Tel
0086-516-83913580
E-mail
[email protected]

Embracing Artificial Intelligence, Dreaming of Exciting Travel, SAIC’s Driverless Taxis Will “Take the streets” Within the year

Pic 1

At the 2021 World Artificial Intelligence Conference “Artificial Intelligence Enterprise Forum” held on July 10th, SAIC Vice President and Chief Engineer Zu Sijie delivered a special speech, sharing SAIC’s exploration and practice in artificial intelligence technology to the Chinese and foreign guests.

 

Technological changes, the automobile industry is on the “new track” of smart electric

 

In recent years, with the rapid development of artificial intelligence, big data, cloud computing and other technologies, the global automotive industry is undergoing disruptive changes. The automotive industry has entered the era of smart electric vehicles from the era of horse-drawn vehicles and fuel vehicles.

 

In terms of automotive products, automobiles have evolved from a “hardware-based” industrialized product to a data-driven, self-learning, self-evolving, and self-growing “soft and hard” intelligent terminal.

 

In terms of manufacturing, traditional manufacturing factories can no longer support the requirements for building smart cars, and a new “data factory” is gradually being formed, enabling the self-evolutionary iteration of smart cars.

 

In terms of professional talents, the automotive talent structure based on “hardware” is also evolving into a talent structure that combines both “software” and “hardware”. Artificial intelligence professionals have become an important force for participation in the automotive industry.

 

Zu Sijie said, “Artificial intelligence technology has fully penetrated into all aspects of SAIC’s smart car industry chain, and has continuously empowered SAIC to realize its vision and mission of “leading green technology and driving dreams”.

 

User relationship, the “new play” from ToB to ToC

 

In terms of user relations, artificial intelligence is helping SAIC’s business model to transform from the past ToB to ToC. With the emergence of young consumer groups born in the post-85s/90s and even post-95s, the traditional marketing models and reach mechanisms of car companies are facing failures, the market becomes more and more segmented, and car companies must more accurately meet the needs of different users. Therefore, auto companies must have a new understanding of users and adopt new ways of playing.

 

Through the CSOP User Data Rights and Interests Plan, Zhiji Auto realizes feedback on user data contributions, allowing users to share the future benefits of the enterprise. SAIC’s passenger car marketing digital business uses data and artificial intelligence algorithms as the core, accurately grasps different user needs, continuously subdivides user needs, and evolves more personalized “feature images” from “standard images” , To make product development, marketing decision-making, and information dissemination more “reasonable” and “targeted”. Through digital marketing, it has successfully helped MG brand sales increase by 7% in 2020. In addition, SAIC has also empowered the R brand online customer service system through a knowledge map based on artificial intelligence, which has greatly improved operational efficiency.

 

Product research and development will “simplify the complex” and “one vehicle with a thousand faces”

 

In product development, artificial intelligence is empowering the user experience of “one vehicle with a thousand faces” and continuously optimizing product development efficiency. SAIC Lingchun took the lead in introducing service-oriented design concepts into the development of smart car software platforms. On April 9th, SAIC held the world’s first automotive SOA platform developer conference, which was held in Baidu, Alibaba, Tencent, JD.com, Huawei, OPPO, SenseTime Witnessed by leading technology companies such as, Momenta, Horizon, iFLYTEK, Neusoft and other leading technology companies, they released SAIC’s zero beam SOA developer platform to “simplify the development of smart cars” and help form a “one car with a thousand faces” user Experience.

 

By decoupling the hardware and software of the smart car, SAIC Automotive has abstracted the hardware into a public atomic service that can be called. Like Lego, it can realize the personalized and free combination of software service functions. At present, more than 1,900 atomic services are online and open. Available for call. At the same time, through opening up various functional domains, combining artificial intelligence, big data and other technologies, it forms a closed loop of experience from data definition, data collection, data processing, data labeling, model training, simulation, test verification, OTA upgrade, and continuous data integration. Training to achieve “let your car know you better”.

