Electric vehicles have entered a new market stage, and intelligence has become the next driving force driving the market demand for electric vehicles. This year's 100 Electric Vehicles Meeting invited many relevant enterprises in the field of autonomous driving and intelligent vehicles in China to express their opinions and talk about the road to intelligentization of electric vehicles in mainland China.
Horizon: Supports the new model of automakers' self-developed chips
Yu Kai, founder and chief scientist of Horizon, mentioned that smart cars are subversive innovations comparable to the "birth of computers" and will become the largest parent ecology of human technological development. The chip and operating system are the digital base of smart cars and the best breaking point for the industry.
There have been many times in history that the chip architecture and the operating system have been closely aligned, such as the alliance of Windows and Intel, the alliance of ARM and Android, and the coupling of NVIDIA GPU architecture and operating system CUDA. "Computing energy efficiency can only be guaranteed if the software and hardware are highly coordinated."
While providing SoC-level high-performance automotive smart chips, Horizon will be further opened, opening up BPU IP authorization to automakers, providing software kits, chip reference designs and technical support, helping some car companies develop self-developed chips, enhance differentiated competitiveness, and accelerate the speed of innovation and research and development. In this open mode, vehicle development will achieve a high degree of synergy from chips to operating systems to software and hardware systems for autonomous driving, greatly improving the iteration speed.
In the era of intelligent cars, under the surge in consumer demand, higher requirements have been put forward for the speed of product iteration, and it is crucial to create a more open innovation ecosystem. Just as Windows and Intel have jointly accelerated the mature development of the PC era, ARM and Android have jointly created a prosperous ecology in the era of smart phones, in the era of smart cars, Horizon expects to create an open source and open innovation ecology, with smart chips and operating systems as the digital base of smart cars, and industry partners to travel with each other, open and win-win, and work together to build an open and collaborative era of automotive intelligent innovation.
Black sesame: Domestic large-capacity car rules chips have entered the year of mass production
Shan Jizhang, founder and CEO of Black Sesame Intelligence, pointed out that 2022 will be the mass production year of domestic large-scale car specification chips, and Black Sesame Intelligence is fully promoting the mass production of its chips to empower intelligent cars and autonomous driving in China and even the world.
From a technical point of view, from the real breakthrough of L2 to L3 will be a relatively long process, which involves software, hardware, data and other technologies with the continuous upgrading of the automatic driving system. The realization of intelligent cars and automatic driving in the future needs to break through the bottleneck of computing power, and the use of large computing power chips is the key to achieving breakthroughs.
At present, there are still several key technologies to be broken through from the mass production of large-scale car specification chips. For example, in terms of chip technology, advanced packaging technology and independent IP technology are required; the system architecture of high computing power chips needs to be broken, and it has a high security tool chain, a high-security operating system and information security; in terms of vehicle regulation certification, a series of core technologies such as functional safety process, functional safety product certification, vehicle regulation reliability certification, and ASPICE software certification need to be broken one by one.
It is not difficult to see that the realization of the mass production of large-scale car specification chips is a difficult process, which requires the necessary stages and investment. Taking the Black Sesame Intelligent Huashan series of autonomous driving chips as an example, it has experienced more than 3 years from the completion of product definition, tape-out, packaging and testing, vehicle regulation certification and algorithm tool chain ready, functional safety certification to end-customer verification, mass production on the car. Huashan No. 2 A1000 chip is in mass production, and it is also the largest and most powerful automatic driving computing chip in China.
SenseTime: Smart cars have the synergy advantage of multi-industry integration business
Sun Dapeng, vice president of SenseTime, said that at present, SenseTime's business can basically be summarized as "one platform, four pillars". "One platform" is the artificial intelligence infrastructure, equipped with a self-developed training framework. The "four pillars" are the four business sectors formed on this basis, namely smart business, smart city, smart life and smart car. Among them, smart cars are the most obvious plate with multi-industry integration business synergy advantages, combined with smart business to form automatic connection products, which can effectively reach commercial hubs and promote business model transformation; cross with smart cities to form smart transportation products, so that the vehicle-road collaboration system can effectively integrate various business scenarios and applications of smart transportation, bringing more choices for travel; combined with smart life, the entertainment application of the mobile phone can be effectively applied to the entire cabin, so that the passenger experience is more beautiful.
The current logic of SenseTime's vehicle-to-road collaboration solution is "smart car + smart road + collaborative cloud", which allows them to be fully integrated and accelerate the continuous evolution of autonomous driving technology.
The car platform mainly includes two categories: one is the traditional vehicles that belong to the network transformation, plus the L4 level automatic driving minibus. The road platform mainly collects the roadside perception ability based on various sensors, plus the more intelligent signal light control system.
The cloud platform mainly includes three points: the cloud control platform mainly provides vehicle-side access, roadside device access and cloud control; the business platform mainly provides visual algorithms, traffic engineering and business applications; and the training platform relies on AI training devices plus training engines to make data more secure and flexible.
