023 Responsible for modelling various machine learning techniques to achieve functions like automated labelling, outlier detection, trace back analysis, graphical representations, predictions/forecasts for the near-future and therefore offer advanced diagnostics to aid decision making and specifically for the case of predictive maintenance. The results of these techniques may be used for monitoring purposes as well as by other applications.
Positions of ships and obstacles Data September 2022 Open datasets with historical data on the positions of ships and obstacles to be available on Zenodo.
Sensor data (weather and air quality) Data September 2022 Open datasets with historical records on the sensor data (weather and air quality) to be available on Zenodo.
Pre-configured 5G slices 5G testbed 2023 5G slices provisioned during the third-party experimenters’ trials. Different network slices (eMBB and uRLLC) to be provisioned in a pre-configured manner.
New 5G slices to be provisioned 5G testbed 2023 5G slices provisioned during the third-party experimenters’ trials. Different network slices (eMBB and uRLCC) to be provisioned in a dynamic manner.
Network KPIs from monitoring platform on 5G testbed Results 2023 Measurement logs on network KPIs from executed experiments provided through the Vital-5G monitoring platform.
Risk analysis data, alarms, relevant analytics info Results 2023 Logs on risk analysis, alarms, and other relevant analytics information from executed experiments provided through the Vital-5G monitoring platform.