I have always felt that the Internet of Vehicles, intelligent driving, and unmanned driving described in the technology house mouth can really be realized on the road.
Wu Hequan, an academician of the Chinese Academy of Engineering, recently conducted an analysis on the Internet of Things in Wuxi, which made reporters rethink the Internet of Vehicles.
He believes that the Internet of Vehicles will be one of the biggest driving forces of the Internet of Things, and the Internet of Things is redefining the car. A good car will have hundreds of sensors. It is predicted that the global Internet of Vehicles is expected to exceed 100 billion euros in 2020, and China will account for 1/3. However, where is a good entry point for the Internet of Vehicles market that the communications industry has always wanted to enter?
At the "Intelligent Networked V2X Technology Summit Forum" hosted by Datang Telecom Group a few days ago, Chen Xiaohua, executive vice president of BAIC Group's New Technology Research Institute, gave a glimpse of the true market. He believes that the V2X solution proposed by the communications industry is It is a supplement to the existing sensor system and can play six roles in the intelligent networked car.
Existing sensors for smart cars have shortcomings
First of all, I need to explain that intelligent networked and connected vehicles mean the same thing, but the big difference is that the main body is different. "Intelligent networked cars" talk about networking from the perspective of car companies, while Internet of vehicles is discussed from the perspective of the Internet. Joint vehicle. The subject is different and the perspective is brand new.
From the perspective of car companies, IT companies and communication companies are currently providing a new generation of information technology for the automotive industry, and the technologies of these two types of companies will be integrated into the car companies’ vehicle solutions to help car companies realize service transformation Upgrades, such as self-driving car services; energy-saving and new energy vehicles, smart cars; different levels of self-driving or intelligent networked applications.
Chen Xiaohua said that sensor-based smart car solutions are generally divided into: long-distance recognition, recognition of people, objects, and signal signs, middle-distance recognition and short-distance recognition. The main sensors include cameras, radar, lidar, and GPS boxes. However, the function and performance of the sensor when the vehicle is fast and violent will be very limited. For example, a highly equipped self-driving car launched a few days ago is equipped with 12 cameras and 8 high-speed radars. However, in the test, it was found that the coverage, recognition rate, recognition accuracy of these sensors on the surrounding environment detection, and the efficient analysis There is a certain degree of deficiency in the maturity of the fusion algorithm of the conditions around the car.
"In extreme weather, there are also very big problems. In heavy rain and heavy snow, some sensors' performance indicators such as target recognition also drop significantly. Moreover, visual sensors still have blind spots in some cases, such as when overtaking a large vehicle, if the vehicle is in front of it. There is also a car, the visual sensor is generally unable to see this situation, which will bring great danger; for example, at an intersection, if there are buildings, trees, etc., it will also affect the accuracy of the visual sensor; In addition, there are often misjudgments when judging the sideways of a large box truck, which leads to the lack of safety of existing sensors." Chen Xiaohua said.
In addition, high-end sensors such as high-definition cameras and high-speed radars for automobiles are not produced in China, and car companies will also encounter difficulties in procurement.
Chen Xiaohua believes that the intelligent networked V2X technology can make up for these shortcomings to a certain extent.
The role of intelligent networked V2X technology
At present, communication-based V2X is slowly being embedded in the entire vehicle industry chain and value chain. Communication technology is one of the key enabling technologies for intelligent networked vehicles. The communication technology for intelligent networked vehicles related applications has also made milestone progress. In September 2016, 3GPP completed the first version of LTE-V2V communication technology. Standard formulation. But what are the needs of vehicle manufacturers for V2X? Chen Xiaohua believes that clarifying this issue is very important for targeted research and development. He believes that V2X can have six roles in smart cars.
One is the role of pipeline. V2X is the communication channel between the car body and the external environment, including the information interaction between the car and the car, between the car and the person, between the car and the infrastructure, and between the car and the background. V2X can also analyze and process the transmitted data, and follow-up Hope to be able to make use of these data. For example, some information on the car terminal can be customized and pushed to the driver.