The three products of vehicle, road and cloud are integrated together and are now widely used in the management of demonstration areas, closed parks, highways and urban traffic. At present, SenseTime's collaborative product system has been connected to more than 1.2 million networked operation vehicles, and has more than 20 intelligent transportation detection capabilities.
One-way technology: "double wisdom" stimulates intelligent driving, vehicle-road collaboration and other markets will grow tremendously
The goal of the coordinated development of "dual intelligence" is to build a safer, more efficient and more intelligent transportation era. As an important part of smart transportation and smart city, the development of intelligent and networked new energy vehicles and the construction of vehicle-road collaborative technology will carry the important mission of empowering the urban transportation revolution and promoting the accelerated landing of smart cities.
Shi Tuo, founder and CEO of Yijing Technology, believes that cost is the main factor restricting the development of autonomous driving, including vehicle-road collaboration and the entire intelligent networking industry, and lidar as a core sensor is the main part of the cost component. In the process of large-scale application, L3 automatic driving may require a vehicle to be equipped with 1-3 lidar, L4 automatic driving needs to be equipped with 3-5, or even 7-9 lidar. A large number of lidar will cause the cost of vehicle sensors to remain high, which requires us to develop high-performance, high-reliability lidar products at the same time, but also to promote the entire industry chain upstream and downstream to work together to jointly achieve lidar cost reduction. On the one hand, performance improvement is achieved through technological innovation, on the other hand, through chip integration, independent chip design to improve the integration of the entire product, and automated manufacturing to achieve the final cost reduction.
In the background of "double intelligence", under the guidance of the direction of intelligence and networking, the markets such as intelligent driving and vehicle-road coordination will usher in tremendous growth. With the rapid development of the market, a number of excellent enterprises have been born, and the industry has been promoted to develop rapidly through technological innovation. In the past few years, the lidar industry has developed rapidly, through technological innovation and chip integration, lidar has rapidly matured, and the cost has dropped rapidly, which has promoted the development of intelligent driving technology; but at the same time, it has also been seen that the lack of core caused by the epidemic has become prominent, and the ecological chain construction of the entire industry needs to be strengthened. We need to increase and adhere to investment in technology research and development, and promote the upstream and downstream of the industrial chain to a higher level and deeper collaboration. Yijing Technology expects to work together with all colleagues to jointly help the rapid development of intelligent driving and truly realize the scale of 'double intelligence'.
Jidu: The smart car 3.0 era pays more attention to software security
Xia Yiping, CEO of Jidu, believes that the era of smart car 3.0 is an era that pays more attention to software security. In the era of fuel vehicles and electric vehicles, the innovation of the automobile industry is driven by energy. Now, electric vehicles are ushering in a comprehensive digital upgrade, AI intelligence has become a new trend, "AI has brought technological innovation, efficiency improvement, experience subversion." We believe that 2023 will be the first year of intelligent competition in automobiles, and the era of intelligent car 3.0 has arrived. ”
In the era of intelligent car 3.0, "free movement, natural communication, and self-growth" is the basic concept that automotive products should have, and behind it requires a set of complex software and algorithms, as well as hardware to meet the corresponding computing power needs. To this end, Jidu has developed its own electrical and electronic architecture and domain controllers, from the perspective of combining software and hardware, to fully ensure the security of the overall architecture. Software research and development, software capability verification, and software operation security and stability are the most concerned issues in Jidu.
2022 will be a watershed year for the commercialization of AI autonomous driving
Hou Jun, COO of Zhixing, believes that the development of automatic driving needs to comply with the three laws of automatic driving, that is, first low speed and then high speed, first load things and then carry people, first commercial and then civilian. This gradual route is more in line with the law of autonomous driving growth, and the industry is now exploring from the feasible and reliable stage to the commercializable stage. At present, the high-level assisted driving of passenger cars for passenger cars has now entered the commercial stage, and the automatic driving of low-speed unmanned logistics vehicles is being explored from feasible and reliable.
The overall trend is that from assisted driving to fully autonomous driving, the market penetration rate will be higher and higher. Taking Great Wall Motors as an example, in 2022, Zhixing will undertake the task of high-level assisted driving development of Great Wall Motors' 34 models to be listed, accounting for nearly 80% of Great Wall Motors' models to be listed in the whole year, of which 30% are standard and the rest are equipped with high equipment.
With the increased penetration rate of China's autonomous driving market, the future market of autonomous driving will also blossom. With the blessing of AI technology, the pace of car driving towards intelligence has been unstoppable, and high-level intelligent driving has been on the eve of large-scale mass production. Hou Jun said that 2022 will be a watershed year for the commercialization of AI autonomous driving, a key year for the entire automotive industry to move towards automation, intelligence and scale, and the industry will explore more in high-speed scenarios, urban open roads, low-speed complex scenarios, etc. The way to commercialize autonomous driving is to choose a simple scene and gradually change from a simple scene to a complex scene; the other is to choose a complex scene to reduce the speed of the car, from low speed to high speed.