 

SAIC Lingshu also provides an exclusive development environment and tools to convert the cold code into a graphical editing tool. With a simple mouse drag and drop, the “engineering novices” can also customize their own personalized applications, allowing suppliers, Professional developers and users are able to participate in the application development of smart cars, not only to realize the personalized subscription service of “thousands of people, but also to realize the “civilization” development and application of cutting-edge technologies such as artificial intelligence, big data, and software design.

 Pic 2

Take the Zhiji L7 to be delivered at the end of the year as an example. Based on the SOA software architecture, it can generate personalized function combinations. By calling the perception data of more than 240 sensors in the entire vehicle, iterative optimization of the functional experience is continuously realized. From this, the Zhiji L7 will Truly become a unique travel partner.

 

At present, the development cycle of a complete vehicle is as long as 2-3 years, which cannot meet the market’s demand for rapid iteration of smart cars. Through artificial intelligence technology, it can help shorten the vehicle development cycle and improve development efficiency. For example, the development of chassis systems has accumulated nearly a hundred years of knowledge accumulation in the automotive industry. The large stock of knowledge, high density, and wide fields have led to certain challenges in the inheritance and reuse of knowledge. SAIC combines knowledge maps with intelligent algorithms and introduces them into the design of chassis parts, supports precise search, and greatly improves the development efficiency of engineers. At present, this system has been integrated into the daily work of chassis engineers to help engineers quickly grasp knowledge points such as part functions and failure modes. It also connects knowledge in different fields such as braking and suspension to support engineers to make better part design plans.

 

Smart transportation, 40-60 unmanned taxis will “take the streets” within the year

 

In smart transportation, artificial intelligence is being integrated into the core links of “digital transportation” and “smart port”. SAIC gives full play to its practical experience and industrial chain advantages in innovative technologies such as artificial intelligence and autonomous driving, and actively participates in Shanghai’s urban digital transformation.

 

In terms of digital transportation, SAIC has created the Robotaxi project of L4 autonomous driving for passenger car scenarios. Combined with the project, it will promote the commercial application of technologies such as autonomous driving and vehicle-road collaboration, and continue to explore the realization path of digital transportation. Zu Sijie said, “We plan to put 40-60 sets of L4 Robotaxi products into operation in Shanghai, Suzhou and other places by the end of this year.” With the help of the Robotaxi project, SAIC will further advance the research of the “vision + lidar” intelligent driving route, realize the implementation of autonomous wire-controlled chassis products, and will also use artificial intelligence technology to realize the continuous upgrade and iteration of the “data-driven” self-driving system, and solve the problem of automation The “long-tail problem” of driving, and plans to achieve mass production of Robotaxi in 2025.

 

In terms of smart port construction, SAIC, in conjunction with SIPG, China Mobile, Huawei and other partners, based on the common scenes in the port and the unique scenes of the Donghai Bridge, and fully applied technologies such as autonomous driving, artificial intelligence, 5G, and high-precision electronic maps to create two major The self-driving vehicle product platform with independent intellectual property rights, that is, the L4 smart heavy truck and the intelligent AIV transfer vehicle in the port, has built an intelligent transfer scheduling solution for the smart port. Using big data and artificial intelligence, SAIC continues to optimize the machine vision and lidar perception capabilities of autonomous driving vehicles, and continuously improves the high-precision positioning level of autonomous vehicles, as well as the reliability and “personification” of vehicles; at the same time, it also By opening up the port business dispatching and management system and the self-driving fleet management system, the intelligent transshipment of containers is realized. At present, the takeover rate of SAIC’s smart heavy trucks has exceeded 10,000 kilometers, and the positioning accuracy has reached 3cm. This year’s takeover target will reach 20,000 kilometers. It is expected that the quasi-commercial operation of 40,000 standard containers will be realized throughout the year.

