GR-TR, Extended Sensors, Extended sensors for assisted border-crossing

Use Case Category: <strong>Extended Sensors</strong><br> User Story: <strong>Extended sensors for assisted border-crossing</strong><br> Location: Greece - Turkey (GR-TR) cross-border corridor According to 3GPP TS 22.186 R16, Extended Sensors “enable the exchange of raw or processed data gathered through local sensors or live video data among vehicles, RSUs, devices of pedestrians and V2X application servers. The vehicles can enhance the perception of their environment beyond what their own sensors can detect and have a more holistic view of the local situation”. User Story: <strong>Extended sensors for assisted border-crossing</strong> By utilizing the detailed data provided by the CCAM enabled truck’s sensors (Lidar, radar, GPS, etc.) as well as the data from surrounding heterogeneous information sources such as traffic cameras, road side sensors, smart phones, wearables and more, increased intelligence can be created based on a cooperative awareness of the borders’ environment. The transmission of these data over reliable, ultra-fast and ultra-low latency 5G network connection combined with modern AI and predictive analytics techniques (at the edge) allows for the creation of a virtual environment of the driver enabling various added-value functionalities. As part of this use case the functionalities that will be showcased at the Greek / Turkish borders are: · Border inspection preparation based on predictive CCAM truck routing · Secure CCAM truck border crossing with increased inspection confidence · Increased border cooperative environment awareness for incoming vehicles · Increased border personnel safety The above functionalities will showcase a significant minimization of inspection times at all European “hard” borders through the collaboration feasible of different 5G network operators which could even offer “zero touch” inspection (no human intervention needed) in optimal cases. The same solution offers increased cooperative awareness for passing vehicles at the chaotic border-crossing environment and taking advantage of the CCAM functionalities of vehicles, such as automated braking, to prevent accidents involving border personnel (customs agents, police officers). This intelligent border control functionality may be realized through the following trial set-up. Data originating from the truck sensors in areas around the borders are transmitted over 5G networks and analysed in a cloud-based AI platform after fusion. Once a trajectory towards the border crossing is predicted, special measures may be taken to facilitate further exchange of information and immediate response to predicted events (e.g. the assisted driving application may be downloaded from the Cloud to the edge server to minimize latency, a slice may be provisioned towards a cloud server on the neighbouring county’s PLMN, etc.). An exchange of available information is commencing towards the border authorities via 5G network (mMTC type of communication from the truck OBU itself or even from the cargo which may be equipped with relevant sensors / transmitters (e.g. NB-IoT)) which will facilitate the border inspection and prepare the customs agents for the appropriate checks. All relevant information is transmitted to the edge / MEC servers available at the trial site where they are processed by the downloaded AI/ML platform instantiating this functionality. Additional information can be exchanged over the 5G networks of the neighbouring countries facilitating the acquisition of relevant information about the specific truck (e.g. driver’s information, travel history, cargo inventory, etc.) which could speed-up the control process. Extra security and control measures can be deployed which are controlled and managed through 5G networks such as drones, street cameras, thermal or x-ray cameras, etc. and which can feed large amounts of data (eMBB functionality) in a very short amount of time. In the case that all the acquired data from on-board as well as surrounding sensors / devices agree with the information that is fetched by national archives regarding this truck (and potentially its driver) and provided material (video, thermal imaging, x-ray imaging) clears the truck of any suspicion, then a case of “zero touch” inspection may be realized in which case the truck may be allowed to cross-the border without any manual inspection performed on it. Additionally, the data originating from other vehicles, road side infrastructure, smart phones and wearables may also be fused and analysed at the edge generating a “live” cooperative update of the surrounding environment which can be fed on to the vehicles navigation system, thus increasing the environmental