Autonomous driving and smart travel have been recognized as one of the future directions of the auto travel industry. With the increasing popularity of intelligent and connected vehicles and the rapid development of autonomous driving technology, many industry participants such as OEMs, software technology companies, and chip companies generally believe that “software-defined vehicles” are no longer an empty concept, but a fact.
At the recent “Chief Intelligent Mobility Officer Conference” held by Auto Byte, a platform focused on the smart travel industry created by the well-known AI information platform Machine Heart, executives from a number of companies once again expressed the meaning of software-defined vehicles, and put forward the saying that “software defines the car, hardware defines the software ceiling”. The main point of this is to clarify the relationship between technology and hardware coexistence rather than substitution and to point out that the “computing power” jointly determined by the two is the major factor in transitioning from low-level autonomous driving to high-level autonomous driving at this stage. particularly critical factor.
Just five or six years ago, the entire auto industry didn’t have a car that could be called self-driving. Some models are only equipped with some functions for driving assistance, such as automatic cruise, lane keeping, lane departure warning, and other functions. A few years later, the so-called “self-driving car” products launched by OEMs have been overwhelming, L2 has been the standard, and L3 and L4 cars have also appeared. At the same time, a large number of new car manufacturers, autonomous driving technology companies, and chip companies have emerged, and they have become the main force in promoting the development of autonomous driving. The guest representatives participating in this “Smart Driving Conference” are mainly from major leading companies in the industry chain.
As one of the representatives of OEMs, Jidu Automobile CEO Xia Yiping said at the conference that the computing power of automotive-grade chips has been lower than that of consumer-grade chips for a long time, which makes AI technology unable to play an advantage in automobiles, but in the era of smart car 3.0, it can endow the car with enough computing power to gradually transform it from a means of transportation into an AI-driven intelligent mobile space, which in turn brings technological innovation, efficiency improvement, and experience subversion. He judged that 2023 will be the start of intelligent automobile competition, and the era of real automobile 3.0 has arrived.
As one of the representatives of autonomous driving technology companies, Gu Weihao, co-founder and CEO of HAOMO.AI, said that in the field of autonomous driving, test mileage and test scenarios are important factors in determining capability and safety of autonomous driving systems. Data intelligence is the most fundamental driving force for the evolution of autonomous driving AI. By further learning, mining, processing, and training the feedback data, more algorithms and service modes OTA can be delivered to the car end, which can bring better system performance to users. In this process, cost and speed are the two most critical aspects, and they are also the ideological stamp of data intelligence.
As one of the representatives from the autonomous driving chip company, Wang Ping, executive president of Cambricon, bluntly said that the large-scale implementation of intelligent driving faces multiple challenges on the chip: the computing power of a single chip is not enough, so two or even more chips are needed to achieve it. However, this also leads to a significant increase in system complexity and power consumption, and increases system costs, making it difficult to popularize fuel vehicles or economical electric vehicles below 100,000 yuan.
He judged that there are two future trends for autonomous driving chips, one is general openness, and the other is large computing power. In the era of L1 and L2 autonomous driving, because the amount of data is relatively small, many car companies can accept closed integrated solutions with strong coupling between chips and algorithms. Only large computing power chips can meet the demand.
Based on the opinions of representatives from various enterprises, large computing power has become a major test restricting the development of current autonomous driving, and it is also a prerequisite for the large-scale implementation of autonomous driving systems and further commercialization.
The test of big computing power
Computing power is usually used to refer to the performance of a chip. A simple understanding means that the greater the computing power, the better the performance. With the introduction of laser sensors, the autonomous driving computing platform has exceeded 1000TOPS, and computing power has become one of the main selling points of more and more automobile manufacturers. However, high computing power means that in order to achieve a breakthrough in the synchronization of hardware and software in technology, it is necessary to retain a certain amount of redundancy and to achieve a commercial balance between technology and business.
According to Yang Yuxin, Chief Marketing Officer of Black Sesame Intelligence, the development of autonomous driving has come to the “second half of the first half”, and computing power has also become an important indicator for judging the degree of car intelligence. Car companies hope that by highlighting the value of computing power, end users will have more awareness of the autonomous driving capabilities of car companies. The current computing power can theoretically meet the needs of L2+ and L3 autonomous driving systems, and the next focus is to make the application and experience better.
He also added that “computing power stacking” is a necessary redundancy for subsequent technology upgrades. In terms of business logic and technology evolution, chip companies also need to help customers with lower costs and higher system concentration, lower power consumption, and better autonomous driving functions. This is what chip companies have been working on, and it is also a point that promotes the evolution of everyone’s technology and product routes.
Li Bo, vice president of Lotus Technology and head of the intelligent driving business line, believes that hardware defines the software ceiling, and reserves enough computing power and sensors to reserve redundancy for the performance requirements of future autonomous driving systems. Otherwise, it’s like the current application logic can work on the old phone, but it can’t really run.