The second is the security role. On the one hand, it makes up for the defect that the sensor safety function is easy to fail, and on the other hand, it directly provides safety information for the car and the driver. Embedding V2X in the car can play a role in some safety application scenarios, such as forward collision warning, intersection warning, vehicle blind spot warning, lane change warning, emergency brake warning, left turn assist, vehicle out of control warning, red light warning, etc.; Some application scenarios that are not related to safety, such as news, weather, congestion guidance, community information prompts, priority vehicle traffic, car near-field payment, etc., can also play a role. It should be said that the current V2X equipment is mainly used as a supplement to the safety functions and performance of intelligent networked vehicles, and cannot be a substitute. If V2X is to replace the existing sensors, auto companies will not accept this view. Chen Xiaohua emphasized this point.
The third is to enhance the role of transportation efficiency. V2X can play a role in vehicle-road coordination, green wave belts on the side of traffic roads, traffic guidance, surrounding dynamic parking, and traffic announcements.
The fourth is the role of wisdom. At the Wuxi Internet of Things Conference, the most mentioned are smart communities and smart cities. The main node is smart cars. If smart driving is combined with V2X, it can have a more comprehensive perception of the outside, share car and road information, and deep learning changes. Get smarter. Vehicles and transportation infrastructure can communicate with each other (V2I) to achieve vehicle-road coordination, traffic scheduling accuracy can be higher, and traffic lights can even be cancelled.
The fifth is to provide driving care. This year, automobile manufacturers, including BAIC, are using new technologies to do research and development to provide drivers with vehicle information care based on intelligent network connection. These cares are mainly reflected in driving, such as information reminders, including traffic jams, weather conditions, traffic control, health conditions, etc. The care for the driver’s health is based on the analysis of the driver’s fatigue, using the information interaction between the car and the cloud to analyze, allowing the cloud to make prompts or provide solutions; it can also be used for safety warnings and crossroads warnings. , The vehicle ahead emergency brake reminder, overtaking reminder, surrounding abnormal alarm, etc.
Sixth is the role of the community. V2X technology can provide a link between the community of car users with the same hobbies and the same car model in the background, helping community car users to understand the dynamic information of the community in real time.
According to Chen Xiaohua, BAIC’s overall development strategy for the development of V2X technology is to start with the definition of application scenarios, and then to define intelligent interconnection functions, scenario logic strategies, find efficient development models, technology and integration verification, and formulate enterprise-level application specifications and standards. Then to the realization of smart interconnected products. "We have been testing related scenarios since July. We will do system stability testing throughout 2017, and we will do a comprehensive deployment on our products in 2018. This year, we plan to complete the testing of 10 scenarios on V2X. Next year, 22 V2X application scenarios will be completed, which is a few more than the 17 V2X scenarios defined in the United States."
Challenges and confusions of independent research and development car companies
Chen Xiaohua said that at present, there are some challenges in the development of intelligent network connection technology. The first is that there is no unified goal, and the customer perception of intelligent network technology is relatively independent, and systematic collaboration has not been formed in the entire vehicle product; the second is that when the intelligent network technology is developed, the function performance of the subsystem is overemphasized and the application scenarios are ignored. The third is that the system cost does not match the ratio of the total price of the vehicle. Because a large number of core components are imported, the system cost is uncontrollable; the fourth is that the intelligent networked subsystem must meet the reliability of the entire vehicle. And long-term performance verification can become mature, which is currently severely lacking; fifth is the lack of systematic product innovation and technological breakthroughs; sixth is the lack of uniformity in the evaluation of intelligent network technology, integration in the vehicle, and the application of the vehicle. Regulations, standards and regulations.
In the development of V2X technology, car companies also encountered some difficulties. The first is the high R&D cost and the high investment pressure. The car companies want to establish a research team for the Internet of Vehicles, but the IT industry is two to three times higher than the automotive industry. The second is the weak integration and verification capabilities and the lack of talents. The third is the lack of completeness. Technical system, less related cooperation, need to build a new technology-oriented R&D, application, and testing system; fourth, advanced R&D equipment and integration schemes are monopolized by foreign countries, and core algorithms are missing; fifth, lack of application guidance and incomplete supporting facilities; sixth, R&D And the demonstration period is long.
Chen Xiaohua said that from an industry perspective, when independent car companies develop intelligent network technology, domestic companies have the right to develop V2X technology first among independent car companies, but if car companies install V2X into cars, they must consider the industrial chain. Whether it is complete, supply capacity, quality control level, market position, talent team, profit model, etc.