awareness of the vehicle (covering blind spots, pedestrian locations and trajectories, assigned inspection lane by the authorities, etc.) and actively contributing to the safety of the border ground personnel (i.e. automated trajectory alignment or braking upon detection of a potential incident). In all cases, the same services continue being provided as the truck passes the border from the neighbouring country’s network, based on exchanged information in such inter-PLMN scenarios. Service continuity during the inter-PLMN HO is of utmost importance in such cases, and the existence of such intelligence deployed at the edge close to the border greatly facilitates continuous service by identifying imminent HO’s and helping the MNOs prepare for it based on the available information. This could lead to the provisioning of a roaming slice before the HO even takes place. To implement this use case a laptop onboard the truck will be acting as the UE/gateway that will connect truck and/or cargo devices/systems (e.g. additional sensors deployed in the cargo hold of the truck) to the rest of the system via 5G connectivity (and 4G / NB-IoT during testing &amp; development). These additional sensors are crucial in this case since they have the capability of raising alarms by cross-checking their data with nominal values. For instance, a thermal camera (or even CO2 sensor) installed in the cargo hold of the truck may provide indications of a human presence in the cargo hold (smuggling / human trafficking attempt) which will enable alerted reaction by the border officers upon the arrival of the truck at the border. Additional measures may take place in case contradicting information is gathered regarding a truck, in which case drones equipped with cameras for live feed may be deployed or thermal or x-ray imaging may be requested to rule out the possibility of smuggling goods and people. The AI based inspection functionality residing in the edge platform will fuse all available information from these heterogeneous sources (potentially originating from different 5G networks in the case of a cross-border scenario) and will locate potential inconsistencies, assigning a certain risk factor to each truck which will affect the degree (and thoroughness) to which border agents will perform a manual inspection. For the realization of this trial a single autonomous truck is needed equipped with additional sensors.用例类别：<strong>扩展传感器（Extended Sensors）</strong> 用户故事：<strong>用于边境通关辅助的扩展传感器</strong> 部署地点：希腊-土耳其（GR-TR）跨境通道 根据3GPP TS 22.186 R16标准，扩展传感器（Extended Sensors）支持在车辆、路侧单元（Road Side Unit, RSU）、行人设备与车万物联网（Vehicle-to-Everything, V2X）应用服务器之间，交换本地传感器采集的原始或处理后数据，以及实时视频流。车辆可借此突破自身传感器的感知局限，提升对周边环境的认知水平，更全面地掌握局部态势。 用户故事：<strong>用于边境通关辅助的扩展传感器</strong> 本场景通过利用支持协同互联自动驾驶（Cooperative Connected and Automated Mobility, CCAM）的卡车传感器（激光雷达、雷达、GPS等）采集的详细数据，以及交通摄像头、路侧传感器、智能手机、可穿戴设备等多源异构信息源的数据，基于边境环境的协同感知构建增强型智能系统。 依托可靠、超高速、超低时延的5G网络传输这些数据，并结合现代人工智能与边缘侧预测分析技术，可构建面向驾驶员的虚拟环境，实现多种增值功能。本用例将在希腊-土耳其边境展示以下功能： · 基于预测性CCAM卡车路由的边境检查准备 · 提升检查置信度的安全CCAM卡车边境通关 · 提升入境车辆的边境协同环境感知能力 · 提升边境工作人员的安全性 上述功能可通过不同5G网络运营商的协同合作，显著缩短欧洲所有实体管控边境的检查时长，在最优场景下甚至可实现"零接触"检查（无需人工干预）。该解决方案还可为混乱的边境通关环境中的过往车辆提供增强的协同感知能力，并利用车辆的CCAM功能（如自动制动），避免涉及边境工作人员（海关人员、警务人员）的事故发生。 此类智能边境管控功能可通过以下试验架构实现： 边境周边区域的卡车传感器数据通过5G网络传输，经融合后在云端AI平台进行分析。当预测到车辆驶向边境通关路线时，可采取特殊措施以促进信息交互并对预测事件做出即时响应（例如：将辅助驾驶应用从云端下载至边缘服务器以降低时延；可向邻国公共陆地移动网（Public Land Mobile Network, PLMN）的云服务器预配置网络切片等）。 通过5G网络（可采用卡车车载单元（On Board Unit, OBU）甚至搭载相关传感器/发射机的货物的海量机器类通信（Massive Machine Type Communications, mMTC）类型通信）向边境管理部门传输可用信息，以助力边境检查工作，协助海关人员做好相应检查准备。 所有相关信息将传输至试验场地的边缘/多接入边缘计算（Multi-Access Edge Computing, MEC）服务器，并由实例化该功能的已部署AI/ML平台进行处理。 可通过邻国的5G网络交换更多信息，以获取该卡车的相关信息（例如驾驶员信息、旅行历史、货物清单等），从而加快管控流程。 还可部署通过5G网络管控的额外安全与管控措施，例如无人机、街头摄像头、热成像或X射线摄像头等，这些设备可在极短时间内传输海量数据（增强移动宽带（enhanced Mobile Broadband, eMBB）功能）。 若车载及周边传感器/设备采集的所有数据均与国家档案中关于该卡车（及其驾驶员）的信息一致，且相关材料（视频、热成像、X射线成像）排除了该卡车的可疑性，则可实现"零接触"检查，允许卡车无需经过人工检查即可通关。 此外，来自其他车辆、路侧基础设施、智能手机及可穿戴设备的数据也可在边缘侧进行融合与分析，生成周边环境的"实时"协同更新信息，并推送至车辆的导航系统，从而提升车辆的环境感知能力（覆盖盲区、行人位置与轨迹、管理部门指定的检查车道等），并主动助力边境地面工作人员的安全（例如：在检测到潜在事件时自动调整轨迹或制动）。 在所有场景中，当卡车从邻国网络完成边境通行时，仍可基于跨PLMN场景下交换的信息持续提供相同服务。跨公共陆地移动网切换（Handover, HO）时的服务连续性在此类场景中至关重要，而部署在边境附近边缘的智能系统可通过识别即将发生的切换，并基于可用信息协助移动网络运营商（Mobile Network Operator, MNO）做好切换准备，极大地促进服务的连续性。此举甚至可在切换发生前预配置漫游切片。 为实现该用例，卡车上的笔记本电脑将充当用户设备（User Equipment, UE）/网关，通过5G连接（测试与开发阶段可采用4G/NB-IoT（Narrow Band Internet of Things））将卡车及/或货物的设备/系统（例如卡车货舱内部署的额外传感器）连接至系统其余部分。此类额外传感器至关重要，因为它们可通过将自身数据与标称值进行交叉比对来触发警报。例如，安装在卡车货舱内的热成像相机（甚至CO₂传感器）可提供货舱内存在人员的迹象（走私/人口贩运企图），使边境官员在卡车抵达边境时能够及时做出响应。 若收集到关于某辆卡车的矛盾信息，还可采取额外措施，例如部署搭载实时馈送摄像头的无人机，或请求进行热成像或X射线成像以排除走私货物与人员的可能性。 部署在边缘平台的基于AI的检查功能将融合来自这些异构源的所有可用信息（在跨境场景中可能来自不同的5G网络），并定位潜在的不一致性，为每辆卡车分配特定的风险等级，该等级将影响边境人员进行人工检查的程度（及细致程度）。 为实现该试验，需配备额外传感器的单台自动驾驶卡车。