In China, chip research and development is not only done by chip manufacturers. Under the current situation of chip shortage, it has also become a major trend for OEMs to develop self-driving chips. Tesla, Xiaopeng, Geely, and other car companies are listed here. Wang Kai, director, and CEO of Geely’s core technology said that on the one hand, the shortage of chips has made OEMs pay more attention to supply chain diversity and supply security. On the other hand, high computing power chips have become the core competitiveness of car companies, it is increasingly difficult for supplier chips to meet the iteration speed, cost, and performance requirements of OEMs.
However, Wang Kai also said that there are many challenges for car companies to develop their own chips: the threshold for autonomous driving chips is high, and once they take a detour, they will face huge financial losses and will also cause planning inconsistencies. At the same time, car-grade chips are different from consumer-grade chips and have higher requirements for performance, power consumption, and reliability. They also need to complete car-grade certification. The cycle is longer and the investment is larger. It needs to pass the application in a variety of vehicles. Popularization to recover the upfront cost, so it is necessary to launch a more inclusive and competitive product system to meet the needs of different car manufacturers.
The current chip shortage has become a huge pain point in the automotive industry. Facing this problem, many corporate guests expressed their views. All in all, the chip shortage will continue for a period of time. Although production capacity has recovered from the epidemic, the demand that was suppressed last year has not been met, and the real solution may have to wait until next year. In addition, chip manufacturers are in a wait-and-see state. At present, the cost of chip expansion is relatively high, and chip manufacturers are blindly expanding production capacity without guaranteeing that there will be the same demand in the next few years.
Commercialization in 2025?
Whether it is intelligent driving or automatic driving, what it wants to achieve is automatic and intelligent travel in various scenarios, so that people, cars, and transportation systems can build a fully autonomous driving society through the link of AI technology, and use technology to change human behaviors.
In the industry, the current commercialization of autonomous driving is divided into two schools of thought. One believes that this is a distant dream and may never be realized; the other believes that the commercialization of autonomous driving is imminent and may be achieved in 2025. The delegates at this “Chief Intelligent Mobility Officer Conference” believe that the commercialization of autonomous driving will take some time, but some people are optimistic that the project will be implemented in 2025.
Among the popular applications of autonomous driving, Robotaxi (autonomous taxi) is the closest track to commercialization. Xiao Jianxiong, founder and CEO of AutoX (Antu), a representative company in this field, said that AutoX has always focused on the L4-level unmanned RoboTaxi without safety officers, and he believes that this is the only way to truly commercialize this route. Only when it achieves the same practicability as the existing online car-hailing and completely removes the safety officer, autonomous driving with unlimited destinations and regions, can it be truly commercialized.
Among them, the coverage area is the most important point for Xiao Jianxiong. Xiao Jianxiong said that the commercial operation of RoboTaxi must have a large enough service area. If it can only run on a few main roads, it is more of a pure technology demonstration without real commercial value. Furthermore, large-scale mass production is also necessary for the commercialization of RoboTaxi, which determines efficiency, consistency, and reliability.
Autonomous driving commercialization projects are mainly divided into two categories: B-end and C-end. Zhou Xin, co-founder and chief product officer of UISEE, and Hao Jianan, co-founder and chief architect of TuSimple, both believe that efficiency and cost are the prerequisites for the commercialization of autonomous driving on the B-side: either be more efficient than humans, or fully autonomous driving. However, in order to realize the final business logic, not only a very high level of security and reliability is required, but also the gradual improvement of regulations.
As an enterprise targeting both B-end and C-end users, Dong Jian, co-founder of Hongjing Zhijia and VP of software algorithm, said that the current landing speed is faster than expected, and more mass-produced models will appear in one or two years. However, subject to legal and regulatory issues, most car companies will launch models with an L3 autonomous driving experience but developed according to the L2+ regulatory system.
The legal and regulatory issues of autonomous driving have been mentioned by many guests, which is directly related to whether the commercialization project can be approved or not, and also involves the issue of the autonomous driving responsibility system.
Dong Jian said that in the past one or two years, the mass-produced models launched by car companies are called L3 and L4, but technically they are still the system of L2. This is because there are no specific L3 regulations in China, so even if the function has been I have achieved L3 and L4 experience, if there is an accident in the responsibility system, the driver is still responsible. She believes that autonomous driving in the true sense of L3 and L4 means that the responsibility for an accident lies with the car, not the driver. Europe already has L3 regulations in the true sense of ALKS autonomous driving, so the introduction of domestic regulations is just around the corner.
In terms of automatic driving promotion methods, representatives of many car companies agree that it is a relatively safe way to gradually realize automatic driving coverage from small scenes to large scenes. It is not in line with actual conditions to achieve automatic driving above L4 level in one step. Closed scenarios such as parks, ports, and mines are the main scenarios for current driverless applications, followed by scenarios such as trunk logistics, urban public transportation, and ultimately autonomous driving for personal mobility.
Dai Zhen, Vice President of Heduo Technology, gave a more specific time point for the C-side landing of autonomous driving – it is expected that 2025 will be a critical time node when the mass production of autonomous driving technology, consumer acceptance, infrastructure, and relevant laws and regulations will be gradually implemented.
