Deep integration, cross-border cooperation, status quo of automotive artificial intelligence

: The human production tool is shifting from a power tool based on matter and energy to an intelligence tool based on data, information, value, and intelligence. The artificial intelligence bonus is replacing the demographic dividend. The smart car industry, represented by autonomous driving, integrates almost all mid-to-high-end technologies in the fields of industry and information. It has a long chain, a strong connection, and a wide range of jobs. It involves a large number of economic sectors and a large consumer demand. It will surely be the future of the national economy. The important pillar industry.

The American Society of Automotive Engineers (SAE) divides autopilot into five phases according to the degree of automation: Level 1 ADAS driving assistance, Level 2 PA partially automated driving, Level 3 CA conditional autopilot, Level 4 HA highly automated driving, Level 5 FA Fully automatic driving. The traditional car enterprises generally gradually achieve automatic driving from the first level. Currently, some car manufacturers have already achieved level 2 production models and are now progressing to Level 3; while Internet companies directly from artificial intelligence to fully automatic driving, such as Google driverless cars Developed according to the fully automated driving level.

The Chinese government proposed the concept of “Internet+” in 2015. The Ministry of Industry and Information Technology subsequently announced that it will vigorously develop smart car manufacturing and promote the transformation and upgrading of the traditional automotive industry in the direction of intelligence and networking. In the field of domestic intelligent network-linked automobiles, many manufacturers compete, automobiles and artificial intelligence are deeply integrated, and the traditional car companies and IT companies cross-border cooperation.

First, SAIC Shake hands Alibaba step by step for a comprehensive layout of the "elite-style" smart network car strategy

At the China Automotive Forum 2016, Cheng Jinglei, chief engineer of SAIC, introduced SAIC's 13th Five-Year Plan, including the smart driving plan with a deadline of 2015 to 2025, to achieve a structured and partially unstructured road within five years. The autopilot function achieves full autonomy in 10 years.

At the same time, the forum presented the IGS intelligent driving concept car, "eye" (perceptive system, composed of various radars, cameras) + "limbs" (executive agencies, including accelerator, brake, gear, electric control chassis and other systems ) + "Brain" (ie, software control, including software engineering, decision analysis, logic algorithm, etc.) Under the triple combination, complete the driver-assisted lateral and longitudinal control, the machine to complete the perception and judgment of the driving environment, can be regarded as SAIC Intelligence Intelligent Level 3 products in driving planning.

On July 6, 2016, SAIC Group and Alibaba Group held a conference on the launch of new Internet-enabled cars in Yunqi Township, Hangzhou. The first production Internet car Roewe RX5 was equipped with the Yun OS smart operating system. Yun OS is a smart operating system under the Alibaba Group and incorporates Alibaba's technological achievements in various fields such as cloud data storage, cloud computing services, and smart device operating systems. Yun OS for Car will open up the car lifecycle demand and Internet life cycle, allowing users to experience an Internet-based, more convenient mobile intelligent ecosystem. The Roewe RX5 is the result of cooperation between SAIC and Alibaba after signing a strategic cooperation agreement in 2014. It is positioned as the beginning of the real industrialization and commercialization of Internet vehicles.

II. BAIC and Baidu announced the strategic transformation and upgrading of smart cars

On January 7, 2017, BAIC and Baidu Smart Automotive jointly announced at the CES 2017 (International Consumer Electronics Show) that both parties had reached a comprehensive strategic cooperation in the field of smart vehicles and will strive to jointly promote the intelligent upgrade of the automotive industry. The cooperation between the two parties includes: in-depth cooperation on high-definition maps; application of smart car solutions based on Baidu CarLife, MyCar and CoDrive; in-depth cooperation based on L3 smart driving technology; establishment of joint laboratories; brand and product marketing; cloud platform and big data Technology and so on. At the same time, the Beiqi intelligent car development strategy named "Nova-PLS" was also made public for the first time at the conference. The goal was to enhance BAIC's three areas of smart-pilot Nova-Pilot, Nova-link, and Nova-Space. Competitive Advantage.

In Beiqi's smart car strategy, Beiqi and Baidu are mutually preferred partners. Baidu’s existing Camera+Radar+HDMap high-precision map sensor solution can automate smart cars in certain driving environments such as highways, parking lots, and closed parks. Driving function. At the same time, BAIC's smart car strategy is open, including Bosch, Huawei and other technology companies. In addition to the full implementation of cross-border strategic cooperation with Baidu, Beijing Automotive will continue to increase investment in the future plans to acquire or share advanced technology start-up companies to create intelligent core technology DNA.

Third, the deep integration of automotive and artificial intelligence development proposals

As we all know, the intelligent upgrade of the automotive industry has become an inevitable trend in the future development of the automotive industry, vehicle companies are committed to transforming from traditional car manufacturers to upgrade solutions providers. The development model of the future automobile industry should be the deep integration, cross-border cooperation, and collaborative innovation of traditional auto companies and Internet companies. Both play their own advantages and minimize the development cycle of autopilot. At the same time, governments, enterprises, research institutes, and industries Institutions should work together to jointly promote the development of smart driving and bring new experiences for future human travel. Specific should start with the following aspects:

1. The lack of resources for the current basic support and frontier technologies restricts the progress and sustainability of the field of automotive artificial intelligence. It is proposed to increase the investment in resources based on the leading frontier theoretical and technological research supported by production, research, and industrial application support. , Strengthen the collaboration and cultivation of cross-disciplinary scientific research resources, increase the resources input and industrial support efforts that focus on the needs of customers, pay attention to the delivery of results, and focus on the application and research and development of basic technologies for enterprises.

2. The current domestic road traffic safety law does not provide relevant regulations for intelligent driving and driverless road qualifications. It is suggested that the government establish a joint working mechanism to promote the definition of cross-functional and cross-industry access regulations and promote the access of automated driving vehicles. The establishment of laws and regulations and their support systems include vehicle qualifications, driver qualifications, site qualification certification, and insurance rule definitions.

3. To meet the special needs of intelligent networked vehicles for R&D and testing of environmental simulations and to ensure the safety of vehicles on the road, it is recommended that multiple resources be collaborated to build a dedicated test base to promote the evaluation and standardization of vehicle test and evaluation systems, and to define and establish comprehensive smart cars. Safety assessment system.