Intel's cloud-to-end solutions promote innovation in smart transportation

Recently, the 3rd SAIC-GM/Pan Asia Automotive Software Quality Conference was held in Shanghai. With the theme of "Gathering Wisdom and Innovation to Drive the Future Together", this conference attracted many leading domestic automotive software and hardware suppliers including chip, software and Tier 1 manufacturers. Intel attended the conference and exhibited a series of leading products and cloud-to-end solutions in the field of smart transportation. Currently, Intel is facing the digital economy, relying on its leading technology and innovation advantages, as well as cloud-to-end solutions for different application scenarios, and working with partners to promote the innovative development of smart transportation in my country and even the world.

Intel showcased its series of smart transportation solutions at the exhibition

During the conference, Intel showcased its leading technologies, including a series of cloud-to-end leading products and solutions from hardware to software, from autonomous driving development platforms to data centers. These technologies also cover multiple application scenarios such as smart cockpits, smart manufacturing and smart logistics.

Apollo Lake-based smart cockpit platform: The hardware of this platform is based on the Intel Atom® processor A3900 series. The Intel Atom® processor A3900 series is a processor designed specifically for in-vehicle experience, with powerful computing power and load integration capabilities, capable of supporting in-vehicle infotainment systems, digital instruments, rear-seat entertainment, and advanced driver assistance systems. At the software level, the platform can support the industry-leading ACRN open source hypervisor virtualization software and third-party commercial hypervisor virtualization technologies such as QNX hyperivsor. In addition, the platform also supports client operating systems such as Android, Linux, and QNX, and can support digital instruments and infotainment systems developed based on the HMI Graphics engine of Kanzi and Qt to provide a more beautiful and smooth HMI interface.

ADAS system based on Apollo Lake: L1 advanced driver assistance systems have shown improved computing power on the Apollo Lake platform. In the actual performance of lane departure warning (LDW), the processing time per frame is less than 10ms, and it has the advantages of low false alarm rate and high accuracy. In the applications of front collision warning (FCW) and front vehicle departure reminder, a monocular camera is used, based on the CNN network optimized by OpenVINO™, and the frame rate can reach 12fps.

L3 autonomous driving development platform based on Car Lake processor and FPGA: Car Lake platform is based on the new generation Atom® processor Deverton AD, which can be used with Arria 10 FPGA for auxiliary computing processing, and comes with a set of sample applications, runtimes and libraries, and middleware. Among them, Intel CPU can be applied to the open platform solution of CPU+FPGA, which has the characteristics of open architecture, high flexibility and high customer autonomy. In addition, this processor also has the advantage of low power consumption. Based on the L3 solution, the maximum power consumption of the CPU is only 30 watts, which can effectively reduce costs; at the same time, this processor can also be used in Mobileye's autonomous driving solution to help customers achieve mass production needs in a short period of time.

In addition, Intel's mature solutions enabled by IoT and AI technologies are being used in the use scenarios of intelligent automotive manufacturing and have brought considerable returns on investment. Although automotive manufacturing is already in a leading position in advanced manufacturing, its marginal room for improvement in intelligent manufacturing is quite limited. Intel and its ecosystem partners combine advanced technology with innovative development capabilities to use the Intel Industrial Edge Insight Platform to transmit machine vision data, using edge-side AI algorithms and powerful real-time computing power to achieve seamless integration of innovative edge technologies based on deep learning and MES systems running in computing centers or cloud servers. This solution can also coordinate the work between machines to generate insights and real-time decisions, bringing new marginal value driven by innovation throughout the entire life cycle of advanced manufacturing, including supply chain, manufacturing, quality inspection and logistics stages.

In terms of artificial intelligence and heterogeneous edge computing, Intel can provide innovative visual solutions OpenVINO™ toolkit. Based on the Intel® OpenVINO™ toolkit, artificial intelligence algorithms for factory production workshop monitoring and quality defect inspection can be fully accelerated and optimized. Thanks to OpenVINO™ and Intel's wide range of hardware platforms, the user's development cycle can be greatly shortened, allowing users to deploy high-accuracy edge measurement artificial intelligence industrial solutions faster and better. The intelligent tire detection and automobile fault diagnosis solutions exhibited by Intel this time use the OpenVINO™ toolkit.

At the conference, Intel also exhibited leading technologies and solutions in the fields of automotive manufacturing and smart transportation, such as the Intel Rack Scale Design data center solution and data center software management solution based on Intel's leading technologies.

Xu Weijie, general manager of Intel's Automotive Product Sales Division in China, said: "Intel has cloud-to-end solutions for different application scenarios, which can provide innovative technical support for the development of partners in various fields such as smart transportation. We look forward to Intel deepening cooperation with ecological partners in the future, combining leading technology with innovative capabilities, and jointly promoting innovation and application in the field of smart transportation in my country and even the world.