• 【Application Story】BRAV-7520-WP: Edge AI Appliance For Tencent Autonomous Car


    JHCTECH Edge AI appliance for Tencent autonomous car


    Introduction

    Smart transportation is revolutionizing mobility in cities. In this March, Tencent’s driverless vehicles officially landed in the Shenzhen Intelligent Connected Transportation Test Zone. As one of Tencent's important suppliers in the field of edge computing, JHCTECH has provided leading edge computing product support for the Shenzhen Intelligent Connected Transportation Test Project.


    This Shenzhen test zone meets the demands of 14 testing projects, including the recognition and response of traffic signs, traffic lights, vehicles in front, obstacles, pedestrians, and other testing projects. It focuses on five elements of “vehicles, road, cloud, network, and map”, combining 5G, autonomous driving, AI and other new technologies.


    Autonomous driving, as one of the main application scenarios of artificial intelligence, has gained extensive attention. To achieve safe autonomous driving, the efficient collaboration of smart transportation and the Internet of Vehicles will be the cornerstone. Traditional driving needs the human brain and eyes to realize driving on the road, while driverless vehicles rely on artificial intelligence, edge computing, radar, monitoring equipment, and global positioning systems, etc.to work together, helping driverless vehicles respond correctly and timely.


    JHCTECH Solution

    Driverless vehicle uses in-vehicle sensors to perceive the traffic environment around the vehicle, and by sensing the obtained information of road, vehicle location and obstacles, the vehicle is able to make judgment and execution of vehicle’s steering, speed, and route. Therefore, in order to meet the harsh driving conditions, the key is to select a reliable edge computing product to realize the rapid processing of data from perception unit, and well realize the goal of safe autonomous driving.

    JHCTECH BRAV-7520A-WP, as a product designed for intelligent edge computing, was eventually selected for this project. The hardware system of driverless vehicles including three vital parts: perception, decision-making and control, while positioning, map, and prediction are auxiliary parts. Here is the C-V2X architecture of vehicle, road and cloud. 


    C-v2X architecture


    The vehicle itself will perceive the status information of the vehicle, such as speed, turning angle and roll, route, etc., as well as road environment perception, such as lidar, camera, millimeter wave radar and position information. C-V2X's roadside system and cloud platform can provide beyond the visual range function. When the vehicle is on the road, it is difficult to find information beyond the sensor range. However, through the C-V2X device, the relevant information will be sent and received, and the vehicle will receive the traffic information beyond visual range. 




     Deployment scenario of MEC platforms for C-V2X
    JHCTECH BRAV AI edge computing for roadside edge computing


    The in-vehicle computing unit MDC for L4 and L5 autonomous driving uses a dual architecture solution of CPU+GPU, and the OBU Communicator uses FPGA architecture. The MDC (BRAV-7520-WP) is connected to sensors such as lidar, millimeter-wave radar, and camera by a switch and fuse the data after deep learning, reasoning, calculating. At the same time, it can connect to the OBU communicator, to get the roadside and cloud information of C-V2X network, and connect to CAN bus. The in-vehicle drive-by-wire control system is connected through the CAN bus. The wire control system can realize the vehicle’s braking, steering, the start and stop of the engine, the control of gearbox, door and window, as well as the automatic control of sound, image and vibration warning systems. 

    JHCTECH BRAV AI edge computing for autonomous car


    Featured Product


    BRAV-7302

    AI edge computing BRAV-7302

    BRAV-7520-WP
    AI edge computing BRAV-7520-WP


    “IoT Computer, connecting the dots.” Now, JHCTECH has successfully cooperated with Tencent, Baidu and Alibaba on several self-driving car projects. In the future, we will continue to provide customers with excellent application solutions in intelligent transportation, machine vision, intelligent manufacturing, factory automation and other fields.  


Copyright:Shenzhen JHC Technology Development Co., Ltd.

ICP, Guangdong: 0903729* 网站建设合优网络