China’s first automatic driving simulation bluebook is released! The original technology of the automotive industry has thriven!
On May 30, 2019, China’s first self-driving simulation blue book “China Autopilot Simulation Technology Research Report (2019)” was released at the 6th International Congress of Intelligent and Connected Vehicles Technology(CICV).
This blue book for automated driving simulation was initiated by 51VR, jointly issued by Suzhou Automotive Research Institute of Tsinghua University, Intelligent Network Technology R&D Center of Guangzhou Automobile Research Institute, Intelligent Automotive Research Office and Automotive Software Evaluation Center of China Automotive Technology Research Center, Jiangsu Intelligent Networking Automobile, Beijing Smart Car Industry Innovation Center, Singulato, and 51VR.
“Annual research report on autonomous vehicle simulation in China(2019)” is the first comprehensive reference book on the development status of China’s automated driving simulation test. It combines the cutting-edge research results of current academic institutions and the latest in leading companies, covering all areas of autonomous driving simulation testing, including market demand analysis, method application, construction technical solution, software status, virtual scene database, demonstration area test method, standard, challenges and development trends. It is compiled from the opinions of many industry experts in the field of automated driving simulation.
Among them, Professor Cheng Bo, Dean of Suzhou Automotive Research Institute of Tsinghua University, served as the chairman of the editorial committee of the Blue Book and checked the overall direction and core points of the Blue Book.
The editorial committee members who participated in the writing of the Blue Book include Dr. Kong Fanzhong (Vice President of Beiqi Research Institute and Dean of New Technology Research Institute), Dr. Guo Jishun (Director of Intelligent Driving Technology Department of Guangzhou Automobile Group Co., Ltd.), Mr. Wang Yu (China Automotive Technology and Research Center, Director of Intelligent Automotive Research Laboratory and Automotive Software Evaluation Center), Dr. Dai Yifan (General Manager of Jiangsu Intelligent Networking Automotive Innovation Center), Dr. Huang Yu (President of Singulato USA, President of American Research Institute), Dr. Zhang Fan (51VR’s Director of Intelligent Vehicle and Transportation Division). The strong lineup of experts is a strong guarantee for the authority and profession of the Blue Book in the industry.
For the significance of this blue book, Professor Cheng Bo said: “The domestic (China) automated driving simulation industry is still in its starting stage, but it is breaking the long-term monopoly of foreign simulation software. The release of the Blue Book is a stage for the rise of Chinese original simulation software.”
Professor Cheng Bo’s comments just confirmed the status of current China. Nowadays, China embraces a new era of the rapid rise of independent original technology.
The rise of China’s original scientific and technological power
Simulation software has always been a weakness in China’s equipment industry software. The same is true for automatic driving simulation software. Foreign automated driving simulation software started early, and it has been bound with hardware, and the development trend is rapid.
China has to be deeply involved in the competition of global automated driving simulation technology and standard setting in order to change the situation. To participate in the development of international standards, China first needs to have its own original simulation software.
The automated driving simulation software represented by 51VR, which has intellectual property rights technology and road data with Chinese characteristics, is now trying to change the monopoly of the automotive industry by foreign software.
Although China’s autonomous driving simulation software started later, it still has its unique advantages. Whether it is derived from scenes, laws and regulations, traffic environment, usage habits, local services, and quick response, foreign automated driving simulation software is hard to compete with domestics (China).
“China will be able to catch up with foreign simulation software in the future in the automatic driving simulation software industry. I am very confident about this. It is really difficult to do automatic driving simulation software in China. But it is really meaningful for the market and the whole industry innovation.” Dr. Dai Yifan, General Manager of Jiangsu Intelligent Network Automotive Innovation Center, said.
Today, China has begun to participate more and more deeply in the competition in global technology and standards development.
In May 2018, the International Organization for Standardization established the Working Group on Automated Driving Test Scenarios. China first served as the convener of the International Standard Working Group and made China’s contribution to the field of international standardization of smart vehicles.
It is crucial to seize the commanding heights of technical specifications. Who can participate more deeply in the simulation test methods and data format standardization work is the one who can pioneer in the future.
As the China-US trade war continues to escalate, the cold war of science and technology is on the verge of exploding, and the issue of original technology is getting hotter and hotter.
As a general-purpose underlying technology, computer simulation technology has become one of the foundations of the complete technology chain of intelligent manufacturing in China. The technical experience of each product development process is iterative, the accumulation of engineers’ practice and the optimization of the product development process are the continuous improvements of Chinese computer simulation software and data information system. The development of an automated driving simulation software platform is also to train talents and establish backup systems.
Someday, if the foreign giants cut off the supply of industrial software, this independent research and development system can be put into use immediately to ensure the continuous development of China’s automated driving technology and smart transportation.
Why is the automated driving simulation more and more important?