 

Intelligent manufacturing enables “double improvement” of economic efficiency and labor productivity

 

In intelligent manufacturing, artificial intelligence is promoting the double improvement of the “economic benefits” and “labor productivity” of enterprises. The “Spruce System”, a logistics supply chain decision-making optimization product based on deep reinforcement learning developed by SAIC Artificial Intelligence Laboratory, can provide functions such as demand forecasting, route planning, matching of people and vehicles (vehicles and goods), and global optimization scheduling to achieve economic benefits for users And labor productivity. At present, the system can reduce the cost and increase efficiency of the automotive logistics supply chain by more than 10%, and increase the processing speed of the supply chain business by more than 20 times. It has been widely used in the supply chain management optimization service inside and outside SAIC.

 

In addition, SAIC Anji Logistics has developed an integrated logistics solution for the LOC intelligent warehousing project of SAIC General Motors Longqiao Road, and realized the first domestic intelligent warehousing application for the entire supply chain of auto parts LOC. “The concept is applied to the auto parts logistics industry, combined with the intelligent brain “iValon” independently developed by Anji Intelligent, to realize the linkage scheduling of multiple types of automated equipment.

 

Smart travel, providing safer and more convenient travel services

 

In terms of smart travel, artificial intelligence is helping SAIC provide users with safer and more convenient travel services. From the beginning of its establishment in 2018, Xiangdao Travel has started to build an artificial intelligence team and self-developed “Shanhai” artificial intelligence hub. Related applications have achieved vertical pricing for special vehicles, enterprise-level vehicles, and time-sharing leasing businesses. , Matchmaking, order dispatch, safety, and experience bidirectional coverage of the entire scene. So far, Xiangdao Travel has released 623 algorithm models, and the transaction amount has increased by 12%. The smart car camera has led and established a model in the online car-hailing industry. At present, Xiangdao Travel is currently the only travel platform in China that uses in-vehicle image AI blessing for risk control to ensure the safety of both the driver and the passenger.

  Pic 3

On the “new track” of smart electric vehicles, SAIC will use artificial intelligence to empower companies to transform into a “user-oriented high-tech company” and make every effort to tackle the technological commanding heights of the new round of development of the automotive industry. At the same time, SAIC will also uphold the values ​​of “user-oriented, partner advancement, innovation and far-reaching”, give full play to its advantages in market scale, application scenarios, etc., and adopt a more open attitude to build more cooperation with more domestic and foreign partners. The close collaboration relationship accelerates the breakthrough of global problems in unmanned driving, network security, data security, etc., and jointly promotes the continuous improvement of the global artificial intelligence industrialization application level, and meets the more exciting travel needs of global users in the era of smart cars.

 

Appendix: Introduction to SAIC Exhibits at the 2021 World Artificial Intelligence Conference

 

The luxury pure electric smart car Zhiji L7 will create a full-scenario and most continuous Door to Door Pilot intelligent driving experience for users. In a complex urban traffic environment, users can automatically complete parking out of the parking lot according to the preset navigation plan, navigate through the city, navigate at high speed, and reach the destination. After leaving the car, the vehicle automatically parks in the parking space and enjoys the entire intelligent assisted driving.

 

The medium and large luxury smart pure electric SUV Zhiji LS7 has a super long wheelbase and super wide body. Its embracing yacht cockpit design breaks the traditional functional cockpit layout, restructures the space, and diversified immersive experience will subvert the user’s interior Imagination of space.

 

R Auto’s “Smart New Species” ES33, equipped with R Auto’s world’s first high-end intelligent driving solution PP-CEM™, to build a “six-fold fusion of laser radar, 4D imaging radar, 5G V2X, high-precision maps, vision cameras, and millimeter wave radars. The “style” perception system has all-weather, beyond visual range, and multi-dimensional perception capabilities, which will raise the technical level of intelligent driving to a whole new level.

 

MARVEL R, the “5G Smart Electric SUV”, is the world’s first 5G smart electric vehicle that can be used on the road. It has realized the “L2+” intelligent driving functions such as intelligent deceleration in corners, intelligent speed guidance, parking start guidance, and intersection conflict avoidance. It also has black technologies such as MR driving remote sensing visual driving assistance system and intelligent calling, bringing more intelligence to users. A safer travel experience.


Post time: Jul-12-2021