In March 2018, a collision with a pedestrian occurred in a self-driving car in Uber, which directly led to the suspension of its drive test permission.
More and more Tesla’s driver deaths are caused by the Autopilot assisted driving system. In January 2016, when a Tesla was driving at the Beijing-Hong Kong-Macau Expressway, a rear-end collision with the road sweeper in front caused the driver to die. Tesla admitted that the vehicle was in autonomous driving at the time of the accident; in June of the same year, there was also an accident in Tesla that collided with a heavy truck crossing the road in Florida, causing the driver to die on the spot.
Automated driving accidents have prompted national vehicle technology regulators to be vigilant and begin to carefully examine autopilot road tests on public roads.
In order to solve the actual road test problem, the state and various regions actively invested in the construction of a number of closed test sites, and conditionally opened a number of designated roads. However, compared to the strong test requirements, the number of autopilot test sites in China is far from satisfactory.
According to the statistics of the US Highway Administration, drivers need to travel an average of 850,000 kilometers to experience a police report accident. Nearly 150 million kilometers will experience a fatal accident. Therefore, each automated driving system requires 16 billion kilometers of driving data for system optimization. In this way, a fleet of 1,000 autopilot test vehicles is required, and it takes about 50 years to complete enough mileage testing. Due to time and expense feasibility, the industry generally recognizes that computer-based simulation technology is the most powerful tool for testing and evaluation of automated driving systems.
The automatic driving simulation system can effectively increase the frequency of high training value scenarios within a limited test mileage by providing a richer and more diverse static environment, continuous dynamic random traffic flow, combined with parameterization and edge case parameter generalization techniques.
The “China Intelligent Networking Automotive Technology Roadmap” released in 2017 pointed out that the test evaluation is one of the basic support technologies for intelligent networked vehicles. The computer simulation of the automated driving system is the key technology for the test and test of the self-driving vehicle, and it is also the basic tool for the future industry to define the development process and technical access standards for the self-driving vehicle.
Autopilot simulation test refers to the establishment of a mathematical model of realistic static environments and dynamic traffic scenes through computer simulation technology, allowing autopilot cars and algorithms to perform driving tests in virtual traffic scenarios.
Traditional computer simulation software has shortcomings in solving the following problems, such as lack of high-precision map production and matching, lack of the establishment of high-fidelity static scenes, lack of physical simulation of advanced sensory sensors, lack of driver model and intelligent traffic model simulation, lack of dynamic Scene library construction, etc.
A new generation of simulation systems with information physics features can not only more efficiently iterate and verify the autopilot algorithm but also better meet the overall testing needs of physical information systems.
At present, there are internal simulation systems such as Baidu, Tencent, and Waymo, which are using self-developed research, and external commercial software for the automotive industry, such as 51VR, AAI, CARLA, Cognata, Panosim and so on. Simulation testing has become a rigid requirement for most companies developing automated driving systems.
In the future, “autopilot software development” and “autopilot system and vehicle verification and integration” will largely rely on the common support provided by computer simulation technology.
For automotive manufacturing and technology companies, the autonomous driving simulation system runs through the whole life cycle of product development and usage: from product concept selection to product operation data collection and system upgrade.
The simulation system serves as a toolbox for engineering technology developers to achieve product security and stability, and as a data repository for business executives, serves to accumulate valuable intangible assets such as design processes, engineering experience, and digital models.
For national regulatory and testing organizations, automated driving system simulation can also be used as an analysis and testing tool to evaluate the safety and intelligence level of automotive functions.
It is estimated that the total size of the international market for simulation software and testing in the next five years will be around $10 billion. With the commercialization, the industry will have more demand for simulation virtual testing. It is expected that the simulation technology for the automatic driving system will also be applied to the development process of automotive products like the mechanical and electronic systems.
In order to give people a more comprehensive understanding of driverless simulation testing technology, the Blue Book covers a wide range of topics, from methods to databases to standards. Whether it is for automated driving development companies or testing organizations, especially newcomers in the industry, this blue book can make them quickly understand the latest practice of automatic driving simulation testing, as the most timely and detailed technical reference in the work.
International Congress of Intelligent and Connected Vehicles Technology (CICV 2019) is the top event in the autonomous driving industry. It is jointly organized by China Automotive Engineering Society, Suzhou Automotive Research Institute of Tsinghua University, Guoqi (Beijing) Intelligent Network Automotive Research Institute Co., Ltd., Beijing Economic and Technological Development District Management Committee.
The 6th CICV conference focuses on key technologies and policies and regulations for autonomous driving and discusses in depth about: environmental awareness, decision and control, V2X (vehicle wireless communication technology), artificial intelligence, information security, new vehicle high-speed network, high-precision maps, intelligent transportation and smart cities, human-machine driving, and HMI (Human Machine Interface).